CHILDREN AND FAMILIES
EDUCATION AND THE ARTS
The RAND Corporation is a nonprofit institution that helps improve policy and
decisionmaking through research and analysis.
ENERGY AND ENVIRONMENT
HEALTH AND HEALTH CARE
INFRASTRUCTURE AND
TRANSPORTATION
This electronic document was made available from www.rand.org as a public service
of the RAND Corporation.
INTERNATIONAL AFFAIRS
LAW AND BUSINESS
Skip all front matter: Jump to Page 16
NATIONAL SECURITY
POPULATION AND AGING
PUBLIC SAFETY
SCIENCE AND TECHNOLOGY
TERRORISM AND
HOMELAND SECURITY
Support RAND
Browse Reports & Bookstore
Make a charitable contribution
For More Information
Visit RAND at www.rand.org
Explore the RAND Corporation
View document details
Limited Electronic Distribution Rights
This document and trademark(s) contained herein are protected by law as indicated in a notice appearing
later in this work. This electronic representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of RAND electronic documents to a non-RAND website is
prohibited. RAND electronic documents are protected under copyright law. Permission is required from
RAND to reproduce, or reuse in another form, any of our research documents for commercial use. For
information on reprint and linking permissions, please see RAND Permissions.
This report is part of the RAND Corporation research report series. RAND reports
present research findings and objective analysis that address the challenges facing the
public and private sectors. All RAND reports undergo rigorous peer review to ensure
high standards for research quality and objectivity.
Research Report
Redirecting Innovation in U.S. Health Care:
Options to Decrease Spending and
Increase Value
Case Studies
Steven Garber, Susan M. Gates, Emmett B. Keeler, Mary E. Vaiana, Andrew W.
Mulcahy, Christopher Lau, Arthur L. Kellermann
RAND Health
RR-308
RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.
is a registered trademark.
Contents
Figures, Boxes, and Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Case Study 1
Avastin for Metastatic Breast Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Antiangiogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Initial FDA Approvals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
FDA’s Emerging Regulatory Dilemma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Accelerated Approval. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
The FDA Reconsiders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Public Backlash. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Perverse Economic Incentives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Benefits and Risks of Accelerated Approval. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Use of Surrogate Endpoints as the Basis for Drug Approval. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Hope or False Hope?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Mandated Payment for “Off-Label” Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Physicians Are Rewarded for Using the Most Costly Drug. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Once Approval Is Granted, the FDA Has Limited Power to Discourage Use of a Low-Value
Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Cost and Health Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
The Regulator’s Dilemma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Case Study 2
A Cardiovascular Polypill. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
No Consensus About the Pill’s Components, Target Population, or Commercial Viability. . . . . . . . . 15
iii
iv
Redirecting Innovation in U.S. Health Care: Case Studies
Costly Phase III Clinical Trials Required for Regulatory Approval. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Proponents of the Polypill See Large Potential Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
There Is Little or No Market for a Cost-Saving Pill. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Cost and Health Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Small-Scale Trials Have Produced Promising but Inconclusive Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Development Block. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Case Study 3
Electronic Health Records. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Studies Identify Potential Value. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
HITECH Provides Financial Incentives for Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Multiple Barriers to Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
HITECH Promotes Adoption of Existing Systems, Not Better Ones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Cost and Health Impact .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Clinical Impact of EHRs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Economic Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Societal Benefits of EHRs.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Barriers to Success. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Two Platforms, Two Approaches: VistA and Epic .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
VistA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Epic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
The EHR “Productivity Paradox” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Case Study 4
Haemophilus influenzae Type b (Hib) Vaccine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Infants Are Most Vulnerable to Hib Disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Initial Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
These First-Generation Vaccines Represented a Major Advance But Had Important
Limitations.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Expansion of Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Initial Studies Produce Mixed Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Despite Its Limitations, Use of the First-Generation Vaccine Was Recommended. . . . . . . . . . . . . . . . . . . 50
Breakthrough: The World’s First Hib Conjugate Vaccine Is Licensed in 1987. . . . . . . . . . . . . . . . . . . . . . . 50
Currently Available Hib Vaccines.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Cost and Health Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Hib Vaccines Have Proven to Be Highly Effective in Developed World. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Contents
v
Hib Conjugate Vaccines Are Also Effective in Developing Countries, But Their High Cost
and Complex Administration Schedule Limits Their Reach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
In High-Income Countries, the Benefits of the Vaccine Outweigh its Costs. . . . . . . . . . . . . . . . . . . . . . . . . . 55
In the Developing World, Cost Is a Barrier to Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
A Comprehensive Vaccination Campaign Could Potentially Eradicate Hib Worldwide. . . . . . . . . . . . 56
A Technology “Home Run”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
The Best of the Best.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Case Study 5
Implantable Cardioverter-Defibrillator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Secondary Prevention of Sudden Cardiac Death. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Primary Prevention of SCD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Not All Groups Benefit from Receiving ICDs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Limitations of Clinical Trials.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Medicare Expands Coverage to Include Primary Prevention.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Medicare Approves Further Expansion but Specifies “Coverage with Evidence Development” .. . . . 72
ICDs Are Sometimes Implanted in Patients Who Are Unlikely to Benefit.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Health and Cost Impact .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Inappropriate Shocks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Surgical Complications.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Infection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Lead Failures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Are ICDs Cost-Effective? .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
“Treatment Creep”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Case Study 6
Prostate-Specific Antigen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
PSA Testing First Used to Monitor Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
FDA Approval for Screening Use Followed.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Professional Organizations Provide Strong Support. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Corporate and Nonprofit Lobbying Promotes Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Cost and Health Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Although Watchful Waiting Is an Acceptable Option, Few Men Choose It. . . . . . . . . . . . . . . . . . . . . . . . . . 86
For Most Patients, PSA Screening Does More Harm Than Good.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
vi
Redirecting Innovation in U.S. Health Care: Case Studies
As a Result of This Evidence, PSA Screening Is No Longer Recommended by the United
States Preventive Services Task Force.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Despite the USPSTF’s Recommendation, Medicare Is Legally Obliged to Pay for PSA
Screening.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Unintended Consequences.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Case Study 7
Robotic Surgery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
The Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Genesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
The Da Vinci Surgical System Dominates the Current Market.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Early Adoption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Robotic Surgery Expanded Rapidly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Robotic Assistance Increases the Cost of a Typical Surgical Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
The Impact of Adoption Is Most Dramatic in Prostatectomy Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Growing Reliance on Surgical Treatment of Prostate Cancer Is Linked to the Growing Use
of Robotic Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Acquisition of Robots Has Boosted Radical Prostatectomy Rates in Hospitals That Get Them.. . . 98
Hospitals with Robots Attract Cases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
After 2005, Robots Triggered Market Consolidation and Intensification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Robots Also Influenced Surgical Training and Practice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Robot-Assisted Prostatectomy Affects Diagnostic Procedures, Disparities, and Inappropriate
Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Impact on Health and Costs.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Hospitals Cite Robotic Surgery in Marketing.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
A Medical “Arms Race”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Case Study 8
Telemedicine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Concept. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Store-and-Forward (Asynchronous) Mode.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Real-Time (Synchronous) Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Development.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Early Adoption .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Subsequent Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Expansion to Clinical and Academic Environments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Affordable Care Act Encourages Telemedicine to Expand Access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Advances in Telecommunications and Computer Technology Facilitate Adoption. . . . . . . . . . . . . . . . .
Barriers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Regulatory Barriers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Financial Barriers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
109
109
109
110
110
110
112
113
113
114
114
114
114
115
Contents
vii
Provider Reluctance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Medico-Legal Concerns.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Uncertain Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cost and Health Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adoption Block. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
116
116
116
117
119
119
Figures, Boxes, and Tables
Figures
5.1.
6.1.
6.2.
7.1.
An ICD Implanted. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Prostate Cancer Incidence Trends (1975–2006) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Benefits and Harms of PSA Screening for Prostate Cancer.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Mean Change in HRR Number of Radical Prostatectomy Procedures Following
Robot Acquisition, Controlling for Baseline HRR Volume and Herfindahl Index .. . . . . . 98
7.2. Trends in Radical Prostatectomy Procedure Volume and Number of Hospitals
Performing the Procedure in New York, New Jersey, and Pennsylvania. . . . . . . . . . . . . . . . . . . 100
Boxes
5.1. Indications for ICD Therapy in the Prevention of SCD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.2. Class I Recommendations for ICD Therapy as Secondary Prevention in Patients
with Heart Failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.3. Class I Recommendations for ICD Therapy as Primary Prevention in Patients with
Heart Failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Tables
3.1. Health IT and Consumer IT Comparisons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.1. Selected Efficacy Studies of the Hib Vaccine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2. Currently Licensed Hib Vaccines in the United States.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
ix
Case Study 1
Avastin for Metastatic Breast Cancer
Benjamin R. Roman, M.D.
Fellow, Robert Wood Johnson Clinical Scholars Program, University of Pennsylvania;
Fellow, Head and Neck Oncologic Surgery, Memorial Sloan Kettering Cancer Center
The Technology
Bevacizumab (brand name: Avastin) is a recombinant humanized monoclonal antibody that
binds to human vascular endothelial growth factor (VEGF), preventing its interaction with
receptors on the surface of endothelial cells. Its effect is to stop angiogenesis, the development
of new blood vessels—a process essential for cancer growth. Developed by Roche/Genentech,
Avastin is widely used as an adjunctive therapy to traditional chemotherapies in the treatment
of several types of cancers.
Rationale
There are approximately 230,000 new cases of invasive breast cancer in the United States
annually and 40,000 breast cancer-related deaths.1 Approximately 29,000 newly diagnosed
patients are already at stage IV at presentation, meaning that they have distant metastases. This
group has a 15-percent five-year survival rate, based on National Cancer Database data from
2000 to 2001.2 As is true with many cancer types that present with distant metastases, few
therapies have been identified that significantly improve survival once breast cancer reaches
this stage. A cure is desperately sought.
Antiangiogenesis
In 1971, Judah Folkman of Harvard University hypothesized that angiogenesis was an important aspect of cancer biology and might be a target for future drug therapy.3 In the ensuing decades, with support from the National Institutes of Health’s (NIH’s) National Cancer
Institute (NCI) and EntreMed, a biotechnology company in Rockville, Md., he began testing
antiangiogenesis compounds in the laboratory setting. By the late 1990s, there was gathering excitement among researchers and the media about the promise of newer compounds for
cancer coming from his lab. Citing developments in Folkman’s research, Richard Klausner,
then the director of NCI, declared in 1998, “I am putting nothing on higher priority than getting this into clinical trials,” and Nobel laureate James D. Watson claimed that “Judah is going
to cure cancer in two years.’’4
1
2
Redirecting Innovation in U.S. Health Care: Case Studies
Development
At the same time that the NCI was announcing Folkman’s breakthrough, a number of biotechnology companies and researchers were also pursuing antiangiogenesis drugs, albeit with
less media attention. However, only Genentech had a drug that targeted a specific growth
factor for angiogenesis: In 1989, a Genentech scientist named Napoleone Ferrara identified and
cloned VEGF, the first specific growth factor to be described that promoted new blood vessel
formation.5 In 1993, his team demonstrated that an antibody against VEGF could slow angiogenesis and tumor growth in a mouse model.6 Thus, while attention was focused on Folkman’s
work, Genentech was quietly making progress. In 1995, it filed a patent for bevacizumab,
an antibody against VEGF. By 1997, the first phase I clinical trial of the drug that would be
brand-named Avastin began recruiting subjects.7
From the outset, it was clear that Genentech believed Avastin was a potential blockbuster. The company not only spent hundreds of millions of dollars in research on angiogenesis to develop the drug; it decided to build a plant to manufacture Avastin before it received
U.S. Food and Drug Administration (FDA) approval to market the drug in 2004.8 Priced at
more than $4,000 a month (about $50,000 a year for metastatic colon cancer),9 Avastin generated sales of approximately $1.4 billion in its first full year after approval in the United States.10
Early sales and the promise for future growth for Avastin’s market attracted investors; between
2003 and 2005, shares of Genentech quadrupled.
In 2006 and in 2009, Avastin won FDA approval to use Avastin for some types of lung
and breast cancer. The price of treating lung or breast cancer with Avastin is about twice as
high as treatment for colon cancer because the drug is given at a higher dose for these cancers.
The higher price point and growing use for multiple indications helped drive global sales as
high as $6.2 billion in 2010.11 When the FDA subsequently revoked of approval of Avastin for
breast cancer, sales dropped to $6.0 billion in 2011, but they rebounded to $6.1 billion in 2012,
in part because of Avastin’s continued approval for use in metastatic breast cancer in Europe
and Japan, and in part because of its expanded use to treat other forms of cancer.12, 13
Genentech’s research and financial successes attracted the attention of the giant Swiss
global health care company F. Hoffmann-La Roche Ltd. (Roche), which in 1990 paid $2.1 billion to acquire a 56-percent share of Genentech. Executives at Roche explained their decision
at the time by noting that Genentech’s pipeline was “the future of [Roche].”14 By 2009, with
Roche’s income decreasing and Genentech’s net income skyrocketing due in large part to Avastin, Roche paid an eye-popping $46.8 billion to make Genentech a wholly owned subsidiary
group.49
Early Adoption
Initial FDA Approvals
The first FDA-approved use for Avastin in oncology came in 2004 after a successful phase III
trial in metastatic colorectal cancer funded by Genentech. The addition of Avastin to chemotherapy as first-line treatment for metastatic colorectal cancer was shown to improve overall
survival by about five months.15 Subsequently, additional FDA approvals were won for the
second-line treatment of metastatic colorectal cancer (2006); first-line unresectable, recurrent,
Case Study 1: Avastin for Metastatic Breast Cancer
3
or metastatic non–small cell lung cancer (2006); second-line glioblastoma (2009); metastatic
renal cell carcinoma (2009); and, later, other forms of cancer.16
FDA’s Emerging Regulatory Dilemma
FDA approval of the use of Avastin in metastatic breast cancer was less straightforward than
for other cancers. A clinical trial published in 2007 became the basis of FDA approval in early
2008. But by 2011, findings from subsequent trials prompted the FDA to withdraw its previous approval.
E2100, the clinical trial that was basis for the initial approval, was published in 2007.
Funded by NCI, it was an open-label, randomized trial conducted by the Eastern Cooperative Oncology Group.17, 18 Patients with newly diagnosed metastatic breast cancer who had not
previously received cytotoxic drugs for metastases were eligible to participate. Patients with
human epidermal growth factor receptor type 2–positive (HER2-positive) tumors, a marker
for aggressive cancer, could only be included in the trial if they had previously received trastuzumab (Herceptin), Genentech’s antibody that specifically targets HER2. Patients were randomized to receive Avastin plus paclitaxel, versus paclitaxel alone.
Patients randomized to the Avastin group experienced an improvement in progressionfree survival (PFS) of 5.5 months compared to those in the paclitaxel-alone group. This means
that during this period, the disease did not seem to spread. However, Avastin-treated patients
did not live any longer. In late 2007, the FDA’s Oncologic Drugs Advisory Committee (ODAC)
met to debate whether the risk-benefit ratio of this trial supported making a recommendation
to the FDA to approve Avastin.
Ralph D’Agostino, a voting member of ODAC, highlighted several concerns with using
E2100 to support approval.19 First, E2100 was not designed to measure the outcomes on which
FDA approval is usually based. Instead of overall survival, the trial was designed to measure
PFS as its primary endpoint; the study made the more traditional outcomes measures, such as
response rate, quality of life, and overall survival, secondary outcomes. Second, quality of life
data were inconclusive, since the study was open-label and information on use of concurrent
medications was not collected. Third, safety and toxicity data were not collected in the control
group, making comparisons to the Avastin (treatment) group impossible. After considering
these concerns, the ODAC voted 5–4 against approval, based on the lack of improvement of
overall survival and uncertainty about other effects of the drug.
Accelerated Approval
Notwithstanding the ODAC’s concerns, the FDA granted accelerated approval to the drug in
early 2008. The stipulation of “accelerated approval” meant that Genentech had to conduct
additional studies to directly address the issue of overall survival. Only then would the agency
grant standard approval.
The FDA Reconsiders
In June 2010, the ODAC met to consider additional evidence on the efficacy of Avastin as a
treatment for metastatic breast cancer. Their primary reference point was the phase III Avastin
and Docetaxel (AVADO) trial, which randomized patients to receive either docetaxel alone or
in combination with different doses of Avastin. Although there were some signs of benefit from
Avastin, such as tumor response, PFS improved by less than one month. As with the earlier
4
Redirecting Innovation in U.S. Health Care: Case Studies
trial, there was no difference in overall survival. Even more concerning, patients randomized
to the Avastin group had a higher incidence of serious adverse events.20, 21
Another phase III study, the Regimens in Bevacizumab for Breast Oncology-1 (RIBBON1), showed similarly negative results: When Avastin was added to conventional chemotherapy,
improvement in PFS was less than reported in the originally favorable trial, E2100; as with the
other two studies, there was no improvement in overall survival. And like the AVADO trial,
RIBBON-1 identified more serious adverse events in the Avastin group.22
With these new data in hand, the ODAC voted 12–1 to revoke FDA approval for Avastin
in metastatic breast cancer.23 Genentech formally protested, citing four reasons why Avastin
should remain approved: a lack of precedent for revocation of accelerated approval, the possibility that the drug might benefit subgroups, the importance of respecting individual choice
in health care, and the possibility that the FDA’s decision would discourage innovation.24 After
considering Genentech’s objections, in November 2011 the FDA revoked approval.
In her final decision memo, FDA commissioner Margaret Hamburg argued that data did
not support Avastin’s efficacy or safety for metastatic breast cancer. She also noted the lack of
clinically meaningful benefit: No demonstrable improvement in survival had been reported in
any of the studies, and an average improvement in PFS of a few months was not thought to be
sufficiently meaningful to warrant the marketing of such a costly drug. Hamburg noted that
Avastin did not improve quality of life; indeed, its use was associated with an increased risk of
serious adverse events, including gastrointestinal perforations, wound healing complications,
and hemorrhage. Taken together, she concluded, these findings suggested that Avastin did
more harm than good for patients with metastatic breast cancer.25
In its decision to revoke Avastin’s accelerated approval, the FDA noted that Genentech
was free to pursue additional studies that might cause the FDA to grant new approval. Several
authors and Genentech itself have noted that Avastin may have a role in treatment of metastatic
breast cancer if biomarkers can be developed that identify a subgroup of patients who respond
well to the drug or a combination of agents that boost Avastin’s beneficial effects without causing added harm.
Public Backlash
Despite strong evidence supporting the FDA’s action, many oncologists and cancer patients
were outraged. All were convinced that Avastin appeared to be holding the cancers of some
women at bay. Advocates for these women felt that the “FDA’s blocking of compassionate
access”26 was “nothing short of a death sentence.”27 Even the deputy chief medical officer of
the American Cancer Society, J. Leonard Lichtenfeld, disagreed with the FDA decision, writing that he hoped insurance companies would still reimburse for Avastin for women “who are
currently on the drug and who are showing a benefit from its use.”26 Opponents of health care
reform, such as Sen. David Vitter (R-La.), argued that the FDA’s revocation of its earlier accelerated approval was a prime example of the government “rationing access [to health care].”27
Subsequent Use
The story of Avastin’s use for metastatic breast cancer underscores how policies influence a
technology’s adoption, uptake, and subsequent use. From the regulator’s perspective, the story
graphically illustrates the peril of using “accelerated approval” as a mechanism for getting a
Case Study 1: Avastin for Metastatic Breast Cancer
5
high-priority drug, such as a cancer drug, to market. The FDA’s policy parallels in many respects
the Centers for Medicare and Medicaid Services’ (CMS’s) “coverage with evidence development” policy for reimbursing promising but unproven technologies. Accelerated approval policies seek to speed the translation of promising research into practice, but they do so at the risk
of raising false hopes about a technology’s benefits before they are clearly proven. Moreover, the
FDA did not anticipate how difficult it would be to remove a product from the market or even
restrict its use once it had developed a constituency among physicians and patients.
Perverse Economic Incentives
From the payers’ perspective, the Avastin story highlights several perverse incentives embedded in reimbursement policy for oncologic drugs, most notably that Medicare must reimburse
even off-label (unapproved) use of drugs if they are listed in one of several drug compendia.
Furthermore, the way oncologists are reimbursed (at a fixed percentage of the drug’s price)
creates a perverse incentive to administer the most expensive drug for any indication in order
to maximize practice revenue. This is because how much physicians are paid is calculated as a
percentage of the drug’s price, not the difficulty of its administration. So even though the FDA
revoked approval of Avastin for metastatic breast cancer in 2011, many doctors still use it to
treat breast cancer patients, and Medicare still pays for its use.
Benefits and Risks of Accelerated Approval
The FDA created the option of granting “accelerated approval” based on surrogate endpoints
to enable early approval of particularly important drugs and to add a measure of flexibility to
the approval process. The accelerated approval mechanism was added to new drug application
regulations in 1992, largely in response to pressure on the FDA to facilitate rapid deployment
of promising treatments for HIV/AIDS.
Accelerated approval allows provisional approval of drugs that treat serious or lifethreatening disease if they appear to provide benefit compared to other therapies. Benefit can
be measured as an impact on a surrogate endpoint that is likely to predict clinical benefit. The
applicant is obliged to perform confirmatory studies with due diligence to confirm the clinical
benefit, at which point the drug can be converted to full approval if warranted by the evidence.
If confirmatory studies are not performed with due diligence or the benefit is not subsequently
proven, the FDA can revoke its accelerated approval.24 This is precisely what happened with
FDA approval for Avastin to treat metastatic breast cancer.
The most recent review of accelerated approval in oncology by the FDA was published in
2011, before the final ruling on the fate of Avastin. At the time, 35 products had been granted
accelerated approval for 47 indications in oncology, and only three had failed to prove some
degree of benefit in confirmatory trials. Of the three, two were removed from the market, and
one was allowed to be implemented with a limited distribution plan.28
Avastin for metastatic breast cancer became the fourth product to fail on confirmatory
studies and the third to have its FDA approval revoked. Daniel Carpenter has argued that the
decision to revoke affirmed the FDA’s credibility and established an important precedent that
was, in his view, as important as the decision itself.24 Only by maintaining its independence
and relying on available data in the face of philosophical and political criticism could the FDA
affirm the integrity of its accelerated approval mechanism. However, the decision carried a
political price—public criticism of the FDA for “rationing” care that was highly desired by
patients who were convinced that it worked, irrespective of the data.
6
Redirecting Innovation in U.S. Health Care: Case Studies
Use of Surrogate Endpoints as the Basis for Drug Approval
The Avastin story also illustrates the peril of relying on a surrogate endpoint in granting accelerated approval. The FDA is required by law to approve only drugs that show evidence of
effectiveness. This requirement is based on a 1962 amendment to the Federal Food Drug and
Cosmetic Act of 1938. After the amendment was enacted, effectiveness was defined in judicial
rulings as having either a clinically meaningful benefit, such as improved survival, function, or
symptoms, or an increased level of an established “surrogate endpoint.”30–32
A surrogate endpoint is a laboratory measure or physical sign used as a substitute for a
more clinically meaningful endpoint—for example, using average blood glucose control rather
than survival to measure the effectiveness of a new treatment for diabetes. If a drug moves a
surrogate endpoint in a favorable direction, it is assumed that this will translate into improved
clinical outcomes.
Reasons to use surrogate endpoints, especially in trials treating metastatic disease, include
the fact that patients tend to live longer than the duration of the trial and often receive many
other types of chemotherapy, confounding measurement of a particular drug’s impact on overall survival.33 Surrogate endpoints in oncology over the past 30 years have included objective
tumor response rates, disease-free survival, time to progression (TTP), and PFS. PFS, often
used interchangeably with TTP, differs from the former because it is concerned only with progression rather than with both progression and death.32
Despite a negative vote from the ODAC, the FDA based its accelerated approval of Avastin for metastatic breast cancer on the manufacturer’s use of PFS as its primary surrogate endpoint in the E2100 trial. Following the initial ODAC hearing on E2100, D’Agostino pointed
out that use of PFS as an endpoint is fraught with problems. It not only requires blinding
of radiologic evaluations, but its measurement and documentation depend on the timing of
evaluations, which can be influenced by uneven assessment in different groups, missed assessments, and incomplete baseline measurements.34
More difficult, perhaps, is the fact that some oncologists, patients, and advocates embrace
such surrogate endpoints as PFS as meaningful outcomes themselves, regardless of the drug’s
effect on actual survival. Richard Pazdur, the FDA’s director of the Office of Hematology
Oncology Products, which is responsible for approval decisions, has been quoted as saying that
delaying progression of disease “may be a direct clinical benefit in itself.”35 European oncologists are also inclined to believe that PFS is an important standalone endpoint, although it
is difficult to support this view without concomitant evidence that a drug improves patients’
quality of life.19
Use of surrogate endpoints in oncology is hotly debated. Recently, a new endpoint, pathologic complete response, has been proposed in the neoadjuvant setting of nonmetastatic breast
cancer treatment. This refers to locally advanced cancers before surgical treatment. Two trials
with Avastin using this surrogate marker have shown promise,36, 37 and the FDA has agreed
that the marker is “reasonably likely to predict clinical benefit” and therefore may be used. At
the same time, however, the FDA has cautioned that the definition of the endpoint needs to
be standardized, appropriate cancer subtypes and subgroups need to be refined, and the magnitude of surrogate improvement that correlates with overall survival must be determined.38, 39
Hope or False Hope?
The FDA’s accelerated approval policy and the companion CMS policy, “coverage with evidence development,” reflect the agencies’ genuine desire to provide access to promising new
Case Study 1: Avastin for Metastatic Breast Cancer
7
technologies while remaining flexible with their subsequent approval and coverage decisions
based on later scientific evidence. The policies are also similar in that they risk raising false hope
that a technology is beneficial before benefit is clearly proven. The example of Avastin demonstrates that providing provisional approval to a drug may make it politically difficult to rescind
approval later based on equivocal or contradictory evidence. The outcry from patients and physicians following the FDA’s revocation of Avastin’s approval for metastatic breast cancer suggests that the political costs of revoking approval, and dashing the hopes that approval raises,
may be more problematic for the agency than requiring stronger evidence before allowing a
drug to be brought to market.
The distinction between offering hope and false hope was cited by Mikkael Sekeres, a
member of the ODAC, as a core reason for the committee’s near-unanimous vote to rescind
approval of Avastin. He argued that offering Avastin for metastatic breast cancer in the face of
evidence that it is ineffective or even harmful amounts to providing false hope. Commenting
on the danger of false hope, bioethicist Adrienne Martin has suggested that pharmaceutical
companies can exploit the desperation of cancer patients by bolstering unreasonable hope of
a highly unlikely outcome. Specifically citing Avastin for metastatic breast cancer, she noted
that “fear of death, fear of cancer, hope for cure . . . all of these and more make cancer patients
overvalue the promise of five additional months free of disease progression.”40
Mandated Payment for “Off-Label” Use
Shortly before the FDA revoked its accelerated approval of Avastin for metastatic breast cancer,
the National Comprehensive Cancer Network (NCCN) added Avastin to its Compendium
of Drugs and Biologics for this indication. This decision meant that even if FDA approval
were subsequently revoked, CMS would remain legally bound to reimburse use of Avastin for
metastatic breast cancer, including paying medical oncologists their drug administration fee.
Put another way, revoking FDA approval for metastatic breast cancer did not remove the drug
from the market; it simply meant that Genentech could no longer promote its use for metastatic breast cancer. Medical oncologists who wanted to keep administering it for this condition could do so, and CMS would pay.
CMS reimbursement of off-label use of cancer drugs is far more common than people
realize. Indeed, the NCCN estimates that 50–75 percent of all cancer drugs are administered
off-label.41 This peculiar approach to reimbursement is written into federal law. The Social
Security Act of 1993 required that Medicare reimburse all anticancer drugs and biologics listed
in certain compendia, such as the NCCN. In 2008, five such compendia met CMS criteria.42
In addition, nearly three-quarters of the U.S. population live in states that require private
insurers to cover certain cancer drugs. Laws in most of these states mandate that plans must
pay if a prescribed drug is listed in a recognized compendium, irrespective of FDA approval.
Some states mandate insurance coverage if a drug is simply mentioned in the peer-reviewed
literature.43
Amy Abernathy has argued that drug compendia such as the NCCN’s are given disproportionate weight in deciding the appropriateness of off-label prescribing. By requiring reimbursement of any drug listed in one of these compendia, Congress and the states are essentially
relying on the organizations that publish them to perform comparative-effectiveness research,
although that is not their purpose and they lack the resources to do it. Furthermore, the subjective processes and lower levels of evidence used by different compendia compromise the valid-
8
Redirecting Innovation in U.S. Health Care: Case Studies
ity of their assessments.44 In the case of Avastin, one-third of NCCN’s breast cancer panel had
financial ties with Genentech. The company also funds the NCCN directly.25
Abernathy has suggested that one way to address the issue of regulatory approval for offlabel use is to require continuous evidence collection and reevaluation of a drug’s use, much as
the FDA and CMS seek to authorize products with provisional approval today. Another potential strategy is to require “pragmatic “ clinical trials that use real-world populations and comparators and measure clinically relevant outcomes, thereby learning from evidence collected
in in actual clinical practice.44 As matters stand, once a compendium adopts a drug, Medicare
and other insurers must pay for its use, irrespective of FDA approval or the subsequent generation of contradictory clinical evidence.
Physicians Are Rewarded for Using the Most Costly Drug
Before 2005, reimbursement for Part B cancer drugs, which are typically physician-dispensed
and are mostly chemotherapeutic agents, was pegged at 95 percent of a drug’s average wholesale price. Manufacturers competed for market share by offering discounts to purchasing physicians so that they could increase the margin between what they paid to purchase the drug
and what they received in reimbursement. Because the prices of most of these drugs were set
quite high, paying the administering physician 95 percent of the product’s listed wholesale
price gave doctors a hefty margin, since manufacturer discounts (often 12–30 percent, but as
high as 86 percent) meant that they were paying far below list price for these drugs.45
In an attempt to reduce these high dispensing margins and rein in Medicare spending, the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 changed
reimbursement for Part B drugs to the manufacturer’s average sales price (ASP) plus 6 percent
and an administrative fee. Although this change decreased Part B drug reimbursements, of
which chemotherapy drugs are the majority, on the short term, it created a perverse incentive
for manufacturers to charge the highest price possible for their drugs, because this meant that
physicians who used them made more money (6 percent of ASP) than if they treated their
patients with less-expensive drugs.45, 46
The ASP plus 6 percent rule quickly altered the behavior of manufacturers. They began
to set high initial launch prices, rather than raising their prices after release as they had before.
This was done because there is a two-quarter lag in adjusting the ASP. Additionally, for compounds that can be dispensed by both physicians and pharmacies, perverse pricing of Part B
drugs created a “spillover” effect on pharmacy prices, since manufacturers are likely to charge
the same price in both settings. Although the 20 percent patient co-pay required from patients
for Part B drugs had the potential to cause sticker shock, this proved not to be a major barrier
to high pricing, since most Medicare patients purchase supplemental Medigap coverage or are
dually eligible for Medicaid. Those who lacked supplemental coverage went without. Finally,
because private insurers often follow Medicare’s lead on payment policy, the ASP plus 6 percent rule quickly drove the entire market in physician-dispensed drugs.46
Once Approval Is Granted, the FDA Has Limited Power to Discourage Use of a Low-Value
Treatment
Securing regulatory approval is a high bar to clear. But once approval is secured and a drug
has the opportunity to establish itself in the minds of doctors and patients, it can be shielded
as well. This is particularly true for cancer drugs. Because CMS and many private insurers are
legally obliged to reimburse the off-label use of Avastin for metastatic breast cancer irrespec-
Case Study 1: Avastin for Metastatic Breast Cancer
9
tive of the FDA’s subsequent revocation of approval, the agency could not stop practitioners
from prescribing the drug for this condition. There is, however, some evidence that the agency’s
decision to revoke approval blunted use. Genentech reported that by the end of 2011, when the
FDA’s final decision to revoke approval was announced, use of Avastin for metastatic breast
cancer declined from 60 percent of eligible patients to 20 percent.27 At an IOM workshop in
2012, United Healthcare reported that by encouraging providers to follow medical evidence,
they cut use of Avastin for this now unapproved condition in half between 2009 and 2012.44
Cost and Health Impact
Despite paying more than $8,000 a month for the treatment, patients who receive Avastin for
metastatic breast cancer do not live longer than patients treated with other chemotherapeutic
regimens. They also have a significantly higher risk of life-threatening complications, such as
gastrointestinal perforations, wound healing complications, and hemorrhage. This is a high
price to pay for one or two additional months of PFS without measurable improvements in
quality of life.
Since the FDA revoked its earlier approval, more specific evidence about the cost effectiveness of Avastin for metastatic breast cancer has been developed. An independent group
performed a cost-effectiveness analysis using data from the three big clinical trials of Avastin for metastatic breast cancer: E2100, AVADO, and RIBBON1. Because Avastin does not
extend life, E2100 data showed an incremental cost of $100,300 to add 0.49 years of PFS,
which equates to 0.135 quality-adjusted life years (QALYs). (QALYS are a measure of disease
burden that includes both the quality and the quantity of life lived.) The calculated incremental cost-effectiveness ratio (ICER) was therefore $745,000 per QALY, an astronomical sum.
RIBBON1 showed that adding Avastin to the chemotherapy agent alone led to an increase
of 0.168 QALYs, at an ICER of $425,000 per QALY. AVADO showed an increase of 0.0375
QALYs, at an ICER of $1,937,000 per QALY.
Thus, by any measure, Avastin is not cost-effective for the treatment of metastatic breast
cancer. Using E2100 data and a threshold of $150,000 per QALY, the price of the drug would
have to be cut by 80 percent, or PFS would have to be boosted by ten months, to make Avastin
reasonably cost-effective compared with other treatment options.47
In a 2006 interview, the company’s president of product development defended the product’s high price, based on “the value of innovation, and the value of new therapies.” Roche/
Genentech’s annual report in 2010, issued in the midst of FDA decisionmaking about Avastin,
further explained the company’s mindset: “ . . . we consider it vital for the future that policymakers and health officials look not only at the costs of new medicines but also at the innovation they embody and the benefits they offer patients . . . medical progress should be viewed
more as a public good that deserves vigorous political support.”11
The Regulator’s Dilemma
Judah Volkman’s discovery did not cure cancer, but the technology it inspired has generated
billions in sales for Genentech and Roche. The FDA’s experience with Avastin illustrates the
difficult balancing act the agency must maintain between those who think it moves too slowly
10
Redirecting Innovation in U.S. Health Care: Case Studies
and those who criticize it for hasty decisions. The FDA granted Avastin accelerated approval
for use in metastatic breast cancer on the basis of a single clinical trial that used a surrogate
endpoint, with the understanding that further studies would be done. By the time that this
additional research showed that Avastin does not increase survival or improve the quality of
patients with metastatic breast cancer, the drug had built a loyal following among oncologists and cancer patients who felt otherwise. Moreover, during the relatively brief time period
in which Avastin held accelerated approval, it secured a spot on a major drug compendium.
Because Avastin is included in the compendium, federal law requires Medicare to keep paying
for doctors to administer this highly expensive and potentially harmful drug for a condition
it does not help, even though the FDA subsequently revoked its approval of Avastin for use in
metastatic breast cancer. Thus, policies and decisions implemented for the best of reasons can
ultimately generate substantial costs without improving human health.
References
1
American Cancer Society, Cancer Facts and Figures 2013, Atlanta, Ga., 2013.
2
American Cancer Society, Breast Cancer Detailed Guide, 2013. As of January 10, 2014:
http://www.cancer.org/acs/groups/cid/documents/webcontent/003090-pdf.pdf
3 Folkman J, “Tumor Angiogenesis: Therapeutic Implications,” New England Journal of Medicine, Vol. 285,
No. 21, November 18, 1971, pp. 1182–1186.
4
Kolata G, “Hope in the Lab: A Special Report; A Cautious Awe Greets Drugs That Eradicate Tumors in
Mice,” New York Times, May 3, 1998.
5
Leung DW, Cachianes G, Kuang WJ, Goeddel DV, and Ferrara N, “Vascular Endothelial Growth Factor Is
a Secreted Angiogenic Mitogen,” Science, Vol. 246, No. 4935, December 8, 1989, pp. 1306–1309.
6
Kim KJ, Li B, Winer J, et al., “Inhibition of Vascular Endothelial Growth Factor-Induced Angiogenesis
Suppresses Tumour Growth in Vivo,” Nature, Vol. 362, No. 6423, April 29, 1993, pp. 841–844.
7
Gordon MS, Margolin K, Talpaz M, et al., “Phase I Safety and Pharmacokinetic Study of Recombinant
Human Anti-Vascular Endothelial Growth Factor in Patients with Advanced Cancer,” Journal of Clinical
Oncology, Vol. 19, No. 3, February 1, 2001, pp. 843–850.
8
Berenson A, “Cancer Drugs Offer Hope, But at a Huge Expense,” New York Times, July 12, 2005.
9 Berenson A, “A Cancer Drug Shows Promise, at a Price That Many Can’t Pay,” New York Times, February 15, 2006.
10
F. Hoffmann-La Roche Ltd., Annual Report 2004, 2005. As of January 10, 2014:
http://www.roche.com/gb04e.pdf
11 F. Hoffmann-La Roche Ltd., Annual Report 2010, 2011. As of January 10, 2014:
http://www.roche.com/gb10e.pdf
12
F. Hoffmann-La Roche Ltd., Annual Report 2011, 2012. As of January 10, 2014:
http://www.roche.com/gb11e.pdf
13 F. Hoffmann-La Roche Ltd., Annual Report 2012, 2013. As of January 10, 2014:
http://www.roche.com/gb12e.pdf
14
Knowledge@Wharton, Anatomy of a Merger: ‘Hostile Deals Become Friendly in the End, Right?’ Wharton
School of the University of Pennsylvania, 2010.
Case Study 1: Avastin for Metastatic Breast Cancer
11
15
Hurwitz H, Fehrenbacher L, Novotny W, et al., “Bevacizumab Plus Irinotecan, Fluorouracil, and Leucovorin for Metastatic Colorectal Cancer,” New England Journal of Medicine, Vol. 350, No. 23, June 3, 2004,
pp. 2335–2342.
16
National Cancer Institute, “FDA Approval for Bevacizumab,” updated July 1, 2013. As of January 10, 2014:
http://www.cancer.gov/cancertopics/druginfo/fda-bevacizumab
17
Miller K, Wang M, Gralow J, et al., “Paclitaxel Plus Bevacizumab Versus Paclitaxel Alone for Metastatic
Breast Cancer,” New England Journal of Medicine, Vol. 357, No. 26, December 27, 2007, pp. 2666–2676.
18
Gray R, Bhattacharya S, Bowden C, Miller K, and Comis RL, “Independent Review of E2100: A Phase III
Trial of Bevacizumab Plus Paclitaxel Versus Paclitaxel in Women with Metastatic Breast Cancer,” Journal of
Clinical Oncology, Vol. 27, No. 30, October 20, 2009, pp. 4966–4972.
19
D’Agostino RB, Sr., “Changing End Points in Breast-Cancer Drug Approval—The Avastin Story,” New
England Journal of Medicine, Vol. 365, No. 2, July 14, 2011, p. e2.
20
Pivot X, Schneeweiss A, Verma S, et al., “Efficacy and Safety of Bevacizumab in Combination with
Docetaxel for the First-Line Treatment of Elderly Patients with Locally Recurrent or Metastatic Breast Cancer:
Results from AVADO,” European Journal of Cancer, Vol. 47, No. 16, November 2011, pp. 2387–2395.
21
Miles DW, Chan A, Dirix LY, et al., “Phase III Study of Bevacizumab Plus Docetaxel Compared with Placebo Plus Docetaxel for the First-Line Treatment of Human Epidermal Growth Factor Receptor 2-Negative
Metastatic Breast Cancer,” Journal of Clinical Oncology, Vol. 28, No. 20, July 10, 2010, pp. 3239–3247.
22
Robert NJ, Dieras V, Glaspy J, et al., “RIBBON-1: Randomized, Double-Blind, Placebo-Controlled,
Phase III Trial of Chemotherapy with or Without Bevacizumab for First-Line Treatment of Human Epidermal Growth Factor Receptor 2-Negative, Locally Recurrent or Metastatic Breast Cancer,” Journal of Clinical
Oncology, Vol. 29, No. 10, April 1, 2011, pp. 1252–1260.
23
Food and Drug Administration Oncologic Drugs Advisory Committee, “Transcript of the July 20, 2010,
Meeting,” 2010. As of January 20, 2014:
http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/
OncologicDrugsAdvisoryCommittee/UCM221998.pdf
24
Carpenter D, Kesselheim AS, and Joffe S, “Reputation and Precedent in the Bevacizumab Decision,” New
England Journal of Medicine, Vol. 365, No. 2, July 14, 2011, p. e3.
25
Food and Drug Administration, “Proposal to Withdraw Approval for the Breast Cancer Indication for
Avastin (Bevacizumab); Decision of the Commissioner,” 2011.
26 Stein
2011.
R, “FDA Revokes Avastin’s Approval for Breast Cancer Treatment,” Washington Post, November 18,
27 Pollack A, “F.D.A. Revokes Approval of Avastin for Use as Breast Cancer Drug,” New York Times, November 18, 2011.
28
Johnson JR, Ning YM, Farrell A, Justice R, Keegan P, and Pazdur R, “Accelerated Approval of Oncology
Products: The Food and Drug Administration Experience,” Journal of the National Cancer Institute, Vol. 103,
No. 8, April 20, 2011, pp. 636–644.
29
Pearson SD and Bach PB, “How Medicare Could Use Comparative Effectiveness Research in Deciding on
New Coverage and Reimbursement,” Health Affairs, Vol. 29, No. 10, October 2010, pp. 1796–1804.
30
Dagher R, Johnson J, Williams G, Keegan P, and Pazdur R, “Accelerated Approval of Oncology Products: A Decade of Experience,” Journal of the National Cancer Institute, Vol. 96, No. 20, October 20, 2004,
pp. 1500–1509.
31
Pazdur R, FDA Memorandum: FDA Background Package for November 8, 2005 ODAC Meeting on Accelerated Approvals, Food and Drug Administration, 2005. As of January 20, 2014:
http://www.fda.gov/downloads/ForConsumers/ByAudience/ForPatientAdvocates/
SpeedingAccesstoImportantNewTherapies/UCM129700.pdf2005
12
Redirecting Innovation in U.S. Health Care: Case Studies
32
McKee AE, Farrell AT, Pazdur R, and Woodcock J, “The Role of the U.S. Food and Drug Administration Review Process: Clinical Trial Endpoints in Oncology,” The Oncologist, Vol. 15, Supplement 1, 2010,
pp. 13–18.
33
Montero AJ, Escobar M, Lopes G, Gluck S, and Vogel C, “Bevacizumab in the Treatment of Metastatic
Breast Cancer: Friend or Foe?” Current Oncology Reports, Vol. 14, No. 1, February 2012, pp. 1–11.
34 Food
and Drug Administration Oncologic Drugs Advisory Committee, “Transcripts of the December 5,
2007 Meeting,” December 2007. As of January 10, 2014:
http://www.fda.gov/ohrms/dockets/ac/cder07.htm#OncologicDrugs
35
Pollack A, “F.D.A. Extends Avastin’s Use to Breast Cancer,” New York Times, February 23, 2008.
36
Bear HD, Tang G, Rastogi P, et al., “Bevacizumab Added to Neoadjuvant Chemotherapy for Breast
Cancer,” New England Journal of Medicine, Vol. 366, No. 4, January 26, 2012, pp. 310–320.
37
von Minckwitz G, Eidtmann H, Rezai M, et al., “Neoadjuvant Chemotherapy and Bevacizumab for
HER2-Negative Breast Cancer,” New England Journal of Medicine, Vol. 366, No. 4, January 26, 2012,
pp. 299–309.
38
Prowell TM and Pazdur R, “Pathological Complete Response and Accelerated Drug Approval in Early
Breast Cancer,” New England Journal of Medicine, Vol. 366, No. 26, June 28, 2012, pp. 2438–2441.
39
Prowell T, Draft Guidance for Industry: Pathologic Complete Response in Neoadjuvant Treatment of High-Risk
Early-Stage Breast Cancer: Use as an Endpoint to Support Accelerated Approval 2, Food and Drug Administration, 2012.
40 Martin
AM, “Hope and Exploitation,” The Hastings Center Report, Vol. 38, No. 5, September–
October 2008, pp. 49–55.
41
Soares M, “‘Off-Label’ Indications for Oncology Drug Use and Drug Compendia: History and Current
Status,” Journal of Oncology Practice, Vol. 1, No. 3, September 2005, pp. 102–105.
42 Abernethy
AP, Raman G, Balk EM, et al., “Systematic Review: Reliability of Compendia Methods for OffLabel Oncology Indications,” Annals of Internal Medicine, Vol. 150, No. 5, March 3, 2009, pp. 336–343.
43
Bach PB, “Limits on Medicare’s Ability to Control Rising Spending on Cancer Drugs,” New England Journal of Medicine, Vol. 360, No. 6, February 5, 2009, pp. 626–633.
44 Institute of Medicine, Delivering Affordable Cancer Care in the 21st Century: Workshop Summary, The
National Academies Press, 2013.
45
Jacobson M, O’Malley AJ, Earle CC, Pakes J, Gaccione P, and Newhouse JP, “Does Reimbursement Influence Chemotherapy Treatment for Cancer Patients?” Health Affairs, Vol. 25, No. 2, March–April 2006,
pp. 437–443.
46 Danzon
PM and Taylor E, “Drug Pricing and Value in Oncology,” The Oncologist, Vol. 15, Supplement 1,
2010, pp. 24–31.
47
Montero AJ, Avancha K, Gluck S, and Lopes G, “A Cost-Benefit Analysis of Bevacizumab in Combination
with Paclitaxel in the First-Line Treatment of Patients with Metastatic Breast Cancer,” Breast Cancer Research
and Treatment, Vol. 132, No. 2, April 2012, pp. 747–751.
48
Brock DW, “How Much Is More Life Worth?” The Hastings Center Report, Vol. 36, No. 3, May–June 2006,
pp. 17–19.
49 Pollack A, “Roche Agrees to Buy Genentech for $46.8 Billion,” New York Times, March 12, 2009. As of
January 12, 2014:
http://www.nytimes.com/2009/03/13/business/worldbusiness/13drugs.html?_r=0
Case Study 2
A Cardiovascular Polypill
Nihar R. Desai, M.D., M.P.H.
Fellow in Cardiovascular Diseases,
Brigham and Women’s Hospital, Boston, Massachusetts
The Technology
There is no single definition of the cardiovascular polypill. The term generally refers to a multidrug combination pill intended to treat multiple physiological variables that increase the
risk of developing cardiovascular disease. In 2000, Wald and Law secured an EU patent for a
cardiovascular polypill consisting of aspirin, an intermediate-dose statin, folic acid, and three
antihypertensive medications at half-dose.1 The first public description of the idea was published in 2003.2 Subsequent and ongoing studies and clinical trials have employed polypills
with somewhat variable components, but most include aspirin, a statin for lipid-lowering, and
several low-dose antihypertensive medications from different therapeutic classes (i.e., diuretic,
angiotensin-converting enzyme inhibitor, beta-blocker, and/or calcium channel blocker).3
Rationale
Cardiovascular disease is the leading cause of morbidity and mortality in the United States
and around the world.4 In the developing world, the burden of cardiovascular disease has
decreased; however, these gains are seriously threatened by growing prevalence rates of obesity and diabetes mellitus. In developing countries, the toll of cardiovascular disease remains
unchecked; by the year 2020, 80 percent of global cardiovascular morbidity and mortality is
projected to occur in low- and middle-income countries.5 Many of the risk factors that lead to
cardiovascular disease and have been identified in rigorous epidemiologic analyses are modifiable through lifestyle interventions and pharmacologic therapy, including hypertension, hyperlipidemia, insulin resistance, and smoking.6 Many drugs have been shown through randomized trials to be efficacious both at reducing these important risk factors and also at lowering
the rate of major adverse cardiovascular events. Therefore, the hope of attenuating the global
burden of cardiovascular disease rests, at least in part, on successful deployment of affordable
pharmacotherapy.
Efforts to manage cardiovascular disease in the developing world have been limited for
myriad reasons, including high costs and poor adherence to complex medical regimens, inconsistent utilization of highly effective secondary preventive pharmacotherapy, and suboptimal
risk stratification for primary prevention of cardiovascular disease.7 A cardiovascular polypill,
consisting of aspirin, statin, and three antihypertensive medications, could potentially represent a simple, easy, and affordable intervention to overcome many of these barriers.
13
14
Redirecting Innovation in U.S. Health Care: Case Studies
Although the idea of an inexpensive cardiovascular polypill is particularly well-suited for
the developing world, it would also be useful in the United States, which has rates of obesity
and type II diabetes that are among the world’s highest, as well as tens of millions of people
with poorly controlled hypertension. Many of these cases are either unrecognized or undertreated, although it is widely known that hypertension increases the risks of heart disease,
stroke, and kidney failure—all leading causes of death, long-term disability, and health care
spending. Despite this knowledge, the Centers for Disease Control and Prevention estimate
that fewer than half of the 68 million Americans with high blood pressure are on an adequate
regimen of treatment.8
The specific rationale for the polypill depends on the population to which it would be
applied. For primary prevention, the polypill approach essentially obviates the need for sophisticated and expensive risk stratification tests. Instead, it would be recommended for all individuals older than a specific age—55 or 60, for example—rather than using multiple risk factors to
identify candidates.9 This is attractive because rather than tailoring treatment toward those at
highest risk, a simple intervention would be applied to the broad population that accounts for a
large proportion of adverse events.10 An added justification is that the relationship between the
severity of risk factors and the incidence of serious adverse events is continuous, not a threshold
phenomenon.11 Moreover, age has a much greater discriminatory effect on predicting the risk
of heart attack and stroke than do such physiological risk factors as hypertension and hypercholesterolemia, as important as the latter factors are.39
Applied to secondary prevention, the case for a polypill is equally, if not more, compelling. Aspirin, beta blockers, statins, and angiotensin-converting-enzyme (ACE) inhibitors are
already used for secondary prevention; each has been shown in numerous, large, randomized
clinical trials to significantly reduce the incidence of cardiovascular deaths, myocardial infarction, and stroke. Significantly, each treatment is professionally endorsed with Class I, Level of
Evidence A recommendations in clinical practice guidelines.12 In fact, it is estimated that about
half of the reduction in cardiovascular disease burden in developed countries like the United
States is attributable to advances in medical therapy.13
Unfortunately, guideline-indicated medications are not prescribed to all individuals for
whom they are indicated. Moreover, even when properly prescribed, adherence to complex and
costly multidrug regimens is distressingly low: about 50 percent.14, 15 If a cardiovascular polypill could be developed that simplifies care delivery, lowers costs through use of generic drugs,
and improves adherence by packaging the benefits of multiple effective drugs into a single pill,
it could have far-reaching implications for reducing the burden of cardiovascular disease in
both high- and low-income countries worldwide.
Development
Multiple polypills have been created over the past decade, most resulting from collaboration
between academic researchers and drug manufacturers. Although there are approximately
50 two-drug combinations for treating hypertension that have received approval from the
U.S. Food and Drug Administration (FDA) or other regulatory agencies, a cardiovascular
polypill containing three, four, or even five drugs has not been reviewed by any regulatory
authority and, thus, is not commercially available. Because the polypill has not been evaluated
with a large-scale randomized trial involving clinical endpoints, it has no chance of gaining
Case Study 2: A Cardiovascular Polypill
15
regulatory approval in the United States for primary prevention of cardiovascular disease. In
the developing world, it might be possible to bypass the conventional national-level approval
process and appeal to the World Health Organization to place a cardiovascular polypill on the
Essential Medicines list. Another option, used by Polypill Ltd in the UK, is to offer the public
access to an unlicensed pill, provided that it is prescribed by willing clinicians.16
The intellectual property issues surrounding the polypill deserve mention. Several patent
applications have been filed, including the original one by Dr. Wald and Dr. Law in 2000
(GB2361186).1, 17 Normally, the granting of a patent facilitates bringing a product to market by
providing some chance of recouping the large investment involved, perhaps hundreds of millions of dollars, to develop a product, secure regulatory approval for its sale, and manufacture
it in a consistent and safe way. Cipla Limited, a pharmaceutical company based in Mumbai,
India, is currently manufacturing a polypill based on the work of Wald and Law.
Early Adoption
The polypill has been tested in several phase II clinical trials (smallefficacy and safety trials
evaluating intermediate endpoints, such as low-density lipoprotein cholesterol [LDL-C] or
blood pressure) and a few modest-sized phase III trials (larger outcomes trials evaluating safety
and efficacy on such clinical endpoints as death, myocardial infarction, and/or stroke). Several
trials are still under way.
The need for further trials is a matter of debate. Critics argue that the safety and efficacy
of a cardiovascular polypill has not been conclusively demonstrated with a large phase III randomized trial. Advocates of the polypill note that its various ingredients have been extensively
tested and are known to be safe and effective. Moreover, many of these drugs are safely prescribed together at doses higher than those used in a polypill. For this reason, they question
whether further trials are needed, given the existing evidence. Nevertheless, under existing
FDA policy a combination pill must demonstrate efficacy and safety for its intended use as
if it were a new chemical entity, even if the individual drugs it contains have been previously
approved and are frequently prescribed together.
Subsequent Use
More than a decade has passed since the concept was proposed. Although an effective polypill could have an immense impact on the burden of cardiovascular disease, it is little closer to
entering mainstream practice than the day it was first described. What is hindering its development and adoption?
No Consensus About the Pill’s Components, Target Population, or Commercial Viability
Critics offer several concerns, beginning with the polypill’s conception and extending to its
development and potential deployment:18
1. There is no consensus on what constituent drugs should be included in the combination therapy, no consensus regarding the target population that would receive it, and
16
Redirecting Innovation in U.S. Health Care: Case Studies
2.
3.
4.
5.
6.
no consensus regarding the principal objective for its use (primary and/or secondary
prevention).
Some worry that there are important pharmaceutical formulation issues related to bioavailability, pharmacokinetics, adverse drug–drug interactions, and other problems that
combination therapy might engender.19
Investors might question the medication’s patentability and the profitability of marketing a low-cost, prevention-oriented pill in the current health care marketplace.
If a cardiovascular polypill were offered in the U.S. market, it is far from clear that providers would embrace it. Many clinicians might resist the notion of abandoning individualized care, which involves the development and meticulous titration of a customized
treatment plan, in favor of prescribing a polypill.20
Some express concern that using a polypill for primary prevention of cardiovascular
disease would “medicalize” prevention and create a moral hazard that would dampen
patients’ desire to make meaningful lifestyle modifications.21
Some assert that, regardless of the strength of evidence on the safety and efficacy of the
pill’s individual components, marketing should not be allowed unless data are generated in large phase III clinical trials to prove that a cardiovascular polypill is efficacious
and safe.
Costly Phase III Clinical Trials Required for Regulatory Approval
Irrespective of the polypill’s intuitive appeal and strong epidemiological rationale, regulators
and health care providers will expect robust evidence that it works before supporting its use
in clinical practice. The “burden of proof” will vary depending on the clinical circumstance.
In the simplest scenario, in which a patient on costly multidrug regimen switches to a polypill
containing the same (or closely related) ingredients, an outcomes trial might be unnecessary.
However, since the polypill is envisioned for first-line therapy of patients with hypertension
and/or type II diabetes rather than “therapeutic consolidation,” costly phase III outcomes trials
may be required. The same considerations will apply if the drug is envisioned for primary prevention of cardiovascular disease in people above a given age.
Even if outcome studies show that a polypill is effective, opponents will note that a key
rationale for the polypill, nonadherence to guideline-recommended multidrug therapy, is multifactorial in etiology. Clearly, therapeutic complexity (multiple drugs, multiple daily administrations, etc.) and cost are important contributors to nonadherence, but they are not the only
factors. Even a simple once-a-day polypill will not address the social, cultural, and behavioral
factors that contribute to nonadherence.22 Findings from clinical trials conducted to date have
reported discontinuation or incomplete follow-up rates ranging from a low of 2.2 percent to
27 percent.23 24 25 This variability suggests that even a simple, once-a-day tablet such as the
polypill is not a panacea for nonadherence.
Proponents of the Polypill See Large Potential Benefits
Defenders of the polypill argue that most of these objections have no merit. They note, for
example, that reaching clinical consensus over the components of the polypill is irrelevant; all
that matters is finding an effective combination of drugs that can be safely administered in a
once-a-day pill. They note that the pharmacology of the various components is well-known
and that hundreds of thousands of patients are already prescribed complex multidrug treat-
Case Study 2: A Cardiovascular Polypill
17
ments at far higher doses than anything envisioned for a polypill. They could also point out
that physicians’ resistance to abandoning “individualized care” has at least as much to do with
financial self-interests as concern for patients, many of whom are inadequately treated at the
present time. They note that although modifying harmful lifestyles is an important goal, this
strategy is largely failing. Moreover, medical prevention should not be viewed as competing
with lifestyle improvement; the two are complementary.
Given the enormous burden of ischemic heart disease and stroke in the U.S. population,
achieving even a modest decrement in the main causal factors of cardiovascular disease would
prevent a huge number of cardiovascular-related deaths, diseases, disabilities, and health care
costs.
There Is Little or No Market for a Cost-Saving Pill
Because the current economics of drug development reward the creation of patentable products
that can command high prices, there is no market rationale for spending large sums of money
to secure FDA approval of a less-expensive pharmaceutical, even one that is patented and
could replace or sharply reduce demand for more-expensive medications. For this reason, it is
unlikely that any pharmaceutical firm, no matter how wealthy, will invest the sums of money
required to fund the multicenter phase III trials that would be required to gain regulatory
approval, regardless of the polypill’s potential value to patients and society at large.
Cost and Health Impact
Wald and Law, who first proposed a cardiovascular polypill, postulated that it could reduce the
global burden of cardiovascular disease by about 80 percent.2 This was based on an estimate
that lowering cholesterol would reduce the risk by about 30 percent, lowering blood pressure
would reduce risk by another 30 percent, inhibition of platelet activation and aggregation
would reduce risk by about another 25 percent, and an additional 16 percent reduction in mortality could be achieved with folic acid. However, there are two limitations to their analysis.
First, use of folic acid has not been shown to improve outcomes in a large, randomized trial.26
In addition, although aspirin is the cornerstone of antiplatelet therapy for the secondary prevention of cardiovascular disease, its use in primary prevention is more questionable because
of the risks it poses for bleeding.27 Due at least in part to the widespread prescribing of statins,
baseline levels of total cholesterol and LDL-C have dropped in the decade since Wald and
Law’s initial publication. Therefore, the projected reductions in LDL-C and severe adverse
cardiovascular events they postulated have recently been amended.28 Even so, likely preventive
effects are still large. Also, although the epidemiologic evidence linking reduction of elevated
blood pressure with clinical benefit is irrefutable, some argue that there is no evidence to support the administration of multiple low-dose antihypertensive agents to patients who do not
have documented hypertension but are at risk for developing the condition.29, 30 This view has
been refuted by a large meta-analysis.31
Notwithstanding these concerns, most experts believe that a multidrug combination pill
that simplifies care delivery, reduces costs through use of generic drugs, and improves adherence would have far-reaching implications for reducing the global burden of cardiovascular
disease. Even if Wald and Law’s projected benefits are reduced somewhat by omitting folic acid
and aspirin, an intervention that cut the global burden of cardiovascular disease by even half
18
Redirecting Innovation in U.S. Health Care: Case Studies
the magnitude that they project would have an enormous impact on morbidity, mortality, and
health care spending.
Small-Scale Trials Have Produced Promising but Inconclusive Findings
The Indian Polycap Study (TIPS)9 was the first published. Funded by the St. John’s Research
Institute (Bangalore, India), the Population Health Research Institute (Hamilton, Canada),
and Cadila Pharmaceuticals Ltd., it enrolled 2,053 individuals aged 45 to 80 years with at least
one risk factor but without overt cardiovascular disease. An unfortunately complex allocation
scheme was used to randomize patients to receive the “polycap” (thiazide 12.5 mg, atenolol
50 mg, ramipril 5 mg, simvastatin 20 mg, and aspirin 100 mg) or to one of eight other groups
that included single drugs or various drug combinations. There was no true placebo group, just
three monotherapy arms with aspirin, thiazide, or simvastatin. Compared with subjects who
were randomized to treatment arms that did not include antihypertensive therapy, the polycap
reduced blood pressure by 7.4 mm Hg systolic and 5.6 mm Hg diastolic. In another analysis
that compared patients who were prescribed atenolol or simvastatin to those who were not, the
polycap lowered heart rate by 7.0 beats per minute and LDL-C by 0.73 mmol/L, respectively.
The degree of platelet inhibition achieved with the polycap was similar to that achieved with
aspirin alone.
The rate of noncompliance in the various treatment arms was approximately 15 percent
over the three-month follow-up period, a little less than a third of which (4 percent) was due
to side effects. Based on these results, the authors projected that widespread and sustained
use of the polycap could reduce the risk of ischemic heart disease by 62 percent and stroke by
42 percent.
Following the publication of these findings, the funders of the first TIPS study sponsored
a second polycap study, TIPS-2.32 This time, investigators randomized 518 subjects with previous vascular disease and/or diabetes to one polycap versus two polycap pills, thereby testing
“half-dose” versus “full-dose” therapy. As before, there was no placebo group. The group randomized to two polycap pills had additional, significant reductions in systolic and diastolic
blood pressure, LDL-C, and total cholesterol without any incremental changes in mean heart
rate or HDL-C. The discontinuation rates at eight-week follow-up were similar in both groups:
about 8 percent.
In 2012, the inventor of the polypill and two colleagues reported the results of a randomized double-blind placebo-controlled crossover trial of a polypill among 86 subjects aged 50
or above without a prior history of cardiovascular disease. To assess the accuracy of their prior
models, they compared observed reductions in blood pressure and low-density lipoprotein
(LDL) with those predicted from published estimates of the effects of the individual drugs.
Participants took a once-a-day polypill containing amlodipine 2.5 mg, losartan 25 mg, hydrochlorothiazide 12.5 mg, and simvastatin 40 mg for 12 weeks and a placebo for 12 weeks;
the order of treatment was randomly assigned. Each subject’s mean within-person difference
in blood pressure and LDL cholesterol was measured each 12-week period. Eighty-four subjects completed both treatment periods, a compliance rate of 98 percent. The polypill reduced
mean systolic blood pressure by 17.9 mmHg, diastolic blood pressure by 9.8 mmHg, and LDL
cholesterol by 1.4 mmol/L (1.2–1.6). The observed reductions were almost identical to those
predicted. Long-term reductions of this magnitude would be expected to have a substantial
impact on the incidence of heart attacks and strokes. In 2103, a fourth study, the Use of a Multidrug Pill in Reducing Cardiovascular Events (UMPIRE)33 trial, was published in the Journal
Case Study 2: A Cardiovascular Polypill
19
of the American Medical Association. It is part of a larger collaborative effort called the Single
Pill to Avert Cardiovascular Events (SPACE) program, which is being supported by Imperial
College London; the George Institute in Sydney, Australia; and Dr. Reddy’s Laboratories Ltd.
The SPACE program includes the Kanyini Guidelines Adherence with the Polypill (GAP)34
study and the Improving Adherence Using Combination Therapy (IMPACT)35 trial, which is
being conducted in New Zealand.
UMPIRE randomly assigned 2,004 patients to a polypill containing aspirin 75 mg, simvastatin 40 mg, lisinopril 10 mg, and either atenolol 50 mg or hydrochlorothiazide 12.5 mg
versus usual care (which might or might not involve use of similar prescription drugs, so it was
not a “pure” placebo group). Most of the study subjects had established cardiovascular disease;
28 percent were diagnosed with type II diabetes. At the completion of the 15-month study,
86 percent of patients in the fixed-dose combination study arm were still taking the medication, compared with 65 percent in the usual-care arm, a statistically significant improvement
in adherence. This translated into a 4.2 mg/dL reduction in LDL cholesterol levels and a
2.5 mmHg reduction in systolic and diastolic blood pressure—relatively modest but still clinically important differences.
Two other prospective studies of cardiovascular polypills are ongoing: the International
Polycap Trial 3 (TIPS-3), sponsored by Wellcome Trust (United Kingdom) and Cadila Pharmaceuticals Ltd., and the Heart Outcomes Prevention Evaluation 3 (HOPE-3) study, funded
by Astra Zeneca and the Canadian Institutes of Health Research.
TIPS-3 (NCT01646437) plans to enroll at least 5,500 participants (women 60 years or
older and men 55 years or older) without known heart disease or prior stroke, with a planned
follow-up period of five years. Subjects will be randomized in a 2x2x2 factorial design to polycap DS (thiazide 25mg, atenolol 100mg, ramipril 10mg, simvastatin 40mg) versus placebo,
aspirin 75 mg versus placebo, and cholecalciferol 60,000 IU monthly versus placebo. The study
is powered to detect important differences in major adverse cardiovascular events.
HOPE-336 (NCT00468923) is not evaluating a polypill per se, but rather rosuvastatin
10 mg or placebo and combination candesartan 16 mg/hydrochlorothiazide 12.5 mg versus
placebo in a 2x2 factorial study. It is projected to enroll 11,000 subjects (women 60 years or
older and men 55 years or older) from 21 countries who have no known heart disease or prior
stroke. Thus, it is oriented around primary prevention. This study is also powered to detect
important differences in major adverse cardiovascular events. It is projected to be completed in
2015. If these studies generate favorable results, they could prompt health ministries in developing countries to embrace the concept and perhaps stimulate additional interest in the United
States.
These findings beg the question: How much evidence is needed to support the adoption of a polypill in the United States? Given the enormous cost of large-scale phase III trials
and substantial evidence of the polypill’s efficacy and safety among patients at risk for serious
consequences of cardiovascular disease, is more research needed? For capitated health systems,
which have a strong interest in achieving better health at lower cost, the answer may be “no.”
Recently, Kaiser Permanente Northern California (KPNC) published the results of a
comprehensive program to improve blood pressure control among its members. The authors
reported that the program achieved “a significant increase in hypertension control compared
with state and national control rates.” Key elements of KPNC’s program, as reported by the
authors, included “a comprehensive hypertension registry, development and sharing of performance metrics, evidence-based guidelines, medical assistant visits for blood pressure measure-
20
Redirecting Innovation in U.S. Health Care: Case Studies
ment, and single-pill combination pharmacotherapy” (emphasis added).37 The pill employed by
the program contains only two agents—hydrochlorothiazide and lisinopril—so it is technically not a polypill. Even so, the authors noted that “Single pill combinations have important
advantages, including improved adherence, and lower patient cost, and are associated with
improved blood pressure control.”36
Based in part on the success of this effort, Kaiser Permanente has launched the ALL initiative (named for the systematic use of aspirin, lisinopril, and lipid-lowering medication). The
organization’s goal is to reduce the incidence of heart attacks and strokes among its members
through secondary prevention. Kaiser has also rolled out the PHASE program (Preventing
Heart Attacks and Strokes Everyday), which seeks to build on ALL by adding a beta blocker to
the regimen and encouraging healthy lifestyle changes. If Kaiser Permanente achieves the outcomes it seeks with these initiatives, it could open the door for other capitated health systems
to follow. In that case, the simplicity of using a polypill to advance secondary or even primary
prevention could be an appealing option.
Development Block
The cardiovascular polypill is an example of “development block.” Despite the fact that this
promising technology offers the prospect of tremendous value to society in terms of health benefits per dollar spent, its pace of development is slow. This appears to be largely due to the lack
of a compelling business model, given the current way new drugs are patented, regulated, and
reimbursed. If, as the polypill’s proponents assert, individuals could cut their risk of sustaining
a stroke or developing ischemic heart disease by half or more simply by taking an inexpensive
once-a-day pill with few or no side effects, it is likely that millions of people would choose to
do so.28 However, there is no obvious path for the inventors of the polypill to get from “here”
to “there.”
The polypill’s advocates note that its various components have been extensively studied and are highly efficacious. All are approved for sale, provided that they are individually
prescribed and swallowed as individual pills. Several antihypertensive pills are marketed as
two-drug formulations. The epidemiological and clinical trial data on the potential benefits
of combination therapy are impressive. Short-term clinical trials of the polypill have produced
reductions in blood pressure and LDL cholesterol comparable to those achieved in prior, individually dosed clinical dose trials. Side effects were uncommon, as might be expected given the
low doses involved. If the physiological improvements observed in these studies were sustained
over time, treated subjects should have far fewer cases of heart attack and stroke. The benefits
this would produce—both in terms of human health and health care spending—would be
profound.
For the same investment pharmaceutical firms routinely make to test a promising diabetes medication or a new cancer drug, a polypill could be rigorously evaluated through a large
phase III randomized trial involving tens of thousands of subjects. If demonstrated to be safe
and effective at preventing ischemic heart disease and strokes, the benefits of promoting this
treatment would dwarf the development costs of getting a polypill to market. The problem is
that under the current structure of pharmaceutical markets and regulators, the financial benefits of the polypill would be realized by patients, payers, and perhaps capitated health care
systems, not the pill’s developer or the doctors who prescribe it.
Case Study 2: A Cardiovascular Polypill
21
As long as the regulatory costs to enter the U.S. market are as high as they are, the drug
development market will remain aligned to reward the creation and marketing of high-priced
pharmaceuticals. In that case, it is unlikely that we will see much work on such dramatically
cost-lowering drugs as the cardiovascular polypill.
No one knows how many promising drugs, biologics, devices, and diagnostic technologies are languishing on the fringes of our health care marketplace. But a process recently
embraced by the Office of Science & Technology in the Department of Homeland Security
(DHS) might be useful way to identify interesting ideas. The approach, known as “technology foraging,” was envisioned to help DHS identify existing but largely overlooked inventions
that can be swiftly adopted or repurposed to meet urgent agency needs.38 If the U.S. Department of Health and Human Services or a large integrated health care delivery system, such as
Kaiser Permanente, the Veterans Healthcare Administration, or the Military Health System,
adopted a similar approach, it might be possible to identify and evaluate promising health care
technologies that offer the potential for dramatic benefits at far lower cost than the options we
have today.
References
1
Polypill Patent, EU1272220, priority date April 10, 2000.
2 Wald NJ and Law MR, “A Strategy to Reduce Cardiovascular Disease by More Than 80%,” BMJ, Vol. 326,
2003, pp. 1419–1424.
3
Guglietta A and Guerrero M, “Issues to Consider in the Pharmaceutical Development of a Cardiovascular
Polypill,” Nature Clinical Practice Cardiovascular Medicine, Vol. 6, 2009, pp. 112–119.
4
World Health Organization, “Table 2: Cause-Specific Mortality and Morbidity,” World Health Statistics 2012, 2013. As of January 12, 2014:
http://www.who.int/whosis/whostat/EN_WHS2012_Table2.pdf
5
World Health Organization, “Table 2: Cause-Specific Mortality and Morbidity,” World Health Statistics 2009, 2010. As of January 12, 2014:
http://www.who.int/whosis/whostat/EN_WHS09_Table2.pdf
6
Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, McQueenM, Budaj A, Pais P, Varigos J,
Lisheng L, and INTERHEART Study Investigators, “Effect of Potentially Modifiable Risk Factors Associated with Myocardial Infarction in 52 Countries (the INTERHEART Study): Case Control Study,” Lancet,
Vol. 364, 2004, pp. 937–952.
7
Lonn E, Bosch J, Teo KK, Pais P, Xavier D, and Yusuf S, “The Polypill in the Prevention of Cardiovascular Disease: Key Concepts, Current Status, Challenges, and Future Directions,” Circulation, Vol. 122, 2010,
pp. 2078–2088.
8
Centers for Disease Control and Prevention, “Vital Signs: Prevalence, Treatment, and Control of Hypertension—United States, 1999–2002 and 2005–2008,” Morbidity and Mortality Weekly Report, Vol. 60, No. 4,
February 4, 2011, pp. 103–108.
9
Lloyd-Jones DM, Leip EP, Larson MG, D’Agostino RB, Beiser A,Wilson PW, Wolf PA, and Levy D, “Prediction of Lifetime Risk for Cardiovascular Disease by Risk Factor Burden at 50 Years of Age,” Circulation,
Vol. 113, 2006, pp. 791–798.
10
Rose G, “Strategy of Prevention: Lessons from Cardiovascular Disease,” BMJ, Vol. 282, 1981,
pp. 1847–1851.
22
Redirecting Innovation in U.S. Health Care: Case Studies
11
Lewington S, Clarke R, Qizilbash N, Peto R, Collins R, and Prospective Studies Collaboration, “AgeSpecific Relevance of Usual Blood Pressure to Vascular Mortality: A Meta-Analysis of Individual Data for
One Million Adults in 61 Prospective Studies,” Lancet, Vol. 360, 2002, pp. 1903–1913.
12
Antman EM, Hand M, Armstrong PW, Bates ER, Green LA, Halasyamani LK, Hochman JS, Krumholz HM, Lamas GA, Mullany CJ, Pearle DL, Sloan MA, Smith SC, Jr., Anbe DT, Kushner FG, Ornato JP,
Jacobs AK, Adams CD, Anderson JL, Buller CE, Creager MA, Ettinger SM, Halperin JL, Hunt SA, Lytle
BW, Nishimura R, Page RL, Riegel B, Tarkington LG, and Yancy CW, “2007 Focused Update of the ACC/
AHA 2004 Guidelines for the Management of Patients with ST-Elevation Myocardial Infarction: A Report of
the American College of Cardiology/American Heart Association Task Force on Practice Guidelines: Developed in Collaboration with the Canadian Cardiovascular Society Endorsed by the American Academy of
Family Physicians: 2007 Writing Group to Review New Evidence and Update the ACC/AHA 2004 Guidelines for the Management of Patients with ST-Elevation Myocardial Infarction, Writing on Behalf of the 2004
Writing Committee,” Circulation, Vol. 117, 2008, pp. 296–329.
13
Ford ES, Li C, Zhao G, Pearson WS, and Capewell S, “Trends in the Prevalence of Low Risk Factor
Burden for Cardiovascular Disease Among United States Adults,” Circulation, Vol. 120, 2009, pp. 1181–1188.
14
Jackevicius CA, Mamdani M, and Tu JV, “Adherence with Statin Therapy in Elderly Patients With
and Without Acute Coronary Syndromes,” Journal of the American Medical Association, Vol. 288, 2002,
pp. 462–467.
15
Kramer JM, Hammill B, Anstrom KJ, Fetterolf D, Snyder R, Charde JP, Hoffman BS, Allen LaPointe N,
and Peterson E, “National Evaluation of Adherence to Beta-Blocker Therapy for 1 Year After Acute Myocardial Infarction in Patients with Commercial Health Insurance,” American Heart Journal, Vol. 152, No. 454,
2006, pp. e451–e458.
16
Polypill.com, home page, 2013. As of January 10, 2014:
https://www.polypill.com/
17
Guglietta A and Guerrero M, “Issues to Consider in the Pharmaceutical Development of a Cardiovascular
Polypill,” Nature Clinical Practice Cardiovascular Medicine, Vol. 6, 2009, pp. 112–119.
18
Lonn E, Bosch J, Teo KK, Pais P, Xavier D, and Yusuf S, “The Polypill in the Prevention of Cardiovascular Disease: Key Concepts, Current Status, Challenges, and Future Directions,” Circulation, Vol. 122, 2010,
pp. 2078–2088.
19
Patel A, Shah T, Shah G, Jha V, Ghosh C, Desai J, Khamar B, and Chakraborty BS, “Preservation of Bioavailability of Ingredients and Lack of Drug-Drug Interactions in a Novel Five-Ingredient Polypill (Polycap):
A Five-Arm Phase I Crossover Trial in Healthy Volunteers,” American Journal of Cardiovascular Drugs, Vol. 10,
2010, pp. 95–103.
20
Cannon CP, “Can the Polypill Save the World from Heart Disease?” Lancet, Vol. 373, 2009,
pp. 1313–1315.
21
Wald DS and Wald NJ, “The Polypill in the Primary Prevention of Cardiovascular Disease,” Fundamentals
of Clinical Pharmacology, Vol. 24, No. 1, 2010, pp. 29–35.
22
Desai NR and Choudhry NK, “Impediments to Adherence Among Post Myocardial Infarction Patients,”
Current Cardiology Reports, Vol. 15, No. 1, 2013, p. 322.
23
Yusuf S, Pais P, Afzal R, Xavier D, Teo K, Eikelboom J, Sigamani A, Mohan V, Gupta R, and Thomas N,
“The Indian Polycap Study (TIPS). Effects of a Polypill (Polycap) on Risk Factors in Middle-Aged Individuals Without Cardiovascular Disease (TIPS): A Phase II, Double-Blind, Randomized Trial,” Lancet. Vol. 373,
2009, pp. 1341–1351.
24
Rastegarpanah M, Malekzadeh F, Thomas NG, Mohagheghi A, Cheng KK, and Marshall T, “A New Horizon in Primary Prevention of Cardiovascular Disease: Can We Prevent Heart Attack by ‘Heart Polypill’?”
Archives of Iranian Medicine, Vol. 11, 2008, pp. 306–313.
Case Study 2: A Cardiovascular Polypill
23
25
Wald DS, Morris JK, and Wald N, “Randomized Polypill Crossover Trial in People Aged 50 and Over,”
PLoS ONE, Vol. 7, No. 7, 2012, p. e41297. As of January 12, 2014:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0041297
26 Clarke
R, Halsey J, Lewington S, Lonn E, Armitage J, Manson JE, Bønaa KH, Spence JD, Nygård O,
Jamison R, Gaziano JM, Guarino P, Bennett D, Mir F, Peto R, and Collins R, for the B-Vitamin Treatment
Trialists’ Collaboration, “Effects of Lowering Homocysteine Levels with B Vitamins on Cardiovascular Disease, Cancer, and Cause-Specific Mortality: Metaanalysis of 8 Randomized Trials Involving 37,485 Individuals,” Archives of Internal Medicine, Vol. 170, 2010, pp. 1622–1631.
27
Baigent C, Blackwell L, Collins R, Emberson J, Godwin J, Peto R, Buring J, Hennekens C, Kearney P,
Meade T, Patrono C, Roncaglioni MC, and Zanchetti A; Antithrombotic Trialists’ (ATT) Collaboration,
“Aspirin in the Primary and Secondary Prevention of Vascular Disease: Collaborative Meta-Analysis of Individual Participant Data from Randomised Trials,” Lancet, Vol. 373, 2009, pp. 1849–1860.
28
Wald NJ and Wald DS, “The Polypill Concept,” Heart, Vol. 96, No. 1, 2010, pp. 1–4.
29
Turnbull F and the Blood Pressure Lowering Treatment Trialists’ Collaboration, “Effects of Different
Blood-Pressure-Lowering Regimens on Major Cardiovascular Events: Results of Prospectively-Designed Overviews of Randomized Trials,” Lancet, Vol. 362, 2003, pp. 1527–1535.
30
Zanchetti A, Grassi G, and Mancia G, “When Should Antihypertensive Drug Treatment Be Initiated and
to What Levels Should Systolic Blood Pressure Be Lowered? A Critical Reappraisal,” Journal of Hypertension,
Vol. 27, 2009, pp. 923–934.
31
Law MR, Morris JK, and Wald NJ, “Use of Blood Pressure Lowering Drugs in the Prevention of Cardiovascular Disease: Meta-Analysis of 147 Randomized Trials in the Context of Expectations from Prospective
Epidemiologic Studies,” BMJ, Vol. 338, 2009, p. b1665.
32
Yusuf S, Pais P, Sigamani A, Xavier D, Afzal R, Gao P, and Teo KK, “Comparison of Risk Factor Reduction and Tolerability of a Full-Dose Polypill (with Potassium) Versus Low-Dose Polypill (Polycap) in Individuals at High Risk of Cardiovascular Diseases: The Second Indian Polycap Study (TIPS-2) Investigators,” Circulation: Cardiovascular Quality and Outcomes, Vol. 5, 2012, pp. 463–471.
33
Thom S, Poulter N, Field J, et al. for the UMPIRE Collaborative Group, “Effects of a Fixed-Dose Combination Strategy on Adherence and Risk Factors in Patients with or at High Risk of CVD: The UMPIRE Randomized Clinical Trial,” Journal of the American Medical Association, Vol. 310, No. 9, 2013, pp. 918–929.
34 Australian
and New Zealand Clinical Trials Registry, “The Kanyini Guidelines Adherence with the
Polypill Study.” As of January 14, 2014:
http://www.anzctr.org.au/trial__view.aspx?ID=83248
35
ClinicalTrials.gov, “Polypill for Prevention of Cardiovascular Disease,” last updated May 25, 2010. As of
January 12, 2014:
http://clinicaltrials.gov/ct2/show/NCT00567307?term=NCT00567307&rank=1
36
ClinicalTrials.gov, “Heart Outcomes Prevention Evaluation-3 (HOPE-3),” last updated February 13, 2014.
As of January 13, 2014:
http://clinicaltrials.gov/ct2/show/NCT00468923?term=NCT00468923&rank=1
37
Jaffe MG, Lee GA, Young JD, et al., “Improved Blood Pressure Control Associated with a Large-Scale
Hypertension Program,” Journal of the American Medical Association, Vol. 310, No. 7, 2013, pp. 699–705.
38
Hopmeier M, “Technology Foraging: A Novel Approach to Technology Problem Solving Within the DHS
Science and Technology Directorate,” in Kamien D, ed., McGraw-Hill Homeland Security Handbook, 2nd edition, 2012. As of January 13, 2014:
http://www.homelandsecuritybook.com/section9
39 Wald
NJ, Simmonds M, and Morris JK, “Screening for Future Cardiovascular Disease Using Age Alone
Compared with Multiple Risk Factors Plus Age,” PLoS One, Vol. 6, 2011, p. e18742.
Case Study 3
Electronic Health Records1
Enesha Cobb, M.D., M.T.S., M.Sc.
Fellow, Robert Wood Johnson Foundation Clinical Scholars Program,
University of Michigan and the Department of Veterans Affairs,
VA Center for Clinical Management and Research, Ann Arbor VA Healthcare System
Kori Sauser, M.D., M.Sc.
Fellow, Robert Wood Johnson Foundation Clinical Scholars Program,
University of Michigan and the Department of Veterans Affairs,
VA Center for Clinical Management and Research, Ann Arbor VA Healthcare System
The Technology
Electronic health records (EHRs) include a variety of medical documentation systems “generally focused on medical care,” including patient information, diagnoses, procedure codes,
and medications.1 There are many types of EHRs, ranging from hospital systemwide records,
ambulatory care records, emergency department records, and medication administration
documents.
EHRs are often called electronic medical records (EMRs). The latter is a legal record documenting the patient encounter and clinical services and is typically controlled by the hospital
or care delivery organization that generates it.2 For this review, we use the term EHRs because
they encompass more types of medical data.
Rationale
EHRs were developed in response to the needs of a range of stakeholders in the health care
system. Because paper charts are often illegible, have missing data, are difficult to organize,
occupy substantial physical space, are sometimes misplaced or lost, and are difficult to share,
stakeholders immediately grasped the potential of using computers to store, retrieve, and process medical record data faster than any paper chart. Initially, the Office of Data Management
and Telecommunications sought computerization for administrative and research purposes
on a national level; locally, Veterans Affairs (VA) medical centers looked to computerization
to improve their efficiency and quality of care.3 Likewise, academic medical centers sought to
“utilize the speed, efficiency and reliability of digital computer processing” to collect, store, and
retrieve patient data.4
1
Acknowledgments: This report was prepared with the assistance of Katherine Nestor, University of Michigan research
assistant, and Marisa L. Conte, translational research liaison at the University of Michigan Taubman Health Sciences
Library. Kenneth Kizer, Andy Gettinger, and Robert Rudin reviewed earlier drafts and offered helpful comments.
25
26
Redirecting Innovation in U.S. Health Care: Case Studies
Development
EHRs were envisioned as early as the 1950s, with the testing of computer-based medical histories.4,5 In the mid-1960s, Dr. Lawrence Weed, inventor of the “problem oriented medical
record,”6 and others began working on electronic versions of a patient’s medical record that
allowed clinicians to focus on manageable bits of information that would be easily computerized and retrieved.7 During the same period, several academic institutions developed computer-based patient records, including hospital and office-based medical record systems.8
In the 1960s and 1970s, with support from the National Center for Health Services
Research and Development (now the Agency for Healthcare Research and Quality), the
Denver Department of Health and Hospitals and Beth Israel Hospital in Boston developed
early medical information systems.8 The Latter Day Saints Hospital and the University of Utah
received funding from the Regional Medical Program to expand its prototype health information system beyond research purposes.8 The Massachusetts General Hospital System secured
grants from the National Institutes of Health (NIH),9 the National Center for Health Services
Research and Development,8 and a contract funded jointly by the NIH and the American
Hospital Association to establish its computer facilities and sponsor the development of computerized hospital records.10
Early Adoption
Almost as soon as EHRs were developed, new capabilities were added, including providing
clinicians with access to data from multiple caregivers, clinical decision support to improve the
management of complex patients, and the capability to track the process of care and outcomes.
In the 1980s, a time when several health care systems were organizing themselves into health
maintenance organizations (HMOs), providers were expected to accurately document their
care so that it could be evaluated for quality, outcomes, and cost. This drove early adoption of
EHRs and broadened their uptake.11
The federal government encouraged adoption by providing some funding for computerbased patient records projects, but most health systems self-funded their EHRs.11, 12 By 1991,
early adopters were reporting a range of benefits, including lower costs, shorter inpatient stays,
improved outcomes, and fewer medical errors.11, 13–16
In 2000, To Err is Human, the Institute of Medicine’s (IOM’s) landmark report on medical error, provided additional impetus for the adoption of EHRs.17 In addition to describing
the magnitude and dimensions of medical error, the IOM proposed solutions. One of its major
recommendations was that health care providers and systems use EHRs and other computerbased systems to reduce medical errors and improve the quality of care.
Subsequent Use
In 2004, the George W. Bush administration initiated a push to transition the nation from
paper to electronic medical records. To jump-start the process, the administration requested
a doubling of federal funding for health information technology (HIT) and created the position of the national health information coordinator (later renamed the national coordinator for
Case Study 3: Electronic Health Records
27
health information technology [IT]). Congress did not approve the administration’s funding
request, but the establishment of the Office of the National Coordinator for Health IT set the
stage for what was to follow.
Studies Identify Potential Value
The next big push for EHR adoption came in 2005, when a team of RAND researchers published a major analysis of the potential economic payoffs of widespread adoption of HIT.19
Using sophisticated models based on empirical observations of IT adoption in other sectors
of the U.S. economy, RAND’s team projected that if “interoperable and interconnected [electronic medical records] are widely adopted and used effectively,” the United States could save
more than $81 billion annually through improved efficiency and enhanced patient safety. The
Center for Information Technology Leadership reached similar conclusions about the potential
value of health information exchanges.20
Despite these optimistic predictions of benefit, adoption of EHRs lagged. In 2008, a
national survey of 2,758 physicians revealed that just 4 percent had a fully functional EHR,
and 13 percent had a basic system. Although these early adopters were few, they reported positive effects of the systems on several dimensions of quality of care and generally high levels
of satisfaction with their systems. Financial concerns were viewed as the biggest obstacle to
adoption.116
HITECH Provides Financial Incentives for Adoption
In 2009, the stimulus funding provided through the American Recovery and Reinvestment
Act (ARRA) provided an opportunity for EHR supporters in the federal government to boost
adoption nationwide. A major section of the ARRA entitled Health Information Technology
for Economic and Clinical Health (HITECH),32 put billions of dollars in federal incentives
into play to spur the adoption and “meaningful use” of EHRs. There was only one catch—
providers had to move quickly to secure the funding.117 HITECH had a dramatic effect: Over
the next few years, roughly 55 percent of eligible physicians and nearly 4,000 hospitals took
advantage of HITECH’s incentives. To date, about $5.9 billion in incentive payments has been
disbursed to doctors, and $8.7 billion has been provided to hospitals.33
Not surprisingly, federal engagement in promoting EHR development drew in private
firms and investors. Between 2007 and 2012, the EHR market doubled from $9.5 billion
to $20.7 billion in annual activity.19 As a result, health care institutions and providers found
themselves with an abundance of options.
Despite the surge in public and private investment, EHR adoption in the United States
continues to lag adoption in other high-income countries.21 Today, about 80 percent of
U.S. hospitals have a basic EHR system,22 but less than 50 percent meet “meaningful use”
criteria.23 Approximately half of health care professionals use some sort of EHR,22 but only 10
percent of physician offices meet meaningful use criteria.24
Multiple Barriers to Adoption
Rapid adoption of EHRs has been hindered by a variety of factors, including a fragmented
marketplace, changing federal incentives, provider uncertainty about the regulatory landscape,
and the striking lack of interoperability between systems. Uptake has been particularly slow
among smaller hospitals and physician groups. Notwithstanding the availability of federal
subsidies, cost remains a concern for many health care providers.25 Many perceive that the eco-
28
Redirecting Innovation in U.S. Health Care: Case Studies
nomic benefits of installing an EHR system largely accrue to payers and vendors (through efficiencies in diagnostic testing and use of medications); however, health care providers must bear
the cost of implementation.25 Although federal incentive payments have helped defray startup
costs, much of the money invested to date has gone to large hospitals and multispecialty physician practices, largely to replace or upgrade previously acquired platforms. Smaller hospitals,
rural facilities, and small physician practices continue to hold back.26–28
A major factor in predicting whether or not a small- to medium-sized practice will adopt
an EHR is physician “technology readiness”—their perception that they are themselves sophisticated technology consumers and their belief that the technology is easy to use, is useful, and
gives the practice a comparative advantage.29 Many physicians fear that installing an EHR will
disrupt clinic workflow, generate high maintenance costs, and, if the system proves difficult to
use, slow them down. Lack of interoperability is a problem,25, 30, 31 as are concerns about assuring data security and patient privacy.25, 30, 31
HITECH Promotes Adoption of Existing Systems, Not Better Ones
The incentive payments offered through the HITECH act undoubtedly accelerated adoption
of EHRs, but physician practices and medical centers had to move too quickly to qualify for
incentive payments. Even the national coordinator for health IT at the time, Farzad Mostashari, referred to the “rush” for adoption.34 If the EHRs that these hospitals and practices purchased were not ideally configured to meet their needs, they could ultimately delay the adoption of more functional health information technology.35
One feature that many current systems lack is interoperability. HITECH’s language
clearly indicated that Congress wanted HIT systems to be interconnected and interoperable so
that they can readily share data between providers. For this reason, HITECH specified that to
qualify for “meaningful use” incentive payments, providers must not only use their EHR in a
meaningful manner, but also demonstrate that it is “connected” in a way the provides for the
electronic exchange of health information.36
Unfortunately, the rules that the U.S. Department of Health and Human Services (HHS)
issued to guide implementation of HITECH watered down the requirement for connectivity.
The practical effect was to promote adoption of existing platforms, rather than encourage the
development of interconnected systems.
Although this decision was welcomed by large vendors and many health care systems, it
was criticized by others. In a subsequent hearing before the House Ways and Means Subcommittee on Health, a national health IT expert offered the following observations:
Unfortunately, the [interoperability] measure specified for 2011 and 2012 is to “perform
at least one test of the EHR’s capacity to electronically exchange information.” There is
no requirement to implement authentication, consent, authorization, disclosure management, or any other services specifically mentioned in the HITECH Act that are essential
to genuinely enable secure electronic exchange of information. Moreover, the test can be
performed with “dummy” data for a “fictional patient.”
My view is that this is not really exchange, and comes nowhere close to what was intended
by Congress when it passed HITECH. It’s not good enough . . . . By eliminating any real
requirement for HIE [health information exchange] until 2013, the final meaningful use
rule dramatically undermines the incentives to innovate and invest in the infrastructure
necessary to enable it.37
Case Study 3: Electronic Health Records
29
At the same hearing, David Blumenthal, M.D., who was the national coordinator for
health IT at the time, was pressed to explain why HHS was not taking more forceful action to
promote interoperability. He offered an unusually candid response:
You need the ‘pipes’ to enable information to flow, and you need agreements among doctors
and hospitals about the terms of the exchange. . . . You also need to verify who’s a doctor
and who’s not; you don’t want information flowing to people who don’t deserve to have it.
. . . We have a system in which there’s virtually no information exchange going on at all.
We’re doing a lot of things to make exchange easier, such as setting standards for interoperability . . . . We are moving toward robust exchange, but we are starting where we think
industry is, and putting them on notice that they are going to have to move fairly rapidly
while we help them along.118
By subsidizing “where the industry is” rather than where it needed to go, HHS rule-makers allowed hospitals and health care providers to use billions in federal subsidies to purchase
EHRs that did not have the level of connectivity envisioned by the authors of the HITECH
act.36 The result was far short of the “interoperable and interconnected” network that the
RAND team envisioned in its 2005 analysis of the benefits of EHRs.61, 35
Cost and Health Impact
Widespread adoption of interoperable and interconnected EHRs could bring substantial benefits to American health care and the U.S. economy.19, 38 Unfortunately, most of the systems in
place today do not “talk” to each other. In addition, few health care systems have redesigned
their work processes to take advantage of the functionality that HIT offers.39 This has blunted
the benefit that EHRs are supposed to provide.35
Clinical Impact of EHRs
The potential of EHRs to improve patient care is a major factor driving the push for adoption.
It is thought that widespread use of EHRs could reduce medical errors by decreasing errors in
communication, inaccurate dosing, and missed drug–drug interactions.40–42 EHRs could also
improve care by issuing patient and provider reminders for prophylactic care, ensuring adherence to clinical guidelines by providing point-of-care decision support, disease surveillance,
and patient monitoring.43–45 Despite these theoretical benefits, studies of the impact of EHRs
on the processes and outcomes of care have generated mixed results.46–52, 53
In 2010, DesRoches and colleagues examined EHR adoption in U.S. hospitals and its
relationship to quality and efficiency. Across the large number of metrics they examined, they
found that the relationships were “modest at best and generally lacked statistical or clinical significance.” Although the authors noted that use of clinical decision support produced
small quality gains, they concluded that simply promoting widespread adoption of EHRs is
unlikely to have a major impact on quality. Instead, they concluded that “policies are needed
that encourage the use of electronic health records in ways that will lead to improvements in
care.”119
30
Redirecting Innovation in U.S. Health Care: Case Studies
Economic Impact
EHRs’ principal economic impact to date has been facilitating clinical documentation to support timely and complete billing.54–57 EHRs can also be programmed to generate reminders
and prompts to boost follow-up visits and routine health care.57 The data on the impact of
EHRs on health care operations are less clear. Some organizations have reported improved
performance,58 but others have reported less-favorable findings.24, 46
Viewed in hindsight, it is not surprising that introducing IT to health care did not immediately generate substantial gains in productivity. Other industries that adopted IT much earlier experienced similar growing pains. In the 1970s and 1980s, a time when the computing
capacity of the U.S. economy grew more than a hundredfold, productivity growth fell to less
than half the average rate over the preceding 25 years.59 The affected industries responded by
modifying their IT systems to make them easier to use and by revising work processes to take
advantage of IT’s capabilities. Once that was done, productivity soared.39 Because health care
did not learn from the early adopters of IT, it appears destined to follow the same path.
Societal Benefits of EHRs
Properly designed, EHRs should enable a robust research infrastructure through increased
depth and breadth of accessible stored data60 and enabling real-time data analysis.61 Population
health could be improved by increasing vaccination rates62, 63 and by enhancing the ability of
health care providers to swiftly spot emerging public health threats.64, 65 Although such benefits
could generate substantial cost savings for the health care system and society as a whole, they
are largely based on modeling and optimistic projections of EHR uptake and performance.19
Barriers to Success
The theoretical case for EHRs is strong, but the technology’s promised impact is currently
blunted by limitations in its design, uptake, and use.
Lack of Interoperability
The health IT systems that currently dominate the market are not designed to talk to each other.35, 120
This problem might have been avoided if the federal government had pressed the issue when
meaningful use standards were being drafted. Instead, interoperability standards were watered
down, and vendors were allowed to apply for meaningful use certification post-market. It is
unclear how many, if any, were rejected.
The practical effect of the HITECH Act was to shift the government’s early focus from
encouraging interoperability to accelerating uptake of existing EHRs, provided they met criteria for “meaningful use.”34 Although the act called for the creation of health information
exchanges, the final rules that the administration issued to guide implementation set a low bar
for interoperability. Instead of using federal incentives to spur development and sale of interoperable EHRs, the Office of the National Coordinator for Health IT sought to bring the health
care industry on board.37, 66
Advocates of interoperability were disappointed. In their view, dropping the requirement
for functional health information exchanges eliminated the industry’s incentive to develop
interoperability. As one critic noted, this created “a vicious cycle where the lack of incentives
breeds lack of investment and innovation, thereby perpetuating the lack of [Health Information Exchange] capabilities.”37
Case Study 3: Electronic Health Records
31
The landscape for interoperability should begin to change in 2014, when EHR provisions included in the Affordable Care Act (ACA) take effect. They will raise the bar for physicians and hospitals to qualify for federal incentives in 2014. The new standards will require
that future EHRs connect to a health information exchange. Over 700 EHR vendors and
thousands of providers must adopt common data standards to enable interoperability.67 The
Office of the National Coordinator (ONC) has established specific goals for health information exchange, including policies to facilitate directed exchange, query-based exchange, and
consumer-mediated exchange.68
If the ACA’s EHR provisions are not weakened or delayed, the move to greater interoperability should substantially increase the value of EHRs to health care providers and their
patients.29, 69–71 The shift will be less welcome to large legacy vendors because it will blur the
competitive edge they currently enjoy.72 Health care systems may be less-than-enthusiastic
adopters because functional HIEs will make it easier for patient to see nonaffiliated healthcare
providers or switch to a competing health care system.35 Irrespective of industry ambivalence,
the Office of the National Coordinator for Health IT is determined to press ahead.34
Some communities are attempting to bridge the interoperability gap by creating regional
health information exchanges to provide for limited information-sharing between systems. In
most cases, this is on a “read-only” basis, but that is far better than nothing. In theory, this
capability should be particularly helpful for emergency care because the time-critical nature
of patients’ problems may compel them to seek care at a hospital other than at their usual
source of care (and therefore, beyond the reach of their EHR). In those cases, in the words of
an emergency physician who is an expert in EHRs, “I have to work with family members to
assemble a medication list (which is often incomplete), order tests that may have already been
done recently, and risk delivering treatments that may have caused previous adverse effects.”121
Recently, a study done in Memphis, Tenn., found that having access to a regional health information exchange reduced repeated emergency department imaging for headache and back
pain.122 But the exchange was used only 12.5 percent of the time, in all likelihood because
physicians had to use a separate, password-protected portal that required substantial time and
effort to access. In another health care system with a more facile portal, emergency department
physicians integrated records across different health care systems in 57 percent of encounters.121
Usability matters.
Uncertain Return on Investment
EHRs were supposed to produce major savings through improved efficiency and better clinical
outcomes.19 To date, the federal government has invested over $14 billion to secure these benefits. However, much of the literature detailing the promised benefits of EHRs was based on a
small number of early EHR adopters, so it may represent “best case” examples.43 Currently, a
large gap remains between the postulated and realized benefits of HIT.46
There are three main reasons why the economic benefits of EHRs have been disappointing to date. First, uptake has been slower than predicted because of high acquisition costs
and concerns about the cost of ongoing maintenance. Many providers also fear the disruption of office workflow and are not convinced that EHRs will generate long-term savings.
Privacy concerns and fear of unintended consequences are limiting uptake as well.38, 54, 73, 74
The combination effects of limited interoperability and slow uptake have limited the “network
effects” that fueled the success of the telephone and, more recently, social media sites like Facebook and Twitter.35
32
Redirecting Innovation in U.S. Health Care: Case Studies
Second, lack of usability is an ongoing problem.123 Many clinicians complain that EHRs
hinder their productivity via nonintuitive interfaces and cumbersome search mechanisms.75 To
be maximally efficient, EHRs should be designed to be as easy to use as a consumer IT system
(Table 3.1). Some of the helpful features of EHRs can be problematic if they are not properly calibrated. For example, some EHRs generate so many alerts about possible drug–drug
interactions that providers tend to disregard them—a practice that increases the risk of medical errors.76 To address concerns like these, the Agency for Healthcare Research and Quality
(AHRQ) commissioned RAND to develop a guide to reducing the unintended consequences
of EHRs. It can be downloaded for free.77
The third major barrier to ROI is the failure of many health care organizations to
redesign workflows, redefine provider roles, empower self-care, and optimize other processes, such as utilization of inpatient beds to increase efficiency and productivity. Once such
changes are made, the promised productivity gains of EHRs and other forms of health IT
should become increasingly apparent.
Determining Ownership of Personal Health Information
Efforts to facilitate the exchange of health information between providers must be balanced
against the need to assure the patient’s privacy. Some states, such as New Hampshire, have
moved proactively to define in statute that patients control their personal health information.
But regardless of who “owns” patient data, current vendor practices make extracting the
data difficult. Many vendors include a clause in their service contracts that precludes independent data extraction from their software.72 Others have complicated the process by requiring
multistep information requests for data transfer. Some institutions have responded by exporting their clinical data to third-party sites for management.78
EHRs can improve care coordination and patient engagement, but only if information
can be exchanged between providers124 and, ideally, accessed by patients.35 The federal government’s “Blue Button” initiative is intended to make it easier for patients to access their own
health information and share it with different health care providers. Although versions of Blue
Table 3.1
Health IT and Consumer IT Comparisons
Information Technology 2013
Health IT 2013
Examples
Facebook, Amazon, Google, Yelp
Epic, Cerner, McKesson
Decision support
Dynamically updated, data-driven
personalized recommendations
Rule-based, poorly integrated into user
experience
Innovation cycle
Low cost of failure
Low interaction complexity
High cost of failure
High interaction complexity
Development
paradigm
Rapid, experiment-driven development for
continuous iterative improvements based
on monitoring user activity
Waterfall model with scheduled releases
and minimal user feedback
Architecture
Distributed, service-oriented architectures
Monolithic enterprise platforms
Interoperability and
collaboration
High degree of cross-platform
Proprietary designs and terminologies not
interoperability – standard integration with well-suited for interface; poorly integrated
mobile and social platforms
information exchange
NOTE: Table devised by Daniella Meeker of RAND, used with permission.
Case Study 3: Electronic Health Records
33
Button are already available to veterans, uniformed service members, Medicare beneficiaries,
and some private insurance company beneficiaries, access to it is not required by law.79 Blue
Button depends on the content being available in the source data. For this reason, the amount
of data that can be accessed by a veteran in the VA health system is different from the amount
available to a Medicare beneficiary.
Blue Button is helpful, but it is only one piece of the puzzle. It does not give patients
personal control over their electronic health information because it does not allow them to
transmit their medical record in a standard, computable format to a health care provider of
their choice. The federal government has recently taken preliminary steps to encourage health
IT vendors to add functionality that will allow patients to view, download, and transmit their
electronic records to a third party. In 2014, view, download, and transmit functionality will
be required to qualify for federal incentive payments.80 This development is overdue. Although
writers have described personal health data as a new “asset class,”81 the truth is that few patients
currently have the ability to access or manage their own health information. Many patients
still come to the ER with a bag of medications, copies of paperwork, or a medical ID bracelet
as their only form of health information.82
Concerns About Patient Safety
Espoused as a powerful tool to advance patient safety, EHRs have inadvertently spawned new
types of medical errors that are inadequately monitored or regulated. In a 2012 report on this
subject, the noted that vendor-imposed non-transparency clauses make it difficult to quantify
and track EHR-related medical errors. The IOM also noted that safety monitoring is hindered
by a lack of consensus regarding EHR quality measures and the absence of an organization
with the power to exercise oversight.18
Vendor gag clauses not only limit the ability of researchers, providers, and regulators to
access data regarding EHR glitches that may jeopardize patient safety,18, 72, 83, 84 but they also
discourage providers from participating in surveys regarding usability. Many EHR vendors
collect user-generated information about software glitches for their own purposes, but they
regard the information as proprietary. Because few vendors share their observations with the
U.S. Food and Drug Administration (FDA), the agency lacks the ability to identify problems
and recommend (or, if need be, require) solutions.72
In a recent effort to promote self-regulation of the industry, some vendors have joined
together to propose an EHR Code of Conduct that drops the clause that hinders patient safety
reporting. However, adherence to the code is voluntary.83 While the industry ponders how to
respond, the number of EHR-related medical errors grows.18, 84
Two Platforms, Two Approaches: VistA and Epic
Two of the largest and most successful EHRs in the nation—the Veterans Health Information
Systems & Technology Architecture (VistA) and Epic, one of the most successful privately
financed EHRs on the market today—provide an interesting study in contrasts. Both systems
are based on versions of the same antiquated but powerful programming language (MUMPS),
but their differing origins, business models, and trajectories offer useful insights into the development, diffusion, and potential future of EHRs.
34
Redirecting Innovation in U.S. Health Care: Case Studies
VistA
The archetype of an enterprise-wide EHR solution, Vista is installed throughout the Veterans
Healthcare Administration, the largest integrated delivery system in the United States. Intensively studied since its inception, VistA has received numerous accolades for reducing costs,
improving patient care, and enhancing clinical outcomes in the VA health care system.3, 85
Because its architecture is designed to enable the ready retrieval and analysis of clinical data, it
is a powerful tool for quality improvement and health services research.
Systemwide Benefits Demonstrated
Developed in its present form in the mid- to late 1990s, VistA was the first EHR to demonstrate the benefits of adopting an enterprise-wide health information system. It also provided
users and observers with the first sense of what EHRs can do for the quality and efficiency of
care. In 2009, nearly half of hospitals with enterprise-wide (inpatient and outpatient) health
IT systems in the United States used VistA or VistA derivatives.86 Today, VistA and derivative
EHRs, such as the Resource and Patient Management System (RPMS), are installed in every
Veterans Health Administration (VHA) facility in the country, as well as in most of the Indian
Health Service and a number of private health care facilities.
VistA originated in the Office of Data Management and Telecommunications in the late
1970s, when individual VA medical centers began simultaneous development of computerized
solutions for prescription management, patient scheduling, patient registration, patient discharge, and radiology.3 This prompted VA developers to come together to agree on a common
coding language and dictionary, as well as programming standards that allowed design code
portability across systems. Known at the time as the Decentralized Hospital Computer Program, this collaborative, nationwide effort was kicked off in 1983. It was implemented in waves
consisting of 25–100 sites at a time that adopted select applications.3
The pace and scope of EHR adoption increased dramatically under the leadership of
Dr. Kenneth W. Kizer, who served as the VA’s Undersecretary for Health from 1994 through
1999. Dr. Kizer considered installation of a major system upgrade to be a core element in his
effort to transform the organization and delivery of care throughout the entire VA health care
system, along with changing the VA’s payment system, implementing a national performance
management system, instituting universal primary care, and other strategies.3, 87, 88 The most
significant aspect of the VA’s IT upgrade was the addition of a graphical user interface for health
care providers known as the Computerized Patient Record System (CPRS), which served as
the “face” of VistA’s EHR. Systemwide implementation of CPRS-VistA began in early 1997.
VistA incorporated, and in some instances pioneered, a number of useful features, including computerized order entry, electronic prescribing, bar code medication administration, and
embedded clinical guidelines. With VistA, a VA health care provider could easily review and
update a patient’s electronic medical record and share this information with any other health
care provider in the VA’s national system.3 VHA-wide implementation of VistA took place
between February 1997 and December 1999. It was (and remains) the largest and most rapid
deployment of an EHR ever done.88
VistA Credited with Improving Veterans’ Health88
The positive impact of VistA has been documented in outpatient chronic disease management;89 in the VHA’s systemwide Quality Enhancement Research Initiative (QUERI) processes;90, 91 and in overall assessments of performance indicators of preventive, outpatient, and
Case Study 3: Electronic Health Records
35
inpatient care.92 The open-source design of the system encouraged innovation without sacrificing interoperability. For example, the VA’s Cardiovascular Assessment, Reporting, and Tracking System for Cath Labs system for supporting cardiac catheterizations in the VA system
improved quality and safety while simultaneously creating a large data repository for research.93
The product of an $8.5 billion federal investment over roughly 30 years, VistA was part
of a broad reorientation and redesign of the VHA.94 The bet paid off. Between 1996 and 2004,
the VHA accommodated nearly 70 percent more patients, lowered its employee-to-patient
ratio by 37 percent, and decreased its overall costs per patient at a time when Medicare and
private insurance costs were dramatically rising.94 VistA was praised for helping the VHA
improve patient care, boost efficiency, lower costs, and strengthen its quality improvement and
research infrastructure.90–94
VistA’s success has been attributed in large part to the collaborative nature of its development. Clinicians and IT experts worked together to design its user interfaces and patient record
system. Equally if not more important, because VistA is built on a standard code and maintains data-sharing capability between sites, it is fully interoperable within the VA.
Independent surveys suggest that physicians are broadly satisfied with VistA compared
with other EHR systems. In a 2011 survey by the American Academy of Family Physicians
(AAFP)125 and a similar 2012 poll by Medscape,126 VistA outscored the large majority of health
IT solutions, particularly those offered by large vendors like Epic and McKesson. When the
AAFP survey asked respondents to express their level of agreement with the statement, “This
EHR enables me to practice higher quality medicine than I could with paper charts,” VistA
received the top score. In both the AAFP and Medscape survey’s, VistA was one of the highestranking systems overall.
OpenVista
In 2002, two venture capital–supported entrepreneurs, with guidance from Dr. Kizer, founded
Medsphere Systems Corporation. Their goal was to market OpenVista, a version of VistA
adapted for commercial use, in the private sector. The company continues to spend millions
annually to refine this system to meet the needs of nonfederal hospitals.
Some prospective users of OpenVista have expressed concern about its antiquated coding
language (interestingly, the same basic coding language used by Epic), potential maintenance
costs, and warnings from competing vendors that the platform was not designed from the
ground up to support fee-for-service billing. A 2005 effort by CMS to offer a free version of
VistA to support private doctors’ offices was unsuccessful.105, 106
Despite these obstacles, the usability, affordability, and documented impact of VistA have
enabled Medsphere to gain a foothold in the private health care sector. Currently, more than
25 private hospitals use OpenVista, including large urban teaching facilities, such as Lutheran
Medical Center in Brooklyn. OpenVista is deployed in a wide variety of facilities, including an 11-bed Guadalupe County Hospital in New Mexico, the Silver Hill psychiatric hospital in Connecticut, and the nine hospitals that comprise the West Virginia state hospital
system. Medsphere also developed and supports the Resource and Patient Management System
(RPMS), a derivative of VistA. It is currently deployed in more than 370 clinical sites of the
federal Indian Health Service.127
36
Redirecting Innovation in U.S. Health Care: Case Studies
Cost Is a Major Differentiator Between OpenVista and Other Private EHRs
According to Medsphere, no hospital has spent more than $3.5 million over a five-year period
on OpenVista implementation and ongoing maintenance. All of Medsphere’s clients who are
eligible for federal incentives have achieved Stage 1 of meaningful use, and almost all will cover
their initial five-year costs for installing OpenVista through their incentive payments.
Epic
Epic is rapidly becoming an industry leader in the EHR marketplace, particularly among
large academic medical centers and multispecialty group practices. A privately held company,
it develops, installs, supports, and owns its proprietary technology. Among its 260-plus clients
are most of the United States’ elite academic medical centers, including the Cleveland Clinic,
Johns Hopkins, Dartmouth-Hitchcock Medical Center, Kaiser Permanente, and nearly the
entire University of California system. Epic has won several awards for both inpatient and outpatient EHRs, as well as for software used in scheduling, billing, and collections. It has been
named “The Top EHR Vendor by Number of Meaningful Use Attestations.”96
Started in 1979 as a small company of 1.5 employees called Human Services Computing, Inc., Epic has been led from the beginning by its founding CEO, Judith Faulkner.97 Over
the subsequent decades, Faulkner turned an initial investment of $70,000 in private funds
into a global company with approximately $1.5 billion in revenue in 2012.98 Epic’s breakout
moment came in 2003, when Kaiser Permanente chose it over two much larger companies—
IBM and Cerner—to provide the EHR for its 36 hospitals and 8+ million members.99 According to a former VA official, Kaiser’s leadership also considered VistA, but soon abandoned
the idea because there was no commercial vendor at the time to support implementation and
maintenance.127
In contrast to many startup companies, Faulkner built Epic without relying on outside
capital and with almost no marketing. Because it has remained a privately held company, it
does not disclose its earnings. But industry analysts note that it is clearly profitable and has
zero debt.98 Much of Epic’s growth comes through word-of-mouth advertising built on a client
base of elite academic institutions.97 The formula has worked brilliantly. According to a recent
article in Forbes magazine, “By next year [2013], 127 million patients or nearly 40% of the
U.S. population will have its medical information stored in an Epic digital record.”98 In Medscape’s 2012 survey of EHRs, Epic was the most-used system by a wide margin.126
Closed Platform Makes Interfacing Epic Challenging
In contrast to VistA, which is in the public domain, Epic is a closed platform. It can be challenging and costly for hospitals to interface their EHR with the clinical or billing software of
other companies. Forbes observed, “In addition to the software, [Epic] customers pay dearly
for hardware, and for the army of Epic-certified technicians that needs to be deployed to get
the system up and running.”98 Although Epic is expensive, it works, and in the conservative
world of health IT, that’s all that matters. Another Forbes article put it this way: “Epic is built
on a traditional client server model, and does individual, customized installations for each
client; a reputation for near-flawless implementation—derived by tightly constraining how
much idiosyncrasy is engineered in each install—has been a prime driver of growth.”101
Given its strong market position, Epic was well positioned to compete for new contracts
when the HITECH Act of 2009 put billions of dollars in federal incentive payments into play.
As the only head of a health IT company appointed to serve on the Obama administration’s
Case Study 3: Electronic Health Records
37
Federal Health IT Policy Committee, Faulkner was able to advise the national coordinator for
health IT and other federal officials as they crafted a policy framework to develop a nationwide health information infrastructure, including the criteria that HHS ultimately adopted to
satisfy HITECH’s requirement that IT platforms be able to exchange health information.98, 102
Today, Epic is a market leader in helping users of its system meet meaningful use criteria.96 Several studies have described various ways in which Epic contributes to improvements
in care. Examples include enabling providers to identify high-risk patients prior to performing
interventions, thereby reducing nosocomial infections,103 and use antibiotics more appropriately.104 In a randomized controlled trial of Epic in primary care sites, Epic’s Best Practice and
SmartSet platforms were able to decrease prescribing of heavily marketed pharmaceuticals and
lead to more–cost-effective care.105 Epic has been credited with playing an important role in
quality improvement efforts in the Kaiser Permanente system, including use of best practice
alerts, standardized order sets, and chart abstraction tools.106 Epic has been shown to decrease
redundancy of testing and utilization.107, 108
However, not all evaluations of Epic’s impact have been positive.43, 47, 104–107, 109–112 An independent evaluation of Epic’s impact in the Kaiser system found that implementation led to
efficiency losses and a persistent two-minute increase in the length of time of an average patient
encounter.111 Implementation of Epic in the Hennepin Health system did not change outcomes
for critically ill patients. Moreover, physicians complained of workflow interruptions, slower
processes of care, and excessive time with the provider’s “back to the patient” because of the
need to focus on computerized order entry.47
Epic’s Limited Interoperability Is a Major Concern
One of the biggest concerns with Epic is its relative lack of interoperability.100 Although the
company has a strict structure and retains tight control of its software and data, it does customized installations for each client. This allows health care systems to tailor Epic’s applications and functionality to meet their own needs. Customizability is attractive to hospital
CEOs, CFOs, and some members of the medical staff, but it can limit interoperability between
sites and the capacity to communicate with providers who use other EHRs.
While there is potential for EHRs with functional data linkages to be used for research
purposes, currently the lack of interoperability between EHRs hinders research capabilities.113
Furthermore, many EHR fields contain free text, which makes subsequent extraction and
analysis difficult.114 To circumvent these limitations, the Cleveland Clinic Neurologic Institute
developed a distinct program that is incorporated into the patient’s workflow in order to facilitate data collection for research.114
To address concerns about lack of interoperability, Epic developed Care Everywhere. The
feature allows Epic users to communicate with other health systems’ EHRs, although it works
best if both parties have Epic systems. According to the company’s website,
Care Everywhere provides a framework for interoperability, so that wherever the patient
goes—between healthcare systems in the same town or across state and national borders—
the clinicians providing care can have the information they need. Information can come
from another Epic system, a non-Epic EMR that complies with industry standards, or
directly from the patient. When an Epic system is on both sides of the exchange, a richer
data set is exchanged and additional connectivity options, such as cross-organization referral management, are available. Regardless of the information source, Care Everywhere connects it to your EpicCare EMR, giving clinicians a more complete clinical record.128
38
Redirecting Innovation in U.S. Health Care: Case Studies
Epic’s hold on America’s top hospitals is powerful and increasingly pervasive. Numerous
theories for Epic’s dominance abound, but its market share among leading academic medical
centers is undeniable.115 It is less clear what clients are getting in return. Last year, a writer for
Forbes offered the following observations:
She [Judy Faulkner, Epic CEO] has quietly convinced them [hospital chief information
officers] that her product is best: a single, seamless database—the fruit of a company that
has grown organically, and shunned acquisitions. And, because it is no small task to deploy,
she is there all the way to hand-hold jittery CIOs, and help them get millions of dollars in
government subsidies by showing meaningful use of her EHR.
Her not-for-profit clientèle will need every penny of those taxpayers’ dollars, but they won’t
cover anywhere near the staggering cost of an Epic EHR. Duke University Health System
will shell out $700 million, so will Boston-based Partners HealthCare; University of California, San Francisco will pay $150 million.97
It is impossible to independently verify the accuracy of the sums quoted in the Forbes
article. Health IT experts contend that a health care system should devote 3–4 percent of its
annual budget to IT. Given the annual revenues of some of these systems, the reported levels
of spending in the Forbes article are more or less in line with that figure. However, it should
also be noted that capital expense is only part of an EHR’s cost. Many hospital administrators
fail to consider the additional spending required to train (and retrain) personnel, the downtime
involved in learning a new system, the costs of maintenance and the inevitable upgrades, and
other hidden costs. As a result, it is difficult to determine the true costs of IT adoption from
public data.129
The EHR “Productivity Paradox”
David Brailer, the first national coordinator for health information technology at HHS, once
described VistA as “probably one of the most stress-tested, and life-tested electronic health
records in the world.”95 It is an ongoing success story in the VA health care system. However,
despite its relative ease of use, interoperability, and impact, VistA’s private-sector counterpart,
OpenVista, has experienced limited success.
While OpenVista has had relatively few takers, Epic has established itself as the enterprisewide solution of choice for large private health care systems and academic medical centers, irrespective of ongoing concerns about its limited interoperability and less-than-ideal usability. As
a result, Epic has been able to capitalize on the recent surge in federal incentive payments to
encourage doctors and health care systems to adopt or upgrade their EHR to meet the government’s meaningful use standards.
If health care systems continue the current trend of consolidating into large integrated
delivery systems, and the small hospitals and physician practices that have been sitting on the
fence get drawn in, Epic could become the dominant vendor of EHRs in the United States. Two
technology writers recently envisioned a day when “Epic will be health IT’s (HIT’s) Roman
Empire: establishing the laws and the language for “known world”, as well as the underlying
infrastructure, and ends up shaping information flows, IT architecture and—potentially—the
eventual configuration of provider systems.”101
Case Study 3: Electronic Health Records
39
Alternatively, American healthcare will adapt to health IT’s “productivity paradox” the
way financial services, airlines, and other industries did before it.39 The same pair of technology
writers described what might happen next:
The second group of scenarios—we call them “Open Data”—assume that progress on
interoperability standards, together with Epic’s focus on convincing large clients to adopt
its functional but hardly delightful software, will create an opportunity for an agile competitor to compete effectively for the long tail customers, especially those who’ve deliberately avoided participation in a large hospital system, and might especially value an EMR
system that offers a great user experience, and doesn’t distance the physician from the
patient the way most [EHRs] tend to do.101
There are signals that pressure is growing on EHR vendors to make their systems more
interoperable, usable, and patient-centered. Recently, several of Epic’s top competitors, including McKesson and Cerner, signed an agreement that they will work together to improve
interoperability between their systems.108 It is too early to know what will happen next, but
two things are clear: (1) EHRs are here to stay; and (2) how they are designed and employed
will profoundly influence the quality, efficiency, and cost of American health care for decades
to come.
References
1
Hayrinen K, Saranto K, and Nykanen P, “Definition, Structure, Content, Use and Impacts of Electronic
Health Records: A Review of the Research Literature,” International Journal of Medical Informatics, Vol. 77,
No. 5, May 2008, pp. 291–304.
2
Bates DW, Leape LL, Cullen DJ, Laird N, Petersen LA, Teich JM, Burdick E, Hickey M, Kleefield S,
Shea B, Vliet MV, and Seger DL, “Effect of Computerized Physician Order Entry and a Team Intervention on
Prevention of Serious Medication Errors,” Journal of the American Medical Association, Vol. 280, No. 15, 1998,
pp. 1311–1316.
3
Brown SH, Lincoln MJ, Groen PJ, and Kolodner RM, “VistA—U.S. Department of Veterans Affairs
National-Scale HIS,” International Journal of Medical Informatics, Vol. 69, No. 2–3, March 2003,
pp. 135–156.
4 Slack WV, Hicks GP, Reed CE, and Van Cura LJ, “A Computer-Based Medical-History System,” New England Journal of Medicine, Vol. 274, No. 4, January 27, 1966, pp. 194–198.
5
Brodman K, Erdmann AJ, Jr., et al., “The Cornell Medical Index; An Adjunct to Medical Interview,” Journal of the American Medical Association, Vol. 140, No. 6, June 11, 1949, pp. 530–534.
6
Jacobs L, “Interview with Lawrence Weed, MD—The Father of the Problem-Oriented Medical Record
Looks Ahead,” The Permanente Journal, Vol. 13, No. 3, Summer 2009, pp. 84–89.
7
Weed LL, “Medical Records That Guide and Teach,” New England Journal of Medicine, Vol. 278, No. 12,
March 21, 1968, pp. 652–657 concl.
8
Collen M, A History of Medical Informatics in the United States, 1950 to 1990, Indianapolis, Ind.: American
Medical Informatics Association, 1995.
9
Balch POM, “Computer Ombudsman: A New Role for the 80’s,” Proceedings of the Annual Symposium on
Computer Application in Medical Care, Vol. 5, No. 3, November 5, 1980, pp. 1709–1714.
40
Redirecting Innovation in U.S. Health Care: Case Studies
10
Castleman P, “On-Line Data Management System for the Massachusetts General Hospital,” Proceedings of
the 2nd Symposium on Computer-Centered Data Bases, Vol. 3, 1965, pp. 143–183.
11
The Computer-Based Patient Record, Washington, D.C.: Institute of Medicine, 1997.
12
Miller RH and Sim I, “Physicians’ Use of Electronic Medical Records: Barriers and Solutions,” Health
Affairs, Vol. 23, No. 2, March–April 2004, pp. 116–126.
13
Garrett TJ and Ashford AR, “Computer-Assisted Instruction in Patient Management for Internal Medicine
Residents,” Journal of Medical Education, Vol. 61, No. 12, December 1986, pp. 987–989.
14
Rogers JL and Haring OM, “The Impact of a Computerized Medical Record Summary System on Incidence and Length of Hospitalization,” Medical Care, Vol. 17, No. 6, June 1979, pp. 618–630.
15
Rogers JL, Haring OM, Wortman PM, Watson RA, and Goetz JP, “Medical Information Systems: Assessing Impact in the Areas of Hypertension, Obesity and Renal Disease,” Medical Care, Vol. 20, No. 1, January 1982, pp. 63–74.
16 United States General Accounting Office, Automated Medical Records Hold Promise to Improve Patient Care,
Washington, D.C., January 1991.
17
Kohn LT CJ and Donaldson MS, To Err Is Human: Building a Safer Health System, Washington, D.C.:
National Academies Press, Institute of Medicine, 2000.
18 Institute of Medicine, Health IT and Patient Safety, Building Safer Systems for Better Care, Washington,
D.C., 2012.
19
Hillestad R, Bigelow J, Bower A, et al., “Can Electronic Medical Record Systems Transform Health Care?
Potential Health Benefits, Savings, and Costs,” Health Affairs, Vol. 24, No. 5, September–October 2005,
pp. 1103–1117.
20
Walker J, Pan E, Johnston D, Adler-Milstein J, Bates DW, and Middleton B, “The Value of Health Care
Information Exchange and Interoperability,” Health Affairs, January–June 2005, Supplemental Web Exclusives, pp. W5-10–W15-18.
21
Jha AK, Doolan D, Grandt D, Scott T, and Bates DW, “The Use of Health Information Technology in
Seven Nations,” International Journal of Medical Informatics, Vol. 77, No. 12, December 2008, pp. 848–854.
22
Jackson D, “80% of Hospitals Embrace Electronic Records,” USA Today, July 17, 2013.
23
DesRoches CM, Charles D, Furukawa MF, et al., “Adoption of Electronic Health Records Grows Rapidly,
But Fewer Than Half of US Hospitals Had at Least a Basic System in 2012,” Health Affairs, July 9, 2013.
24
DesRoches CM, Audet AM, Painter M, and Donelan K, “Meeting Meaningful Use Criteria and Managing Patient Populations: A National Survey of Practicing Physicians,” Annals of Internal Medicine, Vol. 158,
No. 11, June 4, 2013, pp. 791–799.
25 Bates DW, “Physicians and Ambulatory Electronic Health Records,” Health Affairs, Vol. 24, No. 5,
September–October 2005, pp. 1180–1189.
26 DesRoches
CM, Worzala C, Joshi MS, Kralovec PD, and Jha AK, “Small, Nonteaching, and Rural Hospitals Continue to Be Slow in Adopting Electronic Health Record Systems,” Health Affairs, Vol. 31, No. 5,
May 2012, pp. 1092–1099.
27
Rao SR, DesRoches CM, Donelan K, Campbell EG, Miralles PD, and Jha AK, “Electronic Health Records
in Small Physician Practices: Availability, Use, and Perceived Benefits,” Journal of the American Medical Informatics Association, Vol. 18, No. 3, May 1, 2011, pp. 271–275.
28
Sittig DF and Singh H, “Legal, Ethical, and Financial Dilemmas in Electronic Health Record Adoption
and Use,” Pediatrics, Vol. 127, No. 4, April 2011, pp. e1042–e1047.
Case Study 3: Electronic Health Records
41
29
Abdolrasulnia M, Menachemi N, Shewchuk RM, Ginter PM, Duncan WJ, and Brooks RG, “Market
Effects on Electronic Health Record Adoption by Physicians,” Health Care Management Review, Vol. 33,
No. 3, July–September 2008, pp. 243–252.
30
Valdes I, Kibbe DC, Tolleson G, Kunik ME, and Petersen LA, “Barriers to Proliferation of Electronic Medical Records,” Informatics in Primary Care, Vol. 12, No. 1, 2004, pp. 3–9.
31
Menachemi N, Burke D, and Brooks RG, “Adoption Factors Associated with Patient Safety-Related Information Technology,” Journal for Healthcare Quality, Vol. 26, No. 6, November–December 2004, pp. 39–44.
32
Title XIII—Health Information Technology for Economic and Clinical Health (HITECH) Act, 2009.
33
Radnofsky L, “Electronic Health Data Gaining Favor,” Wall Street Journal, May 22, 2013, p. A2.
34 Mostashari
F, Brailer D, and Mostashari F, “Two National Health IT Czars Compare Notes,” Health
Affairs, Vol. 31, No. 3, March 2012, pp. 475–480.
35
Kellermann AL and Jones SS, “What It Will Take to Achieve the As-Yet-Unfulfilled Promises of Health
Information Technology,” Health Affairs, Vol. 32, No. 1, January 2013, pp. 63–68.
36
American Recovery and Reinvestment Act of 2009, 2009.
37
Hare J, “Hearing on Efforts to Promote the Adoption and Meaningful Use of Health Information Technology,” testimony before the Subcomittee on Health, Committee on Ways and Means, U.S. House of Representatives, Washington, D.C., July 20, 2010.
38
Menachemi N and Collum TH, “Benefits and Drawbacks of Electronic Health Record Systems,” Risk
Management and Healthcare Policy, Vol. 4, 2011, pp. 47–55.
39
Jones SS, Heaton PS, Rudin RS, and Schneider EC, “Unraveling the IT Productivity Paradox—Lessons for
Health Care,” New England Journal of Medicine, Vol. 366, No. 24, June 14, 2012, pp. 2243–2245.
40 Bates
DW, Leape LL, Cullen DJ, et al., “Effect of Computerized Physician Order Entry and a Team Intervention on Prevention of Serious Medication Errors, Journal of the American Medical Association, Vol. 280,
No. 15, October 21, 1998, pp. 1311–1316.
41
Evans RS, Pestotnik SL, Classen DC, and Burke JP, “Evaluation of a Computer-Assisted Antibiotic-Dose
Monitor,” Annals of Pharmacotherapy, Vol. 33, No. 10, October 1999, pp. 1026–1031.
42 Evans
RS, Pestotnik SL, Classen DC, et al., “A Computer-Assisted Management Program for Antibiotics and Other Antiinfective Agents,” New England Journal of Medicine, Vol. 338, No. 4, January 22, 1998,
pp. 232–238.
43
Chaudhry B, Wang J, Wu S, et al., “Systematic Review: Impact of Health Information Technology on
Quality, Efficiency, and Costs of Medical Care,” Annals of Internal Medicine, Vol. 144, No. 10, May 16, 2006,
pp. 742–752.
44 Kucher
N, Koo S, Quiroz R, et al., “Electronic Alerts to Prevent Venous Thromboembolism Among Hospitalized Patients,” New England Journal of Medicine, Vol. 352, No. 10, March 10, 2005, pp. 969–977.
45
Willson D, Ashton C, Wingate N, et al., “Computerized Support of Pressure Ulcer Prevention and Treatment Protocols,” Proceedings of the Annual Symposium on Computer Application in Medical Care, 1995,
pp. 646–650.
46 Black
AD, Car J, Pagliari C, et al., “The Impact of eHealth on the Quality and Safety of Health Care: A
Systematic Overview,” PLoS Medicine, Vol. 8, No. 1, 2011, e1000387.
47
Brunette DD, Tersteeg J, Brown N, et al., “Implementation of Computerized Physician Order Entry for
Critical Patients in an Academic Emergency Department Is Not Associated with a Change in Mortality Rate,”
Western Journal of Emergency Medicine, Vol. 14, No. 2, March 2013, pp. 114–120.
48
Cebul RD, Love TE, Jain AK, and Hebert CJ, “Electronic Health Records and Quality of Diabetes Care,”
New England Journal of Medicine, Vol. 365, No. 9, September 1, 2011, pp. 825–833.
42
Redirecting Innovation in U.S. Health Care: Case Studies
49
Keyhani S, Hebert PL, Ross JS, Federman A, Zhu CW, and Siu AL, “Electronic Health Record Components and the Quality of Care,” Medical Care, Vol. 46, No. 12, December 2008, pp. 1267–1272.
50
Linder JA, Ma J, Bates DW, Middleton B, and Stafford RS, “Electronic Health Record Use and the Quality of Ambulatory Care in the United States,” Archives of Internal Medicine, Vol. 167, No. 13, July 9, 2007,
pp. 1400–1405.
51
Romano MJ and Stafford RS, “Electronic Health Records and Clinical Decision Support Systems: Impact
on National Ambulatory Care Quality,” Archives of Internal Medicine, Vol. 171, No. 10, May 23, 2011,
pp. 897–903.
52
Zhou L, Soran CS, Jenter CA, et al., “The Relationship Between Electronic Health Record Use and
Quality of Care Over Time,” Journal of the American Medical Informatics Association, Vol. 16, No. 4, July–
August 2009, pp. 457–464.
53
Bailey JE, Pope RA, Elliott EC, Wan JY, Waters TM, and Frisse ME, “Health Information Exchange
Reduces Repeated Diagnostic Imaging for Back Pain,” Annals of Emergency Medicine, Vol. 62, No. 1,
July 2013, pp. 16–24.
54
Sidorov J, “It Ain’t Necessarily So: The Electronic Health Record and the Unlikely Prospect of Reducing
Health Care Costs,” Health Affairs, Vol. 25, No. 4, July–August 2006, pp. 1079–1085.
55
Agrawal A, “Return on Investment Analysis for a Computer-Based Patient Record in the Outpatient Clinic
Setting,” Journal of the Association for Academic Minority Physicians, Vol. 13, No. 3, July 2002, pp. 61–65.
56
Erstad TL, “Analyzing Computer Based Patient Records: A Review of Literature,” Journal of Healthcare
Information Management, Vol. 17, No. 4, Fall 2003, pp. 51–57.
57
Mildon J and Cohen T, “Drivers in the Electronic Medical Records Market,” Health Management Technology, Vol. 22, No. 5, May 2001, pp. 14–16, 18.
58
Bhattacherjee A HN, Menachemi N, et al., “The Differential Performance Effects of Healthcare Information Technology Adoption,” Information Systems Management, Vol. 24, No. 1, 2007, pp. 5–14.
59
Brynjolfsson E and Hitt L, “Beyond the Productivity Paradox: Computers Are the Catalyst for Bigger
Changes,” Communications of the Association for Computing Machinery, Vol. 41, 1998, pp. 49–55.
60 Aspden
P, Patient Safety: Achieving a New Standard for Care, National Academies Press, 2004.
61
Belletti D, Zacker C, and Mullins CD, “Perspectives on Electronic Medical Records Adoption: Electronic Medical Records (EMR) in Outcomes Research,” Patient Related Outcome Measures, Vol. 1, July 2010,
pp. 29–37.
62
Dexter PR, Perkins S, Overhage JM, Maharry K, Kohler RB, and McDonald CJ, “A Computerized
Reminder System to Increase the Use of Preventive Care for Hospitalized Patients,” New England Journal of
Medicine, Vol. 345, No. 13, September 27, 2001, pp. 965–970.
63
Ledwich LJ, Harrington TM, Ayoub WT, Sartorius JA, and Newman ED, “Improved Influenza and Pneumococcal Vaccination in Rheumatology Patients Taking Immunosuppressants Using an Electronic Health
Record Best Practice Alert,” Arthritis and Rheumatism, Vol. 61, No. 11, November 15, 2009, pp. 1505–1510.
64 Kukafka
R, Ancker JS, Chan C, et al., “Redesigning Electronic Health Record Systems to Support Public
Health,” Journal of Biomedical Informatics, Vol. 40, No. 4, August 2007, pp. 398–409.
65
Institute of Medicine Committee on Effectivness of National Biosurveillance Systems, Biowatch and Public
Health Surveillance: Evaluating Systems for the Early Detection of Biologic Threats, Washington, D.C., 2010.
66 Hare
J, “It’s About the Network, Not the App,” Health Management Technology, 2011. As of January 14,
2014:
http://www.healthmgttech.com/articles/201105/how-do-we-get-to-meaningful-use-part-2.php
67
The Direct Project Overview, undated. As of January 14, 2014:
http://wiki.directproject.org/file
Case Study 3: Electronic Health Records
43
68
Williams C, Mostashari F, Mertz K, Hogin E, and Atwal P, “From the Office of the National Coordinator:
The Strategy for Advancing the Exchange of Health Information,” Health Affairs, Vol. 31, No. 3, March 2012,
pp. 527–536.
69
Kohane IS and Altman RB, “Health-Information Altruists—A Potentially Critical Resource,” New England
Journal of Medicine, Vol. 353, No. 19, November 10, 2005, pp. 2074–2077.
70 Miller AT, “Privacy Protection and Technology Diffusion: The Case of Electronic Medical Records,” Management Science, Vol. 55, No. 7, July 2009, p. 17.
71
Mandl KD and Kohane IS, “Escaping the EHR Trap—The Future of Health IT,” New England Journal of
Medicine, Vol. 366, No. 24, June 14, 2012, pp. 2240–2242.
72 McDonnell CW and Wendel L, Electronic Health Record Usability, Rockville, Md.: Agency for Healthcare
Research and Quality, May 2010.
73
Cearnal L, “Electronic Medical Records Link to Upcoding Under Fire,” Annals of Emergency Medicine,
Vol. 61, No. 4, April 2013, p. 3.
74
Goldzweig CL, Towfigh A, Maglione M, and Shekelle PG, “Costs and Benefits of Health Information Technology: New Trends from the Literature,” Health Affairs, Vol. 28, No. 2, March–April 2009,
pp. w282–w293.
75
Grabenbauer L, Skinner A, and Windle J, “Electronic Health Record Adoption—Maybe It’s Not About
the Money: Physician Super-Users, Electronic Health Records and Patient Care,” Applied Clinical Informatics,
Vol. 2, No. 4, 2011, pp. 460–471.
76
Phansalkar S, Desai A, Bell DS, et al., “High-Priority Drug-Drug Interactions for Use in Electronic Health
Records,” American Medical Informatics, Vol. 19, No. 5, September 2012, pp. 735–743.
77 Jones
SS, Koppel R, Ridgely MS, Palen TE, Wu S, and Harrison MI, “Guide to Reducing Unintended
Consequences of Electronic Health Records,” Santa Monica, Calif.: RAND Corporation, EP-201100-261,
2011. As of January 14, 2014:
http://www.rand.org/pubs/external_publications/EP201100261.html
78
Weber GM, Murphy SN, McMurry AJ, et al., “The Shared Health Research Information Network
(SHRINE): A Prototype Federated Query Tool for Clinical Data Repositories,” Journal of the American Medical Informatics Association, Vol. 16, No. 5, September–October 2009, pp. 624–630.
79
HealthIT.gov, “The Blue Button Connector,” undated. As of January 15, 2014:
http://www.healthit.gov/bluebutton
80
Mosquera M, “Mostashari Calls for Vendors to Add Blue Button Quickly,” Government Health IT, 2012. As
of January 14, 2014:
http://www.govhealthit.com/news/mostashari-calls-vendors-add-blue-button-quickly
81
World Economic Forum, Personal Data: The Emergence of a New Asset Class, Geneva, Switzerland, 2011.
82
Potini VC, Weerasuriya DN, Lowery-North DW, and Kellermann AL, “Commercial Products That
Convey Personal Health Information in Emergencies,” Disaster Medicine and Public Health Preparedness,
Vol. 5, No. 4, December 2011, pp. 261–265.
83
Conn J, “EHR Trade Group Issues Code of Conduct to Improve Patient Safety,” ModernHealthcare.com,
2013.
84 Magrabi
F, Ong MS, Runciman W, and Coiera E, “Patient Safety Problems Associated with Heathcare
Information Technology: An Analysis of Adverse Events Reported to the US Food And Drug Administration,” AMIA Symposium, 2011, pp. 853–857.
85
Rundle R,” In the Drive to Mine Medical Data, VHA Is the Unlikely Leader,” Wall Street Journal, 2001,
p. 1.
44
Redirecting Innovation in U.S. Health Care: Case Studies
86
Jha AK, DesRoches CM, Campbell EG, et al., “Use of Electronic Health Records in U.S. Hospitals,” New
England Journal of Medicine, Vol. 360, No. 16, April 16, 2009, pp. 1628–1638.
87
Iglehart JK, “Reform of the Veterans Affairs Heath Care System,” New England Journal of Medicine,
Vol. 335, No. 18, October 31, 1996, pp. 1407–1411.
88
Kizer KW and Dudley RA, “Extreme Makeover: Transformation of the Veterans Health Care System,”
Annual Review of Public Health, Vol. 30, 2009, pp. 313–339.
89
McMahon GT, Gomes HE, Hickson Hohne S, Hu TM, Levine BA, and Conlin PR, “Web-Based
Care Management in Patients with Poorly Controlled Diabetes,” Diabetes Care, Vol. 28, No. 7, July 2005,
pp. 1624–1629.
90 Hynes
DM, Weddle T, Smith N, Whittier E, Atkins D, and Francis J, “Use of Health Information Technology to Advance Evidence-Based Care: Lessons from the VA QUERI Program,” Journal of General Internal
Medicine, Vol. 25, Supplement 1, January 2010, pp. 44–49.
91
Stetler CB, Mittman BS, and Francis J, “Overview of the VA Quality Enhancement Research Initiative
(QUERI) and QUERI Theme Articles: QUERI Series,” Implementation Science, Vol. 3, 2008, p. 8.
92
Jha AK, Perlin JB, Kizer KW, and Dudley RA, “Effect of the Transformation of the Veterans Affairs
Health Care System on the Quality of Care,” New England Journal of Medicine, Vol. 348, No. 22, May 29,
2003, pp. 2218–2227.
93
Box TL, McDonell M, Helfrich CD, Jesse RL, Fihn SD, and Rumsfeld JS, “Strategies from a Nationwide
Health Information Technology Implementation: The VA CART Story,” Journal of General Internal Medicine,
Vol. 25, Supplement 1, January 2010, pp. 72–76.
94
Evans DC, Nichol WP, and Perlin JB, “Effect of the Implementation of an Enterprise-Wide Electronic
Health Record on Productivity in the Veterans Health Administration,” Health Economics, Policy, and Law,
Vol. 1, Part 2, April 2006, pp. 163–169.
95 De LaGarza P and Nohlgren S, “World Turning to Software VA Shuns,” St. Petersburg Times, 2005. As of
January 14, 2014:
http://www.sptimes.com/2005/05/31/news_pf/Worldandnation/World_turning_to_soft.shtml
96
Richards B, “EHR Market Share Report Shows Top MU Vendors,” Journal of AHIMA, February 6, 2013.
97
Moukheiber Z, “The Staggering Cost of an Epic Electronic Health Record Might Not Be Worth It,” Forbes,
June 18, 2012.
98
Moukheiber Z, “Epic Systems’ Tough Billionaire,” Forbes, April 20, 2012.
99
Freudenheim M, “Digitizing Health Records, Before It Was Cool,” New York Times, January 14, 2012. As
of January 14, 2014:
http://www.nytimes.com/2012/01/15/business/
epic-systems-digitizing-health-records-before-it-was-cool.html?pagewanted=all&_r=1&
100 Allphin
M, Epic HIE 2014: Everywhere, Elsewhere, or Nowhere Else? KLAS Research, April 2014. As of
January 26, 2015:
http://www.klasresearch.com/Vendor/54#/krms/97/0
101Shaywitz
D and Wolf T, “Epic Challenge: What the Emergence of an EMR Giant Means for the Future of
Healthcare Innovation,” Forbes, June 9, 2012. As of January 20, 2014:
http://www.forbes.com/sites/davidshaywitz/2012/06/09/
epic-challenge-what-the-emergence-of-an-emr-giant-means-for-the-future-of-healthcare-innovation/
102Health
IT Policy Committee, “Federal Advisory Committees (FACAs),” undated. As of January 15, 2014:
http://www.healthit.gov/policy-researchers-implementers/health-it-policy-committee
103Cook
PP, Rizzo S, Gooch M, Jordan M, Fang X, and Hudson S, “Sustained Reduction in Antimicrobial
Use and Decrease in Methicillin-Resistant Staphylococcus Aureus and Clostridium Difficile Infections Fol-
Case Study 3: Electronic Health Records
45
lowing Implementation of an Electronic Medical Record at a Tertiary-Care Teaching Hospital,” Journal of
Antimicrobial Chemotherapy, Vol. 66, No. 1, January 2011, pp. 205–209.
104Kullar
R, Goff DA, Schulz LT, Fox BC, and Rose WE, “The ‘Epic’ Challenge of Optimizing Antimicrobial
Stewardship: The Role of Electronic Medical Records and Technology,” Clinical Infectious Diseases, June 6,
2013.
105Fortuna
RJ, Zhang F, Ross-Degnan D, et al., “Reducing the Prescribing of Heavily Marketed Medications: A Randomized Controlled Trial,” Journal of General Internal Medicine, Vol. 24, No. 8, August 2009,
pp. 897–903.
106Whippy
A, Skeath M, Crawford B, et al., “Kaiser Permanente’s Performance Improvement System, Part
3: Multisite Improvements in Care for Patients with Sepsis,” Joint Commission Journal on Quality and Patient
Safety/Joint Commission Resources, Vol. 37, No. 11, November 2011, pp. 483–493.
107Chen
C, Garrido T, Chock D, Okawa G, and Liang L, “The Kaiser Permanente Electronic Health Record:
Transforming and Streamlining Modalities of Care,” Health Affairs, Vol. 28, No. 2, March–April 2009,
pp. 323–333.
108O’Neill
L and Klepack W, “Electronic Medical Records for a Rural Family Practice: A Case Study in Systems Development,” Journal of Medical Systems, Vol. 31, No. 1, February 2007, pp. 25–33.
109Garrido
T, Raymond B, Jamieson L, Liang L, and Wiesenthal A, “Making the Business Case for Hospital
Information Systems—A Kaiser Permanente Investment Decision,” Journal of Health Care Finance, Vol. 31,
No. 2, Winter 2004, pp. 16–25.
110Gonzales
R, Anderer T, McCulloch CE, et al., “A Cluster Randomized Trial of Decision Support Strategies for Reducing Antibiotic Use in Acute Bronchitis,” JAMA Internal Medicine, Vol. 173, No. 4, February 25,
2013, pp. 267–273.
111Krall
MA, “Acceptance and Performance by Clinicians Using an Ambulatory Electronic Medical Record in
an HMO,” Proceedings of the Annual Symposium on Computer Application in Medical Care, 1995, pp. 708–711.
112Lindholm
C, Adsit R, Bain P, et al., “A Demonstration Project for Using the Electronic Health Record
to Identify and Treat Tobacco Users,” WMJ: Official Publication of the State Medical Society of Wisconsin,
Vol. 109, No. 6, December 2010, pp. 335–340.
113Bradley
CJ, Penberthy L, Devers KJ, and Holden DJ, “Health Services Research and Data Linkages: Issues,
Methods, and Directions for the Future,” Health Services Research, Vol. 45, No. 5, Part 2, October 2010,
pp. 1468–1488.
114Katzan
I SM, Doper C, Urchek J, Bielawski K, Dunphy C, Jehi L, Bae C, and Parchman A, “The Knowledge Program: An Innovative, Comprehensive Electronic Data Capture System and Warehouse,” AMIA
Annual Symposium Proceedings, 2011, pp. 683–692.
115Halamka
J, “The Era of Epic,” 2013. As of January 18, 2014:
http://geekdoctor.blogspot.com/2013/07the-era-of-epic.html
116DesRoches
CM, Campbell EG, Sowmya RR, et al., “Electronic Health Records in Ambulatory Care—A
National Survey of Physicians,” New England Journal of Medicine, Vol. 359, July 3, 2008, pp. 50–60.
117Blumenthal
pp. 382–385.
D, “Launching HITECH,” New England Journal of Medicine, Vol. 362, No. 5, 2010,
118Frieden J, “Congress Dings EHR Privacy Regs,” MedPage Today, July 21, 2010. As of January 18, 2014:
http://www.medpagetoday.com/PracticeManagement/InformationTechnology/21285
119DesRoches
CM, Campbell EG, Vogeli C., et al., “Electronic Health Records in Ambulatory Care—A
National Survey of Physicians,” Health Affairs, Vol. 29, No. 4, 2010, pp. 639–646.
120O’Malley
AS, Grossman JM, Cohen GR, Kemper NM, and Pham HH, “Are Electronic Medical Records
Helpful for Care Coordination? Experiences of Physician Practices,” Journal of General Internal Medicine,
Vol. 25, No. 3, 2010, pp. 177–185.
46
Redirecting Innovation in U.S. Health Care: Case Studies
121Halamka
J, “Health Information Exchange for Emergency Department Care Is on the Right Trajectory,”
Annals of Emergency Medicine, Vol. 62, 2013, pp. 25–27.
122Bailey
JE, Pope RA, Elliott EC, et al., “Health Information Exchange Reduces Repeated Diagnostic Imaging for Back Pain,” Annals of Emergency Medicine, Vol. 62, 2013, pp. 16–24.
123Whitney
E, “Electronic Health Records Breed Digital Discontent for Some Docs,” Colorado Public Radio,
December 12, 2012. As of January 15, 2014:
http://www.kaiserhealthnews.org/stories/2012/december/04/doctors-discontent-electronic-health-records.aspx
124Rudin
RS, “How Electronic Health Records Can Lead to Care Coordination,” the RAND Blog,
August 13, 2013. As of January 15, 2014:
http://www.rand.org/blog/2013/08/how-electronic-health-records-can-lead-to-care-coordination.html
125Robert
L, Edsall RL, and Adler KG, “The 2011 EHR User Satisfaction Survey: Responses from 2,719
Family Physicians,” Family Practice Management, Vol. 18, No. 4, July–August 2011, pp. 23–30. As of January 16, 2014:
http://www.aafp.org/fpm/2011/0700/p23.html
126Medscape,
“EHR Report 2012: Physicians Rank Top EHRs,” August 23, 2012. As of January 16, 2014:
http://www.medscape.com/features/slideshow/EHR2012
127Kizer
K, personal communication, August 18, 2013.
128Epic,
“Epic Interoperability,” 2013. As of January 16, 2014:
http://www.epic.com/software-interoperability.php
129Gettinger
A, former CFO of Dartmouth Hitchcock Medical Center, personal communication, 2013.
Case Study 4
Haemophilus influenzae Type b (Hib) Vaccine
Ashaunta Tumblin, M.D.
Assistant Professor In-Residence of Pediatrics,
University of California, Riverside, School of Medicine;
Health Policy Researcher, RAND Corporation
The Technology
Haemophilus influenzae type b (Hib) vaccine is a childhood vaccine that protects against Hib
infection, a bacterial disease that was a major cause of death and disability in young children. The vaccine is composed of the capsular polysaccharide of the Hib bacterium chemically
bound to a carrier protein. When injected, the vaccine elicits a protective antibody response,
even in infants with immature immune systems. Protection is lifelong.
Rationale
In a third of a century, the development, adoption, and widespread use of Hib vaccine has
dramatically reduced the incidence of Hib infection in the United States and subsequently
in many other parts of the world. Before the 1970s, Hib disease affected one in 200 children
in the United States under the age of 5 years.1 Even after the advent of antibiotic treatment,
5 percent of American children who contracted systemic Hib disease died.2 Hib was the most
common cause of childhood meningitis; it also caused several other serious diseases, including
septicemia, epiglottitis, pneumonia, septic arthritis, osteomyelitis, and cellulitis. In the prevaccine era, one in 2,000 U.S. children under the age of 5 years contracted Hib meningitis.
Rates of Hib meningitis among African Americans were four times higher than among whites,
and among Native Americans, the rates were ten times higher than among whites.3–5 Approximately 30 percent of those who survived Hib meningitis suffered severe neurological sequelae,
including mental retardation, deafness, and seizures.6, 7 In fact, for many years, Hib was the
leading cause of acquired mental retardation in the United States.6
Meningitis was the most common manifestation of Hib disease, followed by epiglottitis.
In developing countries, Hib pneumonia is more common than meningitis.8 Compared with
children in developed nations, those in the developing world have higher incidences of Hib
infection at younger ages. They also experience far higher rates of mortality (20–40 percent in
sub-Saharan Africa).9 In these regions, 30–60 percent of children who contract Hib meningitis
die.10–12 It was estimated that in the pre-Hib vaccine era, nearly half a million cases of systemic
Hib disease occurred each year worldwide, with an average mortality rate of 23 percent.13 And
30–50 percent suffered lifelong morbidity, mostly due to central nervous system complications of meningitis.13, 14 Recently, scientists have pointed out that seemingly “cured” meningi47
48
Redirecting Innovation in U.S. Health Care: Case Studies
tis patients, including those who are successfully treated for Hib meningitis, suffer slight but
important lifelong disabilities.15 The Hib vaccine changed this grim picture.
Early Adoption
Foundational discoveries stretching back to the 19th century enabled the development of prototype Hib vaccines in the 1970s. In 1892, Pfeiffer described Haemophilus influenzae as a
Gram-negative bacterium.16 Forty years later, Pittman classified strains of Haemophilus influenzae into two groups: those with capsules and those without (non-typable).17 There are six
capsular types, distinguished by their polysaccharide structures.
Beyond establishing a common language, this classification system has clinical relevance;
pathogenicity and protection are type-specific. Those with invasive Haemophilus influenzae
infection were infected with type b at least 95 percent of the time.18 Type b is so much worse
than the others because, unlike the other capsular types, the type b polysaccharide “shields”
the bacterium from the lytic activity of serum complement alone (innate immunity).19, 20 Consequently, type-specific antibodies (acquired immunity) must be present to initiate complement killing of Hib bacteria.
Infants Are Most Vulnerable to Hib Disease
As the molecular basis for Hib disease became clearer, scientists solved the mystery of why
the organism preferentially attacks infants and young children. Antibodies capable of killing
Hib are not present in this age group.19 Newborns have maternally derived protective antibodies, but their levels decline during the first few months of life. Once these antibodies clear
a newborn’s system, the child is susceptible to infection. This explained why, in the United
States, unvaccinated 6- to 24-month-old infants sustain the highest rates of Hib disease.18, 21
By 6 years of age, most unvaccinated children develop Hib antibodies, either by surviving Hib
infection or, mostly, through repeated contact with nonpathogenic microorganisms that carry
cross-reactive polysaccharides similar to the Hib capsular polysaccharide.22
To protect vulnerable infants, researchers needed to create a vaccine capable of inducing
protective levels of antibody to the Hib capsular polysaccharide at a very young age. Because of
the immaturity of an infant’s immune system, this proved to be a daunting task.
The high incidence and severity of Hib disease led to vaccine development in the 1960s.
The work was initiated by two groups of researchers, one at the Albert Einstein College of
Medicine and then at the Eunice Kennedy Shriver National Institute of Child Health and
Human Development (NICHD) and the U.S. Food and Drug Administration (FDA), and
one at Children’s Hospital Medical Center, Boston, Massachusetts. Both were supported by
the National Institute of Allergy and Infectious Diseases (NIAID) at the National Institutes of
Health (NIH). The Centers for Disease Control and Prevention (CDC) provided surveillance,
and the FDA reviewed the data, approved clinical studies of the candidate vaccines, and finally
licensed them to be marketed to the public.
Two intramural NIH researchers, John B. Robbins and Rachel Schneerson—both
employees of NICHD—dentified the Hib capsular polysaccharide as an essential virulence
factor and a protective antigen as shown previously by Heidelberger for pneumococci;23 their
work aimed to provide a vaccine based on this polysaccharide. In 1969, Robbins and Schneerson initiated a series of studies that ultimately led to the development of the first Hib capsu-
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
49
lar polysaccharide vaccine.19 Their NIAID-supported colleagues, Porter Anderson and David
Smith, independently developed a Hib polysaccharide vaccine at the Children’s Hospital of
Boston during the same time.24 The two groups then conducted clinical studies of the Hib
polysaccharide vaccine.
In order to study the Hib vaccine in humans, scientists needed to first quantify the
normal amount of Hib antibody in humans. Robbins and Schneerson measured the amount
of Hib polysaccharide antibody in a volunteer who received the vaccine. This allowed other
scientists throughout the world to compare Hib antibody levels to the reference level. Robbins
and Schneerson went on to study Hib antibody levels in patients with an immune deficiency
disorder called X-linked hypo-gammaglobulinemia. Preparations of the full complement of
human antibodies were routinely injected to protect these patients from infections like Hib.
Because they knew the amount of Hib antibody in each injection, Robbins and Schneerson
were able to determine the amount of Hib antibody that afforded protected from Hib infection. A Finnish study confirmed their estimate, and the scientists agreed to use this protective
threshold for assessing vaccines.25, 26
Now scientists could match protective levels of Hib antibody to efficacy in preventing
disease in large numbers of individuals. Two controlled, randomized efficacy studies were conducted in 1974: in Mecklenburg County, North Carolina, involving 16,000 children,27 and
in Finland, where nearly 100,000 children were immunized.26 Both studies confirmed safety,
few side effects, and sufficiently protective levels of antibody in children two to five years old.
However, both studies also showed low antibody levels and insufficient protection against
Hib in children less than 18 months old. The Finnish study demonstrated vaccine efficacy of
90 percent at one year in children older than 18 months (95-percent confidence interval, 56 to
96 percent).26, 28
While this work was being done, Anderson and Smith advanced science on the Hib
polysaccharide vaccine. They translated their scientific work into a commercial pharmaceutical company, Praxis Biologics, in 1983.29 As a result of the efforts of this team, Robbins and
Schneerson’s work, and the contributions of others, three nearly identical polysaccharide vaccines against Hib were licensed in 1985: b-Capsa I® by Praxis Biologics, HibVAX® by Connaught Laboratories, and Hib-Immune® by Lederle Laboratories.30
Initial Use
These First-Generation Vaccines Represented a Major Advance But Had Important
Limitations
The new Hib vaccines induced protective levels of Hib antibodies in older children and adults
and prevented systemic Hib disease without significant adverse reactions. However, they failed
to stimulate protective Hib antibody levels in children younger than 18 months old. This was
a serious shortcoming, because this is the age group with the highest risk of contracting lifethreatening Hib disease. Around this time, researchers determined that crowded conditions
are an important risk factor for transmitting Hib disease, because the infection spreads by
droplets from one child to another.31–33 This was a particular concern at day care centers.34–36
Despite these shortcomings, the first-generation Hib vaccines represented a major breakthrough in science. They prevented life-threatening disease in the 25 percent of cases that
occur after the first two years of life.37 Although the basis for licensing Hib vaccine for children
50
Redirecting Innovation in U.S. Health Care: Case Studies
24–59 months of age was largely supported by the 1974 Finnish and Mecklenburg County
studies, which showed protective effects at two years of age,1, 26, 27 the CDC’s Advisory Committee on Immunization Practices (ACIP) recommended administering the vaccine to children as young as 18 months if the child attended day care or was otherwise high-risk.37
Expansion of Use
Initial Studies Produce Mixed Results
Following the licensure and adoption of first-generation Hib polysaccharide vaccines, placebo-controlled trials were no longer ethical. To evaluate the efficacy of these vaccines among
U.S. children younger than five years of age, five case-control studies were performed.37 Two
were supported by the CDC: One sampled day care attendees (vaccine efficacy 41 percent
with a statistically nonsignificant 95-percent confidence interval of –6 to 68 percent); the
other sampled young children without day care attendance (vaccine efficacy 62 percent with
95-percent confidence interval of 7 to 84 percent).38 The third study evaluated members of
the Northern California Kaiser-Permanente Health Plan (vaccine efficacy 69 percent with statistically nonsignificant 95-percent confidence interval of –13 to 91 percent).39 The final two
studies were originally a single study but were later divided into two groups: the Connecticut,
Pittsburgh, and Dallas sites (vaccine efficacy 88 percent with 95-percent confidence interval
of 74 to 96 percent)40 and the Minnesota site (vaccine efficacy –69 percent with statistically
nonsignificant 95-percent confidence interval –285 to 26 percent).41
Based on these unpublished mixed results, concerns arose that the first-generation Hib
vaccines were not uniformly effective in different populations. Therefore, in 1987, the American Academy of Pediatrics (AAP) Committee on Infectious Diseases advised no further use
of the Hib polysaccharide vaccine in areas that had failed efficacy studies, such as Minnesota.
Despite Its Limitations, Use of the First-Generation Vaccine Was Recommended
There were population differences in response, but the FDA and the Advisory Committee on
Immunization Practices recommended continued use of Hib vaccine throughout the United
States.37 Later case-control studies revised the record to five studies with positive estimates of
Hib polysaccharide vaccine efficacy (41 to 88 percent) and three with insignificant or negative
findings.42 These data are presented in Table 4.1.
More than 10 million doses of first-generation Hib polysaccharide vaccine were administered in the United States between 1985 and 1989.13 At that point, the first-generation vaccines were swiftly supplanted by more effective second-generation Hib vaccines. Ultimately,
the first-generation vaccines had a relatively modest effect on the incidence of Hib infection
because they were effective in infants.19 But they set the stage for the remarkable developments
that followed.
Breakthrough: The World’s First Hib Conjugate Vaccine Is Licensed in 1987
The Hib polysaccharide vaccine was better than nothing, but it did not afford protection for
the age groups with the highest incidence of infection. Therefore, the researchers in Bethesda
and in Boston continued to work, hoping to develop a more effective vaccine. The breakthrough came when Robbins and Schneerson extended earlier work done by Karl Landsteiner,
Oswald Avery, and Walther Goebel. Landsteiner had shown that binding a small molecule that
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
51
Table 4.1
Selected Efficacy Studies of the Hib Vaccine
Hib Vaccine
PRPa
Study
Period
Site
Study Type
1974–1976
Finland
Prospective
controlled
trial
1985–1987
Connecticut; Case control
Pittsburgh,
Pennsylvania;
Dallas, Texas
PRP-CRMc
(HbOC)
49,295/
98,272
Vaccine
Schedule
95%
Efficacy Confidence
(%)
Interval (%) Reference
3 mo.–5 yr.,
90
booster 3 mo. in 18–71
later if 3–17
mo.
mo. at first
dose
56–96
26
152/
228
–
88
74–96
40
1985–1987
Northern
California
Case control
166/
201
–
62
–44 to 90
69
1985–1987
Minnesota
Case control
176/
264
–
–55
–283 to 29
70
1986
Mo., N.J., Case control
129/
Okla., Tenn.,
203,
Wash., Los
no day care
Angeles
–
70
17–89
71
1986
Mo., N.J., Case control
291/
Okla., Tenn.,
417,
Wash., Los
in day care
Angeles
–
45
–1 to 70
72
–
–
–6
–184 to 60
73
212/
291
–
–58
–309 to 39
74
–
–
82
38–94
75
1988–1989
Minnesota
Case control
1988–1989
Los Angeles Case control
County
1988–1990 Connecticut; Case control
Pittsburgh,
Pennsylvania
PRP-Db
N
(Control/
Total)
1984–1988
Alaska
Prospective
controlled
trial
1,048/
2,102
2, 4, 6 mo.
35
–57 to 73
49
1985–1989
Finland
Prospective
controlled
trial
56,000/
114,000
3, 4, 6,
14–18 mo.
(intervention)
vs. 24 mo.
(controls)
94
83–98
51, 52
1988–1991
Finland
Prospective 60,500 PRP-D 4, 6, 14–18
controlled 56,500 PRPmo.
trial
CRM/
117,000
100
88–100
53
1988–1990
Northern
California
Prospective
controlled
trial
2, 4, 6 mo.
100
68–100
54
1988–1991
Finland
Prospective 60,500 PRP-D 4, 6, 14–18
controlled 56,500 PRPmo.
trial
CRM/
117,000
100
87–100
53
94
68–99
55
1991–1992
Los Angeles Case control
County
30,680/
61,080
20,800 3
doses
767/
872
2 ,4, 6 mo.
52
Redirecting Innovation in U.S. Health Care: Case Studies
Table 4.1—continued
Hib Vaccine
Study
Period
PRP-OMPd
PRP-Te
a
N
(Control/
Total)
Vaccine
Schedule
95%
Efficacy Confidence
(%)
Interval (%) Reference
Site
Study Type
1988–1990
Navajo
Reservation
Prospective
controlled
trial
2,602/
5,190
2, 4 mo.
93
53–98
56
1991–1993
United
Kingdom
Prospective
controlled
trial
28,000/
56,000
2, 3, 4 mo.
95
74–100
57
1992–1995
Chile
Prospective
controlled
trial
36,741/
72,005
2, 4, 6 mo.
92
65–100
58
1993–1996
Gambia
Prospective
controlled
trial
21,358/
42,848
2, 3, 4 mo.
95
67–100
59
Polyribosylribitol phosphate, the polysaccharide-only Hib vaccine
b Diphtheria toxoid Hib conjugate vaccine
c Mutant diphtheria toxin Hib conjugate vaccine
d Meningococcal outer membrane protein Hib conjugate vaccine
e Tetanus toxoid Hib conjugate vaccine
Notes: mo. = months, yr. = years.
does not provoke an immune response—called a hapten—to a protein could elicit a strong
immune response against the hapten. Avery and Goebel had used the technique to elicit an
immune response to a pneumococcus polysaccharide hapten by binding it to a carrier protein.
Robbins and Schneerson applied the technique to produce a robust immune response to the
Hib polysaccharide by binding it to a different carrier protein. Smith and Anderson in Boston
did the same thing using a different method of conjugation.43, 44
The result of this research was the development of a new, second-generation “conjugate”
Hib vaccine. It was better than the earlier, polysaccharide vaccine in two important ways: First,
the new protein-linked vaccine was more highly immunogenic, so it produced protective levels
of antibodies.6 Attaching a protein to the polysaccharide converted the immune response from
a T-cell–independent process into a T-cell–dependent process, which enables immunologic
memory. As a result, when immunized children were subsequently exposed to Hib or received
a second dose of Hib vaccine, they responded with a strong booster effect.31, 37 Second, studies
showed that the new conjugate Hib vaccine reduced nasopharyngeal carriage of the bacteria.
This conferred a degree of “herd immunity” to unvaccinated children because they were less
likely to catch Hib disease from an infected child.13, 31 One study documented a Hib carriage
rate of zero in vaccinated three-year-olds, compared with a 3.5-percent carriage rate in unvaccinated children of similar age.31 The superiority of the new conjugate Hib vaccines was so
obvious that manufacturers withdrew their remaining stock of first-generation vaccines from
the market in 1989.45 Thereafter, only conjugate Hib vaccines were sold.
Ultimately, four types of Hib conjugate vaccines were brought to market. All contain
the Hib capsular polysaccharide, polyribosylribitol phosphate (PRP), but each links the PRP
molecule to a different protein. The first Hib conjugate vaccine to reach the market was the
diphtheria toxoid conjugate (PRP-D) created by Robbins and Schneerson. It was later modi-
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
53
fied and produced by Connaught Laboratories as ProHIBiT.®30 In 1987, the FDA licensed
PRP-D for children 18 to 60 months old.6 Next came the mutant diphtheria toxin conjugate
(PRP-CRM), which was approved for children 18 months and older in 1988 and subsequently
approved for use in infants in 1990.6, 42 This vaccine was developed by Anderson with production by Praxis Laboratories and distributed by Lederle Laboratories as HbOC (HibTITER®).30
The third vaccine, a meningococcal outer membrane protein conjugate (PRP-OMP; PedvaxHIB®), uses a protein component developed by Merck Sharp and Dohme.30 It was licensed in
1989 for 15- to 18-month-olds and approved for use in infants in 1990.6, 42 With three vaccines
on the market, in April 1990, ACIP recommended immunization of children 15 months and
older with one of the first three conjugate vaccines. Later that same year, both ACIP and the
American Academy of Pediatrics recommended infant immunization with either PRP-CRM
(HbOC) or PRP-OMP. By 1991, ACIP recommended initiation of the primary series of Hib
vaccine at 2 months and a booster dose at 12 to 15 months.46
The fourth variety of Hib conjugate vaccine to reach the market, a tetanus toxoid conjugate (PRP-T), was developed by Schneerson and Robbins and licensed in 1993. It is manufactured by Sanofi Pasteur as ActHIB.®30 At various points, it has been marketed under the
brand names ActHIB® and OmniHIB.® Another version of PRP-T, HIBERIX® produced by
GlaxoSmithKline, was licensed in 2009 for infants and older children.13, 47
Currently Available Hib Vaccines
Today, PRP-T and PRP-OMP are the only Hib conjugate vaccines currently available in the
United States.48 PRP-D (ProHIBIT®) was discontinued in 2000, and PRP-CRM (HbOC;
HibTITER®) was phased out in 2007. HibTITER® combined with diphtheria and tetanus toxoids and a whole-cell pertussis vaccine (DTP) was marketed for a time as TETRAMUNE® by
Lederle-Praxis Biologicals, but it is no longer available. Aventis Pasteur marketed a combination vaccine called TriHIBit (ActHIB® combined with a diphtheria and tetanus toxoids and
acellular pertussis vaccine [Tripedia®]) from 1996 to 2011.45 It was used only as the fourth dose
in a Hib vaccination series.
The vaccines on the market today are produced in both single-agent and combination forms by Sanofi Pasteur (PRP-T; ActHIB® and Pentacel®), GlaxoSmithKline (PRP-T;
HIBERIX® and Menhibrix®), and Merck (PRP-OMP; PedvaxHIB® and Comvax®).48 One
PRP-T vaccine, OmniHIB®, by SmithKline Beecham (now GlaxoSmithKline) is no longer
available.45 Details regarding current Hib conjugate vaccines are summarized in Table 4.2.
Cost and Health Impact
Hib Vaccines Have Proven to Be Highly Effective in Developed World
All but one of the four versions of Hib conjugate vaccine work well in infants. PRP-D was the
exception: It did not perform as well in certain high-risk groups. In particular, PRP-D did not
demonstrate optimal efficacy in a sample of Alaskan Eskimo children (N=2102) with a high
burden of Hib disease.49 This observation was important because the incidence of Hib disease
in 0- to 12-month-old infants was particularly common among Native Americans. In fact,
three-fourths of all cases of Hib disease in Native Americans occurred during infancy, compared with roughly half of black and white children.18 Eventually it was determined that of the
54
Redirecting Innovation in U.S. Health Care: Case Studies
Table 4.2
Currently Licensed Hib Vaccines in the United States
Hib Vaccine
Manufacturer
Brand Name
Date Licensed
Indicated Age
Group
PRP-T
Sanofi Pasteur
ActHIB®
March 1993
2–18 mo.
Pentacel® b
GlaxoSmithKline
PRP-OMP
Merck
June 2008
6 weeks–4 yr.
Regimen
Dose
Age (mo.)
1
2a
2
4
3
6
4
15–18
1
2
2
4
3
6
4
15–18
HIBERIX®
August 2009
15 mo.–4 yr.
4
15–18 c
MENHIBRIX® b
June 2012
6 weeks–18 mo.
1
2
2
4
3
6
4
12–15
1
2
2
4
3
12–15d
1
2
2
4
3
12–15d
PedvaxHIB®
COMVAX® b
December 1989
October 1996
2–71 mo.
6 weeks–15 mo.
a Only two doses eight weeks apart with booster at 15–18 months if first dose at 7–11 months of age; only one
dose with booster two months later if first dose at 12–14 months of age.
b Combination vaccine containing Hib conjugate vaccine with other vaccines
c in the United States, HIBERIX is only licensed as the fourth or booster dose, typically administered at 15–18
months of age.
d No third (booster) dose if the first and second doses (primary series) were received prior to 11 months of age;
only one dose after 15 months of age.
Notes: mo. = months, yr. = years
four conjugate vaccines, PRP-D produced the smallest antibody response among infants—the
group with the highest disease prevalence.50
In a much larger trial conducted in Finland (N=114,000), a nation where Hib disease is
less frequent, PRP-D had an efficacy of 94 percent (95-percent confidence interval of 83 to
98 percent) in children older than 12 months.51, 52 A follow-up Finnish trial comparing PRP-D
to PRP-CRM found efficacies of 100 percent for both (95-percent confidence intervals of 88 to
100 percent and 87 to 100 percent, respectively).53 The alternate PRP-CRM conjugate vaccine
was also found to have efficacy rates of 100 percent (95-percent confidence interval of 68 to
100 percent) and 93 percent (95-percent confidence interval of 70 to 99 percent) in studies
conducted in Northern54 and Southern55 California, respectively.
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
55
Much like Alaskan Eskimo children did in the late 1980s, Navajo children experienced
high rates of Hib disease. The PRP-OMP conjugate vaccine was tested in a sample of Navajo
children under the age of 18 months of age and was determined to have an efficacy of 93 percent (95-percent confidence interval of 53 to 98 percent).56 In this study, rather than administer the usual three doses of the primary series, injections were given at two and four months
only to improve compliance. Importantly, substantial protection against Hib disease was documented after the first dose.
Another innovative regimen was used in a United Kingdom trial of PRP-T.57 This was the
first randomized double-blind efficacy study of PRP-T. In this study, three doses were administered on an accelerated schedule at two, three, and four months of age without a subsequent
booster injection in the second year of life—the common practice at the time. The outcome
of interest was development of invasive Hib disease. Based on “intention to treat” analysis,
vaccine efficacy was 95 percent (95-percent confidence interval of 74 to 100 percent) after
three doses (85 percent [95-percent confidence interval of 49 to 97 percent]). This means that
researchers included all children who were randomly assigned to this treatment group, whether
or not they got all three doses of the PRP-T vaccine.
Hib Conjugate Vaccines Are Also Effective in Developing Countries, But Their High Cost
and Complex Administration Schedule Limits Their Reach
Research done in developing countries has shown that Hib conjugate vaccines are highly effective in these settings as well.19 The problem in low-income countries is not vaccine activity but
cost. Compared with some other childhood vaccines, Hib is relatively expensive and requires a
more complex, multi-dose vaccination schedule to confer a high level of protection. As a result,
many low-income countries have concluded that they cannot afford it. This is unfortunate,
because the vaccine could make a big impact if it was widely administered in these countries.
To determine the impact of Hib vaccination, the Chilean government conducted its own
efficacy studies in Santiago.58 Subsequent research with a Hib conjugate vaccine demonstrated
efficacy of 91.7 percent and effectiveness of 90.2 percent to prevent invasive Hib disease,
91.3 percent to prevent meningitis, and 80 percent to prevent bacteremic pneumonia.58 Similar
findings were reported in Gambia,59 which was able to use these data to secure outside support
for integrating Hib conjugate vaccine into its national immunization program.13 Later studies in Chile demonstrated that the conjugate Hib vaccine also prevented many cases of Hib
pneumonia. This is important, because Hib pneumonia is a bigger problem in the developing
world.60
In High-Income Countries, the Benefits of the Vaccine Outweigh its Costs
Rapid uptake of Hib conjugate vaccines has profoundly changed public health in the developed world.19 When the second-generation conjugate Hib vaccines were brought to market, the
incidence of Hib disease plummeted faster than is typically seen when a new vaccine is introduced.13 By 2002, fewer than 100 cases of Hib disease were reported in the United States. This
represents a greater than 99-percent reduction in disease burden.6 As far back as 1992, reduction
of Hib-related morbidity and mortality was estimated to save the United States $500 million
per year.61 The savings today are likely to be substantially greater, owing to the rise in health
care costs and the sharp reduction in long-term care costs for disabled individuals.
As disease rates fell and the benefits of preventing cognitive and neurological sequelae
grew, it became clear that the costs of Hib immunization were greatly outweighed by the vac-
56
Redirecting Innovation in U.S. Health Care: Case Studies
cine’s health and economic benefits. Similar impact has been noted in other Western countries
with high rates of Hib vaccination, such as Canada, the United Kingdom, Ireland, Austria,
Spain, Switzerland, France, Germany, Finland, Netherlands, Scandinavia, Israel, and Iceland.13, 19 Only three years after Iceland initiated its Hib vaccination program in 1989, no
further cases of Hib disease were documented in that country.50 Hib disease, for all practical
purposes, had been eradicated.
In the Developing World, Cost Is a Barrier to Adoption
Impact to date in the developing world has been much more modest—not because the vaccine
does not work, but because few nations believe they can afford it. According to estimates based
on several studies, Hib immunization efforts in poor nations have reduced the incidence of
early childhood meningitis by only 6 to 8 percent.13 Inaction has carried a high price. In these
countries, Hib disease kills an estimated 600,000 infants every year.
The problem is the vaccine’s cost. Its price—$2 per dose62—plus the need to administer
multiple injections on a set schedule make Hib conjugate vaccine too expensive for many poor
countries to adopt.13 This is why, despite its exceptional effectiveness, four out of five countries
tracked by the World Health Organization have not added Hib vaccine to their routine immunization programs.63
A Comprehensive Vaccination Campaign Could Potentially Eradicate Hib Worldwide
Worldwide eradication of Hib might be possible if effective vaccination programs were established throughout the world.6 The biology of Hib makes it an apt target for eradication. Humans
are the only host for Hib, so widespread use of the conjugate vaccine has the potential to end
Hib disease across the globe. Adults do not need to receive repeat doses of Hib vaccine, as is
typical of other protein vaccines, because their immune systems are stimulated on an ongoing
basis by benign cross-reactive organisms. Thus, adults maintain protective levels of Hib antibodies throughout life. H. influenzae type b has not only shown no propensity to mutate into
a vaccine-resistant form; it is highly unlikely to elicit resistance because the vaccine targets the
polysaccharide that makes it pathogenic. Thus, eradication of this deadly and disabling disease
is technically possible.
The principal obstacle is the vaccine’s cost and cumbersome administration schedule. In
the United States, administration of Hib vaccine costs public payers an average of $5.23 and
private health plans $18.65.64 A more efficient vaccine delivery schedule using smaller quantities of vaccine could make a difference for developing countries, but adoption would still be
hindered by unreliable distribution, supply maintenance, and the complexities of administering multiple doses of vaccine.
Vaccine research continues in the hopes that production techniques will evolve in ways
that maintain the extraordinary clinical value of the Hib vaccine while further reducing its
costs. For example, instead of purifying Hib capsular polysaccharide from Hib bacteria, a biosynthetic version may be possible. This could enable production of anti-Hib conjugate vaccines
on an industrial scale.65 A synthetic conjugate vaccine has already been found to be as safe and
immunogenic as currently licensed Hib conjugate vaccines.65 Further research may reduce its
cost.
Conjugate technology has enabled production of vaccines that have activity to other capsulated bacteria as well, including pneumococci, meningococci, group B streptococci, Salmonella typhi, and E. coli.19 In time, it may be possible to control or perhaps even eradicate
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
57
these pathogens as well. At this point in time, a host of candidate conjugate vaccines are being
explored or are under development.66
A Technology “Home Run”
The CDC designated “vaccines universally recommended for children” as one of the United
States’ ten great public health achievements of the 20th century.67 Hib is one of the leading
agents on the CDC’s list.68 In a subsequent issue of the Morbidity and Mortality Weekly Report
(MMWR Weekly) devoted to childhood immunization, the CDC concluded that, “In less than
a decade, the use of the Hib conjugate vaccines nearly eliminated Hib invasive disease among
[U.S.] children.”67
The Best of the Best
NICHD considers development of the Hib vaccine one of its greatest contributions to public
health. The case of Hib conjugate vaccine development exemplifies the value of partnerships
between public research agencies (the intramural and extramural research programs of the
NIH’s NICHD and NIAID), public health (the CDC, which conducted important epidemiological and efficacy studies), the FDA (which reviewed and approved the marketing of both
the first- and second-generation vaccines), and private industry (which provided the technical
sophistication to develop additional protein conjugates, the resources to test them, and the production and marketing capacity to sell Hib conjugate vaccines throughout the United States,
other high-income countries, and, perhaps in time, the developing world). Since Hib vaccine was brought to market, it has saved the lives of thousands of American children; sharply
reduced the incidence of catastrophic Hib disease; and, by doing so, virtually eliminated the
leading cause of acquired mental disability in the United States. This has spared countless families the heartache and costs of raising children with severe physical and cognitive disabilities.
References
1 Cochi
SL, Broome CV, and Hightower AW, “Immunization of US Children with Hemophilus influenzae
Type b Polysaccharide Vaccine—A Cost-Effectiveness Model of Strategy Assessment,” CORD Conference Proceedings, Vol. 253, No. 4, 1985, pp. 521–529.
2 Eunice
Kennedy Shriver National Institute of Child Health and Human Development, “NICHD Mission
and Scientific Accomplishments: Hemophilus influenzae Type b (Hib) Vaccine,” last updated November 30,
2012. As of May 2013:
http://www.nichd.nih.gov/about/overview/mission/Pages/hib-vaccine.aspx
3 Gilsdorf JRJ, “Bacterial Meningitis in Southwestern Alaska,” American Journal of Epidemiology, Vol. 106,
No. 5, 1977, pp. 388–391.
4 Fraser
DW, Darby CP, Koehler RE, Jacobs CF, and Feldman RA, “The Epidemiology of Haemophilus influenzae Meningitis,” in Sell SH and Karzon DT, eds., Haemophilus influenzae, Nashville, Tenn.: Vanderbilt
University Press, 1973, pp. 231–239.
5 Ward
JI, Margolis HS, Lum MK, Fraser DW, Bender TR, and Anderson P, “Haemophilus influenzae Disease in Alaskan Eskimos: Characteristics of a Population with an Unusual Incidence of Invasive Disease,”
Lancet, June 1981, Vol. 1, No. 8233, pp. 1281–1285.
58
Redirecting Innovation in U.S. Health Care: Case Studies
6 Eunice
Kennedy Shriver National Institute of Child Health and Human Development, home page, undated.
As of January 7, 2014:
http://www.nichd.nih.gov
7 Sell
SH, “Long-Term Sequelae of Bacterial Meningitis in Children,” Pediatric Infectious Disease Journal,
Vol. 2, No. 2, March–April 1983, pp. 90–93.
8 Greenwood
BM, “Epidemiology of Acute Lower Respiratory Tract Infections, Especially Those Due to
Haemophilus influenzae Type b, in the Gambia, West Africa,” Journal of Infectious Diseases, Vol. 165, Supplement 1, June 1992, pp. S26–S28.
9 Bjilmer
HA, “Epidemiology of Haemophilus Influenzae Invasive Disease in Developing Countries and Intervention Strategies,” in Ellis RW and Granoff DM, eds., Development and Uses of Haemophilus B Conjugate Vaccines, New York: Marcel Dekker, 1994, pp. 247–264.
10 Bryan JP, de Silva HR, Tavares A, Rocha H, and Scheld WM, “Etiology and Mortality of Bacterial Meningitis in Northeastern Brazil,” Clinical Infectious Diseases, Vol. 12, No. 1, 1990, pp. 128–135.
11 Gratten
M and Montgomery J, “The Bacteriology of Acute Pneumonia and Meningitis in Children in Papua
New Guinea: Assumptions, Facts and Technical Strategies,” Papua New Guinea Medical Journal, Vol. 34,
No. 3, September 1991, pp. 185–198.
12 Zadik
PMP, “A Bacteriological Study of Meningitis in Kumasi, Ghana,” Journal of Infection, Vol. 13, No. 3,
November 1986, pp. 305–306.
13 Peltola
H, “Worldwide Haemophilus influenzae Type b Disease at the Beginning of the 21st Century:
Global Analysis of the Disease Burden 25 Years After the Use of the Polysaccharide Vaccine and a Decade
After the Advent of Conjugates,” Clinical Microbiology Reviews, Vol. 13, No. 2, April 2000, pp. 302–317.
14 Bijlmer
HA and van Alphen L, “A Prospective, Population-Based Study of Haemophilus influenzae Type b
Meningitis in the Gambia and the Possible Consequences,” Journal of Infectious Diseases, Vol. 165, Supplement 1, June 1992, pp. S29–S32.
15 Roed C, Omland LH, Skinhoj P, Rothman KJ, Sorensen HT, and Obel N, “Educational Achievement and
Economic Self-Sufficiency in Adults After Childhood Bacterial Meningitis, Journal of the American Medical
Association, Vol. 309, No. 16, April 24, 2013, pp. 1714–1721.
16 Pfeiffer
R, “Die Aetiologie der Influenza,” Z Hyg Infektionskr, Vol. 1893, No. 13, pp. 357–386.
17 Pittman
MM, “Variation and Type Specificity in the Bacterial Species Hemophilus influenzae,” Journal of
Experimental Medicine, Vol. 53, No. 4, March 31, 1931, pp. 471–492.
18 Broome
CV, “Epidemiology of Haemophilus influenzae Type b Infections in the United States,” Pediatric
Infectious Disease Journal, Vol. 6, No. 8, August 1987, pp. 779–782.
19 Robbins
JB, Schneerson R, Anderson P, and Smith DH, “Prevention of Systemic Infections, Especially
Meningitis, Caused by Haemophilus influenzae Type b,” Journal of the American Medical Association, Vol. 276,
No. 14, October 9, 1996, pp. 1181–1185.
20 Sutton A, Schneerson R, Kendall-Morris S, and Robbins JB, “Differential Complement Resistance Mediates Virulence of Haemophilus influenzae Type b,” Infection and Immunity, Vol. 35, No. 1, January 1982,
pp. 95–104.
21 Fothergill
LD and Wright J, “Influenzal Meningitis,” Journal of Immunology, Vol. 24, 1933, pp. 273–284.
22 Argaman M and Handzel ZT, “Haemophilus influenzae Type b: Disease and Immunity in Humans,”
Annals of Internal Medicine, Vol. 58, February 1973, pp. 259–269.
23 Heidelberger
M, MacLeod CM, Hodges RG, Bernhard WG, and di Lapi MM, “Antibody Formation in
Men Following Injection of Four Type-Specific Polysaccharides of Pneumococcus,” Journal of Experimental
Medicine, Vol. 85, No. 3, February 28, 1947, pp. 227–230.
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
59
24 Peter
G, “Responses of Children Immunized with Capsular Polysaccharide of Hemophilus influenzae Type b, by
David H. Smith, MD, et al, Pediatrics, 1973;52:637–644; and Haemophilus influenzae Type b Capsular Polysaccharide Vaccine in Children: A Double-Blind Field Study of 100 000 Vaccinees 3 Months to 5 Years of Age in Finland, by Heikki Peltola, MD et al, Pediatrics, 1977;60:730–737,” Pediatrics, Vol. 102, No. 1, Part 2, July 1998,
pp. 252–254.
25 Robbins
JB, Parke JC, Schneerson R, and Whisnant JK, “Quantitative Measurement of ‘Natural’ and
Immunization-Induced Haemophilus influenzae Type b Capsular Polysaccharide Antibodies,” Pediatric
Research, Vol. 7, No. 3, March, pp. 103–110.
26 Peltola
H, Käyhty HH, Sivonen A, and Mäkelä PH, “Haemophilus influenzae Type b Capsular Polysaccharide Vaccine in Children: A Double-Blind Field Study of 100,000 Vaccinees 3 Months to 5 Years of Age in
Finland,” Pediatrics, Vol. 60, November 1977, pp. 730–737.
27 Parke JC, Schneerson R, Robbins JB, and Schlesselman JJ, “Interim Report of a Controlled Field Trial of
Immunization with Capsular Polysaccharides of Haemophilus influenzae Type b and Group C Neisseria Meningitidis in Mecklenburg County, North Carolina (March 1974–March 1976),” Journal of Infectious Diseases,
Vol. 136, Supplement, August 1977, pp. S51–S56.
28 Robbins JB, Schneerson R, and Parke JC, “A Review of the Efficacy Trials with Haemophilus influenzae Type b Polysaccharide Vaccines,” in Sell SH and Wright PF, eds., Haemophilus influenzae: Epidemiology,
Immunology, and Prevention of Disease, Elsevier Science Publishing Company, 1982, pp. 255–263.
29 The
Science Coalition, “Praxis Biologics,” 2013. As of July 20, 2013:
http://www.sciencecoalition.org/successstories-list
30 Shapiro
ED and Ward JI, “The Epidemiology and Prevention of Disease Caused by Haemophilus influenzae
Type b,” Epidemiologic Reviews, Vol. 13, 1991, pp. 113–142.
31 Takala
AK, Eskola JJ, Leinonen MM, Käyhty H, Nissinen A, Pekkanen E, and Mäkelä PH, “Reduction of
Oropharyngeal Carriage of Haemophilus influenzae Type b (Hib) in Children Immunized with an Hib Conjugate Vaccine,” Journal of Infectious Diseases, Vol. 164, No. 5, November 1991, pp. 982–986.
32 Cochi
SL, Fleming DW, Hightower AW, Limpakarnjanarat K, Facklam RR, Smith JD, Sikes RK, and
Broome CV, “Primary Invasive Haemophilus influenzae Type b Disease: A Population-Based Assessment of
Risk Factors,” Journal of Pediatrics, Vol. 108, No. 6, June 1986, pp. 887–896.
33 Singleton
RR, Holve SS, Groom AA, McMahon BJ, Santosham M, Brenneman G, and O’Brien KL,
“Impact of Immunizations on the Disease Burden of American Indian and Alaska Native Children,” Archives
of Pediatrics and Adolescent Medicine, Vol. 163, No. 5, May 2009, pp. 446–453.
34 Glode
MP, Schiffer MS, Robbins JB, Khan W, Battle CU, and Armenta E, “An Outbreak of Hemophilus
influenzae Type b Meningitis in an Enclosed Hospital Population,” CORD Conference Proceedings, Vol. 88,
No. 1, January 1976, pp. 36–40.
35 Ginsburg
CM, McCracken GH, Rae S, and Parke JC, “Haemophilus influenzae Type b Disease: Incidence in a Day-Care Center,” Journal of the American Medical Association, Vol. 238, No. 7, August 15, 1977,
pp. 604–607.
36 Ward
JI, Gorman G, Phillips C, and Fraser DW, “Hemophilus influenzae Type b Disease in a Day-Care
Center. Report of an Outbreak,” Journal of Pediatrics, Vol. 92, No. 5, May 1978, pp. 713–717.
37 Ward JI, Broome CV, Harrison LH, Shinefield H, and Black S, “Haemophilus influenzae Type b Vaccines:
Lessons for the Future,” CORD Conference Proceedings, Vol. 81, No. 6, June 1, 1988, pp. 886–893.
38 Harrison
LH, Hightower AW, and Gaventa S, “Case-Control Efficacy Study of the Polysaccharide Haemophilus influenzae, Type b (Hib) Vaccine,” 27th Interscience Conference of Antimicrobial Agents and Chemotherapy, 1987.
39 Black
SB, Shinefield HR, and Hiatt RA, “Efficacy of Haemophilus influenzae Type b Polysaccharide Vaccine in 87,541 Children,” Pediatric Infectious Disease Journal, Vol. 7, 1988, pp. 149–156.
60
Redirecting Innovation in U.S. Health Care: Case Studies
40 Murphy
TV, Shapiro ED, and Wald ER, “The Protective Efficacy (PE) of Haemophilus influenzae Type b
(Hib) Polysaccharide Vaccine (PV),” 27th Interscience Conference of Antimicrobial Agents and Chemotherapy,
1987.
41 Osterholm
MT, Rambeck JH, and White KE, “Lack of Protective Efficacy and Increased Risk of Disease
Within 7 Days After Vaccination Associated with Haemophilus influenzae Type b (Hib) Polysaccharide (PS)
Vaccine Use in Minnesota (MN),” 27th Interscience Conference of Antimicrobial Agents and Chemotherapy,
1987.
42 Adams
WG, Deaver KA, Cochi SL, Plikaytis BD, Zell ER, Broome CV, and Wenger JD, “Decline of
Childhood Haemophilus influenzae Type b (Hib) Disease in the Hib Vaccine Era,” Journal of the American
Medical Association, Vol. 269, No. 2, January 13, 1993, pp. 221–226.
43 Schneerson
R, Barrera O, and Robbins JB, “Preparation, Characterization, and Immunogenicity of Haemophilus influenzae Type b Polysaccharide-Protein Conjugates,” Journal of Experimental Medicine, Vol. 152,
No. 2, August 1, 1980, pp. 361–376.
44 Anderson
PW, Pichichero ME, Insel RA, Betts R, Eby R, and Smith DH, “Vaccines Consisting of Periodate-Cleaved Oligosaccharides from the Capsule of Haemophilus influenzae Type b Coupled to a Protein
Carrier: Structural and Temporal Requirements for Priming in the Human Infant,” Journal of Immunology,
Vol. 137, No. 4, August 15, 1986, pp. 1181–1186.
45 Centers
for Disease Control and Prevention, “Selected Discontinued U.S. Vaccines,” November 2011. As of
July 21, 2013:
http://www.cdc.gov/vaccines/pubs/pinkbook/downloads/appendices/b/discontinued_vaccines.pdf
46 “Haemophilus
b Conjugate Vaccines for Prevention of Haemophilus influenzae Type b Disease Among
Infants and Children Two Months of Age and Older—Recommendations of the Immunization Practices
Advisory Committee (ACIP),” Morbidity and Mortality Weekly Report Recommendations and Reports, Vol. 40,
No. RR-01, January 11, 1991, pp. 1–7.
47 U.S. Food and Drug Administration, “Accelerated Approval of Hiberix to Help Sustain Adequate Vaccine
Supply,” August 19, 2009. As of July 21, 2013:
http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2009/ucm179533.htm
48 National Vaccine Information Center, “Haemophilus influenza Type b Vaccine (HIB),” 2013. As of July 21,
2013:
http://www.nvic.org/Vaccines-and-Diseases/HIB.aspx
49 Ward
JJ, Brenneman GG, Letson GW, Heyward WL, and the Alaska H. influenzae Vaccine Study Group,
“Limited Efficacy of a Haemophilus influenzae Type b Conjugate Vaccine in Alaska Native Infants,” New
England Journal of Medicine, Vol. 323, No. 20, November 15, 1990, pp. 1393–1401.
50 Heath
PT, “Haemophilus influenzae Type b Conjugate Vaccines: A Review of Efficacy Data,” Pediatric
Infectious Disease Journal, Vol. 17, No. 9, Supplement, September 1998, pp. S117–S122.
51 Eskola JJ, Peltola HH, Takala AK, Käyhty H, Hakulinen M, Karanko V, Kela E, Rekola P, Rönnberg P,
Samuelson JS, Gordon LK, and Mäkelä PH, “Efficacy of Haemophilus influenzae Type b PolysaccharideDiphtheria Toxoid Conjugate Vaccine in Infancy,” New England Journal of Medicine, Vol. 317, No. 12, September 17, 1987, pp. 717–722.
52 Eskola
JJ, Käyhty HH, Takala AK, Peltola H, Rönnberg PR, Kela E, Pekkanen E, McVerry PH, and
Mäkelä PH, “A Randomized, Prospective Field Trial of a Conjugate Vaccine in the Protection of Infants and
Young Children Against Invasive Haemophilus influenzae Type b Disease,” New England Journal of Medicine,
Vol. 323, No. 20, November 15, 1990, pp. 1381–1387.
53 Peltola HH, Eskola JJ, Käyhty HH, Takala AK, and Mäkelä PH, “Clinical Comparison of the Haemophilus
influenzae Type b Polysaccharide-Diphtheria Toxoid and the Oligosaccharide-CRM197 Protein Vaccines in
Infancy,” Archives of Pediatrics and Adolescent Medicine, Vol. 148, No. 6, June 1994, pp. 620–625.
54 Black
SB, Shinefield HR, Fireman BB, Hiatt RR, Polen MM, and Vittinghoff EE, “Efficacy in Infancy of
Oligosaccharide Conjugate Haemophilus influenzae Type b (HbOC) Vaccine in a United States Population
Case Study 4: Haemophilus influenzae Type b (Hib) Vaccine
61
of 61,080 Children—The Northern California Kaiser Permanente Vaccine Study Center Pediatrics Group,”
Pediatric Infectious Disease Journal, Vol. 10, No. 2, February 1991, pp. 97–104.
55 Vadheim
CM, Greenberg DP, Eriksen EE, Hemenway L, Christenson P, Ward B, Mascola L, and Ward JI,
“Protection Provided by Haemophilus influenzae Type b Conjugate Vaccines in Los Angeles County: A CaseControl Study,” Pediatric Infectious Disease Journal, Vol. 13, No. 4, April 1994, pp. 274–280.
56 Santosham
MM, Wolff MM, Reid RR, Hohenboken M, Bateman M, Goepp J, Cortese M, Sack D, Hill
J, Newcomer W, Capriotti L, Smith J, Owen M, Gahagan S, Hu D, Kling R, Lukacs L, Ellis RW, Vella PP,
Calandra G, Matthews H, and Ahonkhai V, “The Efficacy in Navajo Infants of a Conjugate Vaccine Consisting of Haemophilus influenzae Type b Polysaccharide and Neisseria Meningitidis Outer-Membrane Protein
Complex,” New England Journal of Medicine, Vol. 324, No. 25, June 20, 1991, pp. 1767–1772.
57 Booy
R, Hodgson S, Carpenter L, Mayon-White RT, Slack MP, Macfarlane JA, Haworth EA, Kiddle M,
Shribman S, and Roberts JS, “Efficacy of Haemophilus influenzae Type b Conjugate Vaccine PRP-T,” Lancet,
Vol. 344, No. 8919, August 6, 1994, pp. 362–366.
58 Lagos
RR, Horwitz II, Toro JJ, San Martin O, Abrego P, Bustamante C, Wasserman SS, Levine OS, and
Levine MM, “Large Scale, Postlicensure, Selective Vaccination of Chilean Infants with PRP-T Conjugate Vaccine: Practicality and Effectiveness in Preventing Invasive Haemophilus influenzae Type b Infections,” Pediatric Infectious Disease Journal, Vol. 15, No. 3, March 1996, pp. 216–222.
59 Mulholland
K, Hilton S, Adegbola R, Usen S, Oparaugo A, Omosigho C, Weber M, Palmer A, Schneider
G, Jobe K, Lahai G, Jaffar S, Secka O, Lin K, Ethevenaux C, and Greenwood B, “Randomised Trial of Haemophilus influenzae Type-b Tetanus Protein Conjugate for Prevention of Pneumonia and Meningitis in Gambian Infants,” Lancet, Vol. 349, April 26, 1997, pp. 1191–1197.
60 Levine
OS, Lagos RR, Muñoz AA, Villaroel J, Alvarez AM, Abrego P, and Levine MM, “Defining the
Burden of Pneumonia in Children Preventable by Vaccination Against Haemophilus influenzae Type b,” Pediatric Infectious Disease Journal, Vol. 18, No. 12, December 1999, pp. 1060–1064.
61 Shinefield H and Black S, “Conjugate Hib Vaccines and Their Combinations: Present Success and Future
Possibilities,” Journal of the American Medical Association, Southeast Asian Edition, Vol. 9, 1993, pp. S20–S23.
62 Children’s
Vaccine Initiative, “Vaccination News: Haemophilus influenzae—Hib Use Rising . . . Slowly,”
Newswatch, Vol. 1, 1999, pp. 7–8.
63 Children’s
pp. 13–15.
Vaccine Initiative, “Hib—Who’s Using It, Who Isn’t and Why Not?” CVI Forum, Vol. 16, 1998,
64 Zhou
FF, Santoli JJ, Messonnier ML, Yusuf HR, Shefer A, Chu SY, Rodewald L, and Harpaz R, “Economic
Evaluation of the 7-Vaccine Routine Childhood Immunization Schedule in the United States, 2001,” Archives
of Pediatrics and Adolescent Medicine, Vol. 159, No. 12, December 2005, pp. 1136–1144.
65 Verez-Bencomo
VV, Fernández-Santana VV, Hardy EE, Toledo ME, Rodríguez MC, Heynngnezz L,
Rodriguez A, Baly A, Herrera L, Izquierdo M, Villar A, Valdés Y, Cosme K, Deler ML, Montane M, Garcia
E, Ramos A, Aguilar A, Medina E, Toraño G, Sosa I, Hernandez I, Martínez R, Muzachio A, Carmenates A,
Costa L, Cardoso F, Campa C, Diaz M, and Roy R, “A Synthetic Conjugate Polysaccharide Vaccine Against
Haemophilus influenzae Type b,” Science, Vol. 305, No. 5683, July 23, 2004, pp. 522–525.
66 Jennings
HJ and Pon RA, “Polysaccharides and Glycoconjugates as Human Vaccines,” in Dumitriu S, ed.,
Polysaccharides in Medicinal Applications, New York: Marcel Dekker; 1996, pp. 443–480.
67 Centers for Disease Control and Prevention, “Achievements in Public Health, 1900–1999: Impact of Vaccines Universally Recommended for Children—United States, 1990–1998,” Morbidity and Mortality Weekly,
Vol. 48, No. 12, April 2, 1999, pp. 243–248. As of January 7, 2014:
http://www.cdc.gov/mmwr/preview/mmwrhtml/00056803.htm
68 Centers for Disease Control and Prevention, “Ten Great Public Health Achievements—United States,
1900–1999,” Morbidity and Mortality Weekly, Vol. 48, No. 12, April 2, 1999, pp. 241–243. As of January 7,
2014:
http://www.cdc.gov/mmwr/preview/mmwrhtml/00056796.htm
62
Redirecting Innovation in U.S. Health Care: Case Studies
69 Vadheim CM, Greenberg DP, and Bordenave N, “Risk Factors for and Effect of Vaccination on Invasive H.
influenzae Type b Disease in Children 18–59 Months of Age,” in Program and Abstracts of the 30th Interscience
Conference on Antimicrobial Agents and Chemotherapy, Atlanta, GA, October 1990, 1990.
70 Osterholm MT, Rambeck JH, White KE, Jacobs JL, Pierson LM, Neaton JD, Hedberg CW, MacDonald
KL, and Granoff DM, “Lack of Efficacy of Haemophilus b Polysaccharide Vaccine in Minnesota,” Journal of
the American Medical Association, Vol. 260, No. 10, September 9, 1988, pp. 1423–1428.
71 Harrison
LH, Broome CV, and Hightower AW, “Haemophilus influenzae Type b Polysaccharide Vaccine:
An Efficacy Study,” Pediatrics, Vol. 84, No. 2, August 1989, pp. 255–261.
72 Harrison LH, Broome CV, Hightower AW, Hoppe CC, Makintubee S, Sitze SL, Taylor JA, Gaventa S,
Wenger JD, and Facklam RR, “A Day Care-Based Study of the Efficacy of Haemophilus influenzae Type b
Polysaccharide Vaccine,” Journal of the American Medical Association, Vol. 260, No. 10, September 9, 1988,
pp. 1413–1418.
73 Osterholm MT, Jacobs JL, and White KE, “Efficacy of Haemophilus influenzae b Plain Polysaccharide
(PRP) Vaccine and Conjugate Vaccine (PRP-D) in Minnesota,” in Program and Abstracts of the 30th Interscience Conference on Antimicrobial Agents and Chemotherapy, Atlanta, GA, October 1990, 1990.
74 Greenberg DP, Vadheim CM, Bordenave NN, Ziontz L, Christenson P, Waterman SH, and Ward JI,
“Protective Efficacy of Haemophilus influenzae Type b Polysaccharide and Conjugate Vaccines in Children
18 Months of Age and Older,” Journal of the American Medical Association, Vol. 265, No. 8, February 27, 1991,
pp. 987–992.
75 Shapiro ED and Wald ER, “Protective Efficacy of PRP-D Conjugate Vaccine Against Haemophilus influenzae Type b,” in Program and Abstracts of the 30th Interscience conference on Antimicrobial Agents and Chemotherapy, Atlanta, GA, October 1990, 1990.
Case Study 5
Implantable Cardioverter-Defibrillator
Erica S. Spatz, M.D., M.H.S.
Fellow of Cardiovascular Medicine, Yale–New Haven Hospital,
Yale University School of Medicine
The Technology
Implantable cardioverter-defibrillators (ICDs) are designed to sense a life-threatening cardiac
(heart) arrhythmia, such as ventricular tachycardia or ventricular fibrillation, and automatically provide a jolt of direct current (DC) electricity to covert the heart rhythm back to a
normal heartbeat (normal sinus rhythm). In contrast to external defibrillators, such as the
devices used in coronary care units, emergency departments, and the back of an ambulance,
the components of an ICD are miniaturized so that they can be implanted inside a patient’s
body.
The ICD is comprised of a pulse generator and one or more leads. The defibrillator lead
is introduced transvenously and implanted inside the right ventricle of the heart. The lead has
one or two coils at its tip and is able to deliver a series of high-energy shocks (5 to 30 Joules
each) to the interior of the heart without damaging the heart muscle; an ICD’s leads also contain bipolar electrodes, which are used for ventricular pacing and sensing. Variations of the
ICD include dual-chamber (atrial and ventricular) and biventricular (right and left ventricular)
devices, which consist of additional leads to sense and pace the heart’s right atrium and left
ventricle, respectively. The lead or leads are connected to a pulse generator, which is typically
implanted in a pocket under the skin in the upper chest below the collarbone but outside the
rib cage (see Figure 5.1).
A typical pulse-generator is approximately 5x5 cm—or about half the size of a
smartphone—and consists of a battery, capacitors, and ports for the leads. Physicians can program the device to monitor abnormal heart rhythms using different thresholds and to deliver
therapy, including antitachycardia pacing, to stop a dangerous tachycardia or deliver countershock therapy to terminate a potentially lethal cardiac arrhythmia. Current ICD models have
the capacity to store electrograms, useful in determining whether patients received appropriate
therapy (i.e., the defibrillator appropriately identified the abnormal heart rhythm as dangerous
and delivered the correct treatment). The generator also collects and reports information on
battery voltage, half-life, and lead impedance. Dual-chamber devices can detect atrial arrhythmias and pace the atrium, while biventricular devices synchronize pacing of the left and right
ventricles (the two main pumping chambers of the heart) to improve cardiac output in patients
with congestive heart failure and to overcome certain cardiac conduction delays—specifically,
a form that impairs the ability of the left and right ventricles to work together. Typically, the
battery of an ICD lasts four to nine years, at which point it must be replaced. The procedure is
63
64
Redirecting Innovation in U.S. Health Care: Case Studies
Figure 5.1
An ICD Implanted
SOURCE: National Heart, Lung, and Blood Institute, “What Is
an Implantable Cardioverter Defibrillator?” revised August 2009.57
performed by cardiologists, primarily electrophysiologists in the United States, under general
anesthesia and is similar to the placement of a standard pacemaker.1
Rationale
ICDs were developed for the secondary prevention of sudden cardiac death (SCD), a condition
caused by two particularly lethal types of cardiac arrhythmias—sustained ventricular tachycardia and ventricular fibrillation. Both cause sudden loss of consciousness and, if untreated,
the latter almost invariably results in death within a few minutes. Most people who sustain an
out-of-hospital cardiac arrest do not survive. In the United States, SCD is the leading cause of
deaths due to heart disease, accounting for approximately 350,000 deaths annually.
Approximately 80 percent of people who suffer SCD have coronary heart disease.2 Men
account for the majority of SCD cases (75%). Most are adults between the ages of 45–75 years.
Major risk factors for SCD include left ventricular heart failure (systolic function ≤30–35%),
hypertrophic or dilated cardiomyopathy, valvular heat disease, certain congenital heart anomalies, and primary electrophysiologic abnormalities.
ICDs have been shown to be effective in preventing SCD in persons with prior SCD who
are found to have a nonreversible, continued risk for recurrent SCD. This is known as secondary
prevention. ICDs have also been studied and found to be effective for the primary prevention
of SCD, yet this group has generated much scientific and political debate (Box 5.1). Although
a growing body of evidence supports the use of ICD therapy in selected patients with decreased
left ventricular systolic function who have no history of SCD, it is not clear that all members
of this much larger group will benefit, or that many would select treatment if they fully understood its potential benefits and risks.
Case Study 5: Implantable Cardioverter-Defibrillator
65
Box 5.1
Indications for ICD Therapy in the Prevention of SCD
Secondary Prevention: Patients in this group have survived a prior cardiac arrest or sustained ventricular tachycardia. Additionally, patients with unexplained syncope thought to
be due to ventricular arrhythmias are considered eligible for secondary prevention of SCD
with an ICD.
Primary Prevention: Patients in this group have no history of SCD or sustained ventricular
tachycardia but have a cardiac condition that places them at risk for SCD. By far, the largest
group of patients who qualify for ICD therapy for primary prevention are patients with congestive heart failure. However, patients with congenital anomalies, other types of structural
heart disease, and conduction defects may also qualify.
Development
The first DC defibrillator was developed in 1962 by Dr. Bernard Lown,3 a renowned physicianscientist who went on to become an equally accomplished humanitarian and social activist.
Years after his groundbreaking work with defibrillators and anti-arrhythmics, Dr. Lown cofounded (along with Dr. Yevgeny Chazov of the former Soviet Union) International Physicians
for the Prevention of Nuclear War. For their work, Drs. Lown and Chazov received the Nobel
Peace Prize in 1985.4
Initially, all defibrillators delivered DC countershocks through external paddles that were
applied to the patient’s chest. Although their use was initially restricted to physicians, the treatment was so time-urgent that training programs were established to teach coronary care nurses
and prehospital emergency medical services personnel (paramedics) to recognize a dangerous
cardiac arrhythmia and attempt defibrillation or DC cardioversion.
In the mid-1980s, advances in microcomputers enabled manufacturers to produce automated external defibrillators (AEDs) that can accurately diagnose a dangerous cardiac rhythm
and advise relatively untrained personnel, such as basic emergency medical technicians, firefighters, police officers, and even lay rescuers to deliver external countershocks through large
adhesive electrodes applied to the victim’s chest. Although AEDs are much simpler to use than
manual defibrillators, they still require an individual to recognize that a cardiac emergency
has occurred, apply the electrodes, activate the device, and follow the automated prompts. Precious time can be lost while these steps are being accomplished. If no device is available at the
time SCD occurs, the best that a bystander can do is perform CPR while emergency medical
services rush to the scene. If the victim’s collapse is not observed, death almost always follows
because treatment is started too late.
ICDs were conceptualized as a way to overcome these logistical shortcomings. Dr. Michel
Mirowski, a Polish-born cardiologist who fled his country during World War II, conceived the
idea for an implantable defibrillator in 1967, after a close friend and mentor died from recurrent bouts of ventricular tachycardia. Dr. Mirowski spent the next decade testing implanted
defibrillator devices in dogs, with variable degrees of success. The first internal defibrillator
was surgically implanted in 1980 at the Johns Hopkins Medical Center by Dr. Levi Watkins
and Vivien Thomas, pioneering African American medical researchers. Thomas had previously
worked as the lab assistant to Dr. Alfred Blalock, the surgeon who had designed the first proce-
66
Redirecting Innovation in U.S. Health Care: Case Studies
dures to correct Tetralogy of Fallot and other congenital defects; however, racial discrimination
and social inequities prevented Thomas from obtaining a medical degree or being credited with
his contributions to Blalock’s success. Watkins was the first African American graduate of the
Vanderbilt University School of Medicine. In 1985, roughly the same time when AEDs were
being rapidly adopted by EMS services, the FDA approved the first implantable defibrillator,
developed for commercial use by Boston Scientific. Until 1993, ICDs were implanted inside
the rib cage. In 1993, the first transvenous defibrillator lead was approved by the FDA.
Early Adoption
Within cardiology circles, the clinical trials that generated supporting evidence for use of ICDs
are as well-known as the device’s founders. (As is true of many drug trials, they are best known
by the study’s acronym.) Collectively, these studies define, for physicians and third-party payers
alike, which patients are most likely to benefit from ICD therapy. Diffusion of ICD therapy for
prevention of sudden cardiac death (SCD) can be benchmarked to the date on which each trial
was published. Because ICDs were originally envisioned for secondary prevention of SCD, the
data supporting this use case are reviewed first, followed by evidence for the far more expansive
use of ICDs as a strategy for primary prevention of SCD.
Secondary Prevention of Sudden Cardiac Death
It has long been known that survivors of SCD due to ventricular fibrillation and patients who
survive one or more bouts of sustained ventricular tachycardia are at very high risk for recurrent SCD. Two-year mortality rates vary from 20 percent to 45 percent, depending on the
cohort. A series of secondary prevention trials conducted in the late 1980s to 1990s demonstrated improved survival among patients randomized to ICD therapy, though only one study
had sufficient power to reach statistical significance. The three studies are Antiarrhythmics
Versus Implantable Defibrillators (AVID),5 the Cardiac Arrest Study Hamberg (CASH),6 and
the Canadian Implantable Defibrillator Study (CIDS) trials.7
AVID (1997)
In this study, supported by the National Institutes of Health’s (NIH’s) National Heart, Lung
and Blood Institute (NHLBI), patients with a history of sudden cardiac death or ventricular
tachycardia requiring cardioversion were randomized to treatment with an ICD or a class III
antiarrhythmic drug, primarily amiodarone. Patients enrolled with a history of ventricular
tachycardia also had to have an ejection fraction of less than 40 percent and to be symptomatic with heart failure. The trial was terminated early when a significant survival benefit was
observed among the 507 patients randomized to the ICD arm as compared with 508 patients
in the medication arm. The survival benefit persisted each year for three years: Relative risk
reductions with ICD therapy compared with medication therapy were 39 percent, 29 percent,
and 31 percent, p<0.02, respectively.
CASH (2000)
In this industry-sponsored trial, patients resuscitated from documented sustained ventricular
arrhythmias (n=288) were eligible for inclusion and were randomized to ICD therapy or one
of three antiarrhythmic medications (propafenone, amiodarone, or metoprolol). The primary
Case Study 5: Implantable Cardioverter-Defibrillator
67
endpoint was all-cause mortality. Enrollment in the propafenone arm was terminated early
after an interim analysis revealed a 61-percent higher mortality as compared with patients
in the ICD arm. The other arms were followed for a mean of two years. There was a nonsignificant, 23-percent relative reduction in mortality in the ICD arm as compared with the
combined amiodarone and metoprolol arms. However, the study was underpowered to find a
significant effect if one existed.
CIDS (2000)
In this study, supported by a combination of government and industry funding, 659 patients
resuscitated from one or more prior events of ventricular fibrillation, ventricular tachycardia,
or unmonitored syncope with evidence of spontaneous (>10 seconds) or readily inducible ventricular tachycardia were randomized to therapy with an ICD or amiodarone. The primary
outcome was all-cause mortality. Similar to findings in the CASH study, there was a nonsignificant, 20-percent relative risk reduction in mortality at five years among patients in the
ICD arm as compared with the amiodarone arm. There was also a nonsignificant, 33-percent
reduction in arrhythmic deaths.
A subsequent meta-analysis using data from these three trials was performed by Connolly et al. (2000).7 It revealed a cumulative 28-percent reduction (hazard ratio [HR] = 0.72;
95% confidence interval [CI] = 0.60–0.87, p = 0.0006) in the risk of all-cause mortality,
almost entirely driven by a 50-percent decrease in the risk of arrhythmic death (HR = 0.50,
95% CI = 0.37–0.67, p<0.0001). Patients with a reduced systolic function, defined as an ejection fraction of less than or equal to 35 percent, benefited more than patients with preserved
systolic function. The results of this meta-analysis and the three trials that informed it provide
the empirical basis for the American Heart Association’s current guidelines for the secondary
prevention of SCD (Box 5.2).
Box 5.2
Class I Recommendations for ICD Therapy as Secondary Prevention in Patients with Heart Failure 8
• ICD therapy is indicated in patients who are survivors of cardiac arrest due to ventricular fibrillation or hemodynamically unstable sustained ventricular tachycardia after
evaluation to define the cause of the event and to exclude any completely reversible
causes (Level of Evidence: A).
• ICD therapy is indicated in patients with structural heart disease and spontaneous sustained VT, whether hemodynamically stable or unstable (Level of Evidence: B).
• ICD therapy is indicated in patients with syncope of undetermined origin with clinically relevant, hemodynamically significant sustained VT or ventricular fibrillation
induced at electrophysiological study (Level of Evidence: B).
SOURCE: Members WC, Epstein AE, DiMarco JP, et al., “ACC/AHA/HRS 2008 Guidelines for Device-Based
Therapy of Cardiac Rhythm Abnormalities: Executive Summary—A Report of the American College of
Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the
ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices):
Developed in Collaboration With the American Association for Thoracic Surgery and Society of Thoracic
Surgeons,” Circulation, Vol. 117, No. 21, May 27, 2008, pp. 2820–2840. 8
68
Redirecting Innovation in U.S. Health Care: Case Studies
Primary Prevention of SCD
Because the pool of survivors of SCD and sustained ventricular tachycardia is relatively small,
researchers examined whether ICDs might benefit patients at risk for this condition prior to
the first such event. Since then, a series of trials have been conducted that demonstrate effectiveness in an expanding group of adults, all of whom have congestive heart failure and left
ventricular systolic dysfunction.
In the late 1990s, two trials investigated the efficacy of ICD therapy in patients with
underlying ischemic heart disease, the Multicenter Automatic Defibrillator Implantation Trial
(MADIT)9 and the Multicenter Unsustained Tachycardia Trial (MUSTT).10
MADIT (New England Journal of Medicine, 1996)
This industry-sponsored (CPI/Guidant) randomized controlled trial enrolled 196 patients
with prior myocardial infarction (MI) (75 percent of whom had an MI less than six months
prior to enrollment) who were deemed to be at high risk for ventricular tachycardia. Eligible
patients had systolic heart failure with an ejection fraction less than or equal to 35 percent;
nonsustained ventricular tachycardia; and inducible, nonsuppressible ventricular tachycardia
on an electrophysiological (EP) study. Patients were randomized to ICD versus standard medical care (i.e., anti-arrhythmic medication), and average follow-up was 27 months. The primary
endpoint was all-cause mortality. The mean age of patients was 63 years. The study detected
a clinically and statistically significant decrease in mortality among patients randomized to
ICD therapy (HR = 0.46, 95% CI = 0.26–0.82, p = 0.009).
MUSTT (New England Journal of Medicine, 1999)
This randomized controlled trial, supported by NHLBI and industry, studied a similar patient
population as those included in MADIT, testing whether the strategy of performing an EP
study to stratify patients at high risk of SCD was effective. The primary endpoint was cardiac
arrest or death from a cardiac arrhythmia. Patients with prior MI; ejection fraction less than
40 percent; and spontaneous, nonsustained ventricular tachycardia were randomized to one
of two arms: therapy guided by an EP study or no antiarrhythmic therapy. Patients in the first
arm, who had inducible ventricular tachycardia on an EP study, received a staged protocol of
anti-arrhythmic drugs, followed by ICD implantation if the medication failed.
At five years, patients with EP-guided antiarrhythmic therapy had a lower mortality than
patients randomized to no antiarrhythmic therapy (relative risk [RR] = 0.73, 95% CI = 0.53–
0.99); however, this difference was entirely driven by the subgroup of patients who received
ICD therapy. Patients who received an ICD had a lower mortality than patients randomized
to no anti-arrhythmic therapy and those who had an EP study that resulted in anti-arrhythmic
medication but no ICD (RR = 0.24, 95% CI = 0.13–0.45).
MADIT II (New England Journal of Medicine, 2002)
In 2002, the findings of MADIT II were released.11 This trial extended evidence of the ICD’s
effectiveness to patients with coronary artery disease and an ejection fraction less than or equal
to 30 percent without requiring clinicians to demonstrate inducible ventricular tachycardia
through an EP study. A total of 1,232 patients with a prior myocardial infarction and ejection
fraction less than or equal to 30 percent were enrolled and randomized to receive a single-lead
ICD versus conventional therapy. The primary endpoint was death from any cause. The mean
age of patients was 64 years; nearly 90 percent had an interval of less than six months from
their qualifying myocardial infarction. Over a mean follow-up period of 20 months, patients
Case Study 5: Implantable Cardioverter-Defibrillator
69
enrolled in the ICD group had improved survival as compared with patients in the non-ICD,
conventional therapy arm (HR = 0.69, 95% CI = 0.51–0.93, p = 0.016). As compared with
MADIT, in which the mortality benefit for ICD therapy over conventional therapy was apparent immediately, the mortality curves for the two arms in MADIT II diverged only after one
year.
SCD-HeFT (New England Journal of Medicine, 2005)
The Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) further extended evidence of
the value of ICD therapy to patients with and without ischemic heart disease as the underlying cause of systolic heart failure.12 It serves as the basis for current guidelines and payment
coverage in this population. This randomized trial, which was sponsored by the NHLBI and
industry, is notable in that it examined patients with and without coronary artery disease who
had systolic heart failure with an ejection fraction less than or equal to 35 percent. It randomized 2,521 patients to standard heart failure therapy, plus either placebo, shock only/singlelead ICD, or amiodarone. The primary endpoint was death from any cause, and the mean
follow-up period was 45.5 months. The mean age of patients was 60 years, and nearly half of
the patients enrolled had no history of coronary artery disease. As compared with placebo,
amiodarone was associated with a similar risk of death (HR = 1.07, 95% CI = 0.86–1.83, p =
0.53); ICD therapy was associated with a lower risk of death (HR = 0.77, 0.62–0.96, 0.07). Of
note, the mortality curves for the three arms did not diverge until approximately 20 months.
DEFINITE (New England Journal of Medicine, 2004)
One year later, the less definitive Defibrillators in Nonischemic Cardiomyopathy Treatment
Evaluation (DEFINITE) trial was released.13 Patients with nonischemic cardiomyopathy and
an ejection fraction less than or equal to 35 percent (n = 458) were randomized to standard care
versus standard care plus ICD implantation. The primary endpoint was all-cause mortality. At
two years, the ICD group had a nonsignificant reduction in mortality compared with the medical therapy arm (8.1 percent versus 14.1 percent, p = 0.06). The authors speculated that the
unexpected high rates of nonarrhythmic death in both groups contributed to the study being
underpowered to find a difference in arrhythmic deaths. Interestingly, the sponsor of the study
is not mentioned in the article or accompanying disclosures. Although the study was published
in a respected, peer-reviewed journal, it is not listed like the others on clinicaltrials.gov.
While the primary endpoint in most of the clinical trials evaluating ICD therapy was
all-cause mortality, the prevention of arrhythmic deaths drives the effect. In the DEFINITE
trial, the authors argued as much when they noted that the nonsignificant association of ICD
therapy with all-cause mortality was due to an unexpectedly low occurrence of arrhythmic
deaths in both treatment and control groups.
The counterargument is that clinical trials are supposed to determine effectiveness. From
the patient’s point of view, death is death, regardless of the cause. This is particularly germane
when an ICD is implanted in a patient with severe disease, such as nonischemic cardiomyopathy and low ejection fraction. Although it is accepted practice to implant ICDs, a close look at
the evidence reveals discordant findings.12, 13
Not All Groups Benefit from Receiving ICDs
There are subgroups for whom ICD therapy provides little or no benefit. Based on these trials,
ICDs should not be implanted in the immediate 40 days following acute myocardial infarction
70
Redirecting Innovation in U.S. Health Care: Case Studies
or in the three-month period following coronary artery bypass surgery or percutaneous transcatheter coronary angioplasty. These studies include the Coronary Artery Bypass Graft Patch
Trial (CABG-Patch),14 the Defibrillator in Acute Myocardial Infarction Trial (DINAMIT),15
and the Immediate Risk-Stratification Improves Survival (IRIS) trials.16
CABG-Patch (New England Journal of Medicine, 1997)
In this study, supported by NHLBI and Guidant/CPI, patients undergoing coronary artery
bypass surgery who had an ejection fraction less than or equal to 35 percent and an abnormal
signal-average electrocardiogram were randomized to receive an ICD during surgery versus
no ICD. The primary endpoint was all-cause mortality. Among 1,055 patients, there was no
evidence of improved survival among patients receiving ICD therapy (HR = 1.07, 95% CI =
0.81–1.42, p = 0.64).
DINAMIT (2004)
224 patients with recent MI (6 to 40 days post-MI), ejection fraction less than or equal to
30 percent, and reduced heart rate variability by Holter monitor were randomized to ICD versus
no ICD. The primary endpoint was death from any cause. Mean follow-up was 30 months.
While there were fewer deaths due to arrhythmias in the ICD group, all-cause mortality did
not differ (HR = 1.08. 95% CI = 0.76–1.55).
IRIS (2009)
This trial enrolled a similar population of patients as the DINAMIT trial, with nearly identical
results. Specifically, 898 patients early post-MI, with ejection fraction less than 40 percent and
either HR >90 or nonsustained ventricular tachycardia, were randomized to ICD versus no
ICD. Similar to the results of DINAMIT, there was a decreased risk of death due to arrhythmia in the ICD group, but no overall survival benefit for patients with ICD (HR = 1.04,
95% CI = 0.81–1.35).
Collectively, these trials form the basis for current AHA guidelines for primary prevention of SCD (Box 5.3).
Box 5.3
Class I Recommendations for ICD Therapy as Primary Prevention in Patients with Heart Failure 8
• ICD therapy is indicated in patients with an LVEF ≤35% due to prior MI, who are at
least 40 days post MI, and who have NYHA Class II or III (Level of Evidence. A)
• ICD therapy is indicated in patients with an LVEF ≤35% due to non-ischemic cardiomyopathy, and who have NYHA Class II or III (Level of Evidence: B).
SOURCE: Members WC, Epstein AE, DiMarco JP, et al., “ACC/AHA/HRS 2008 Guidelines for Device-Based
Therapy of Cardiac Rhythm Abnormalities: Executive Summary—A Report of the American College of
Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the
ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices):
Developed in Collaboration With the American Association for Thoracic Surgery and Society of Thoracic
Surgeons,” Circulation, Vol. 117, No. 21, May 27, 2008, pp. 2820–2840. 8
NOTES: LVEF = left ventricular ejection fraction, MI = myocardial infarction, NYHA = New York Heart
Association, VT = ventricular tachycardia.
Case Study 5: Implantable Cardioverter-Defibrillator
71
Limitations of Clinical Trials
Although current guidelines for use of ICDs are based on clinical trial data, participants in
clinical trials do not always resemble the community-based populations that subsequently
receive treatment.17 Clinical trials enroll carefully selected study candidates and typically
manage them under ideal conditions that are difficult to replicate in community settings.
Certain patient subgroups (e.g., the elderly and patients with multiple comorbidities) are often
underrepresented in clinical trials.18 For this reason, ICD therapy may not be appropriate for
some patients who technically meet American Heart Association (AHA) guidelines.
Subsequent Use
Because ICDs are expensive (the typical price is about $20,000–$30,000), and their implantation is costly as well (typical procedure fees are about $35,000), diffusion has been largely
driven by whether or not public and private payers will reimburse for the technology and its
use.19, 20 Without insurance coverage, few patients could afford an ICD. And because private
insurance companies generally follow the lead of Medicare, the Centers for Medicare and
Medicaid Services (CMS) has played a highly important role in enabling the diffusion of ICDs.
Once ICDs were shown to be beneficial in the relatively small number of individuals who
survive a prior episode of SCD or protracted bouts of ventricular tachychardia, Medicare’s
decision to cover their use in this patient population was relatively straightforward. This is why,
only one year after the FDA approved the sale of ICDs, CMS issued its first Medicare National
Coverage Determination (NCD) in 1986.
Because the benefits were less clear-cut and the financial consequences were much greater,
the decision regarding whether or not to expand coverage of ICD therapy from secondary to
primary prevention was much more controversial. Nevertheless, as more clinical trials were
reported, CMS expanded its coverage policies in 1991, 1999, 2003, and 2004.
Medicare Expands Coverage to Include Primary Prevention
A pivotal moment occurred in 2003, when the Medicare Coverage Advisory Committee
(MCAC) convened to discuss ICD therapy for primary prevention.21 At the time, the results
of MADIT-II were available; however, the SCD-HeFT trial, which was testing ICD therapy
in an expanded population, was still ongoing. MCAC recommended adopting the same inclusion and exclusion criteria as the MADIT-II trial, which included patients with ischemic heart
disease and reduced systolic function (ejection fraction less than 30 percent).
Concerned by the cost implications of offering ICDs to such a large number of patients,
CMS ultimately opted to cover ICDs to a subgroup of the MADIT-II population, which the
analysis suggested would derive the most benefit. Many in the research and clinical communities objected to this decision, since subgroup analyses are prone to bias and random effects.
Others questioned whether MADIT-II provided sufficient evidence to move forward with
expansion of coverage. The remarks of Mark Hlatky, who presented to CMS, illustrate this
perspective:
I think the big question is whether an EF below 30 percent in and of itself is sufficient to
put in an ICD, and I would say that the question here is whether the evidence is adequate.
I would say MADIT II is suggestive, it’s highly suggestive, but it doesn’t really prove the
72
Redirecting Innovation in U.S. Health Care: Case Studies
case completely for this. The word that was used earlier by Dr. Moss and the representative
of the company was a paradigm shift, a paradigm shift to say that we don’t need any additional markers of patients with low EF. And I question that because this is a single study;
it’s very well done, but it’s only a single study. And I think we have 25 years of research that
says that there are other markers that are important and for that reason I am concerned
that an indication from Medicare that says that ejection fraction alone is necessary to put
in an ICD is overly broad, and would expose many patients who would not benefit from
this device to risks, to say nothing of the large costs to the program.21
Another public presenter, Dr. Joanne Lynn, had similar concerns:
In sum, I would recommend that the Medicare Coverage Advisory Committee do the following: First, advise CMS to issue a national coverage determination for ICDs only for the
populations where evidence is strong that they actually gain desired outcomes, which may
mean that only a very small part of the Medicare population should be covered now, and
certainly does not now include elderly who have multiple comorbidities and competing
causes of death. Second, we should call on CMS to insure that Medicare patients have a
high standard of informed consent. We should recommend that CMS institute methods to
monitor outcomes, that they require evidence about all of the outcomes, including quality
of life. That they monitor changes in the performance over time, and call on various parties
to take up discussion of the priorities and values that are at stake.
As CMS was considering this cautionary testimony, Medtronic Inc. released preliminary
data from the SCD-HeFT trial, which included patients with ejection fractions less than or
equal to 30 percent due either to nonischemic or ischemic heart disease. Based on these findings, the company pressed CMS to expand coverage to include a larger group of patients. In
the course of considering Medtronic’s request, CMS met with the makers of the other ICDs,
including Boston Scientific and Guidant, as well as the Heart Rhythm Society and the American College of Cardiology. Additionally, they solicited public comment on three separate occasions. Ultimately, CMS determined that ICDs were “reasonable and necessary” for primary
prevention of SCD in patients with a reduced ejection fraction due either to ischemic or nonischemic heart failure.22 This decision dramatically expanded the number of patients eligible
for ICDs.
Medicare Approves Further Expansion but Specifies “Coverage with Evidence
Development”
To ensure that its decision produced the benefits that manufacturers and experts promised,
CMS added an important caveat to its decision: It would reimburse providers for ICD implantations, but only if patients who received an ICD for primary prevention of SCD were enrolled
in an FDA-approved clinical trial, a clinical trial managed by CMS, or another organization’s
approved treatment registry.23 The policy, known as “coverage with evidence development,”
is similar to the FDA’s “accelerated approval” process (see the case study in this appendix on
Avastin for metastatic breast cancer).
To help its members comply with CMS’s requirement, the American College of Cardiology established an ICD registry under the National Cardiovascular Data Registry (NCDR).
This registry began collecting data in 2006. Hospitals are only required to submit data to
NCDR on ICD implants involving Medicare beneficiaries. However, less than 75 percent of
Case Study 5: Implantable Cardioverter-Defibrillator
73
participating hospitals report data on all of their ICD implantations, including those provided
to privately insured patients.
As a result of progressively expanded coverage, implantations of ICDs grew by 20–30 percent annually between 1985, the year ICDs secured FDA approval, and the mid-2000s.24
A total of 415,780 ICDs were implanted in the 12-year interval between 1990–2002.25 In
the four years between 2006–2010, over half a million ICD implants were registered in the
National Cardiovascular Data Registry, which captures 90 percent of ICDs implanted in the
United States. This prodigious growth was primarily driven by the expansion of coverage to
primary prevention of SCD, which now accounts for four out of every five ICD implants.
Interestingly, despite Medicare’s decision to further expand the range of patients eligible
for ICDs in 2005, sales of devices started to decline the following year. They have continued to
fall through 2012, the most recent year for which data are available.26 The reasons for this trend
are not entirely clear. There are plenty of eligible patients; Hernandez et al. noted in 2007 that
only 35 percent of hospitalized patients with a class I indication for ICD therapy have received
an ICD or had a plan for future implantation.27 Women and African American patients with
class I indications were less likely to receive ICD therapy than white men, even after controlling for age, insurance status, and comorbidities.
The decline in enthusiasm for ICD therapy may, in part, reflect (1) growing recognition that the types of patients who are being treated today with ICD therapy are quite different than the patients who participated in the original clinical trials, so the effectiveness of
ICD treatment in these populations may be less; (2) growing recognition that the devices can
inappropriately deliver one or more shocks without warning (i.e., the device falsely senses an
arrhythmia and tries to treat it; ICD shocks, particularly false ones, are painful, costly, and
substantially degrade a patient’s quality of life); and (3) growing fears of legal liability from
complications of treatment (such as infection) and device malfunctions due to lead failures.28
ICDs Are Sometimes Implanted in Patients Who Are Unlikely to Benefit
Today, more than 40 percent of ICDs are implanted in patients over 70 years of age. Approximately 10 percent are implanted in patients 80 years of age or older.29 This age distribution is
older than that of the early clinical trials establishing the efficacy of ICDs for prevention of
SCD. In the SCD-HeFT trial, which is the basis for current treatment guidelines and CMS
payment policy, approximately a third of subjects were 65 years of age or older, and 9 percent
were over 75 years of age.
ICD implantation is less beneficial in older patients than younger ones. In the MADIT-II
trial, patients over 65 had less risk reduction, and in the SCD-HeFT trial, the older age group
had no survival benefit from ICD implantation. Even when data were pooled in a meta-analysis
combining data from MUSST, MADIT-II, DEFINITE, and SCD-HeFT, ICDs were associated with a nonsignificant trend of lower all-cause mortality in the age 65 and older group (n =
3,562, HR = 0.66, 95% CI: 0.5–0.87, p = 0.15) that became statistically significant in the over
age 75 group (n = 579; HR = 0.73, 95% CI 0.51–0.97, p =–0.03). However, this study found
no evidence that ICDs improved rates of survival in older patients from SCD, the condition
the devices are intended to treat.
Based on the findings of clinical trials, AHA guidelines recommended against implanting
ICDs in certain clinical contexts. Unfortunately, clinicians often disregard this guidance. For
example, the AHA’s guidelines recommend against ICD therapy in patients who are undergoing bypass surgery or in the early period after a myocardial infarction (because there is no
74
Redirecting Innovation in U.S. Health Care: Case Studies
clinical benefit). The guidelines also advise against ICD use in the first three months following coronary revascularization and in patients with New York Heart Association (NYHA)
class IV symptoms or newly diagnosed heart failure (populations that were excluded from the
above clinical trials). Despite the clarity of this guidance, a recently published study showed
that among 25,000 patients who received an ICD and were enrolled in the National Cardiovascular Data Registry-ICD Registry, nearly one in four—22.5 percent—received an ICD for
a non–evidence-based indication.19
In Canada, ICD implants are more restricted than in the United States because of their
high cost. As a result, only about a fifth as many eligible patients receive ICD therapy in
Canada as in the United States.30 The impact of these differences on outcomes has not been
examined.
There is emerging evidence that high-risk, “real-world” patients treated in community
settings derive similar benefits to those documented among participants in clinical trials. For
example, Chan et al. analyzed a prospective cohort of 965 “real-world” patients with ischemic
and nonischemic heart failure enrolled in a multisite registry.31 They found that ICDs were
effective in reducing mortality and that the magnitude of risk reduction was similar to that
observed in clinical trials (HR = 0.69, 95% CI = 0.05–0.96, p = 0.03). Furthermore, they
found that the survival benefit persisted in patients of all age groups, including those over age
75, as well as in patients with multiple comorbidities.32 An analysis of Medicare beneficiaries
hospitalized for heart failure during 2003–3005 compared patients with heart failure and an
ICD with a propensity score–matched group of patients hospitalized for heart failure but with
no ICD.19 The mean age was 76 years in the ICD group and 77 years in the non-ICD group.
Similarly, the authors found a significant survival benefit among patients with an ICD (HR
= 0.62, 95% CI = 0.58–0.67). The authors concluded that ICD therapy provides substantial
health benefits to Medicare beneficiaries.
Interestingly, both of these studies conclude that although patients of older age and greater
comorbidity burden are at a higher risk for all-cause mortality, this does not preclude them
from deriving benefit from ICD therapy. In fact, the absolute risk reduction in these high-risk
populations may be greater. The authors acknowledge, however, that the cost-effectiveness
of ICDs in these high-risk patients is less, due to the substantially higher rate of all-cause
mortality.
Despite evidence that ICDs can benefit appropriately selected elderly patients, there is
little enthusiasm for the technology in this age group.33, 34 One reason is the concept of “competing mortality”—the realization that as patients grow older, they may die from a cause
unrelated to their ICD. Yung et al. evaluated rates of overall mortality in a prospective cohort
of adults who received ICDs in Ontario, Canada.35 They found increasing risk of mortality in
each respective age group, although the rates of appropriate shocks were similar in each cohort.
A recent analysis determined that a third of Medicare recipients who received an ICD for primary prevention died within three years of implantation. In this study, the number of hospitalizations for heart failure was highly predictive of prognosis. Among patients with no prior
hospitalizations, the three-year mortality rate was 27 percent. It was 63 percent in patients with
3 or more prior hospitalizations.36
Case Study 5: Implantable Cardioverter-Defibrillator
75
Health and Cost Impact
ICDs are intended to recognize and automatically treat life-threatening arrhythmias, namely
ventricular tachycardia or fibrillation. As noted in earlier sections, they have been shown to
significantly reduce rates of death from SCD when implanted in properly selected patients.12, 37
Although one might expect, given the large number of devices implanted over the last 20 years,
that ICDs have had a favorable impact on population rates of SCD, there are no published
studies that quantify the contribution of this technology to an overall change in the incidence
of this condition.
Inappropriate Shocks
Although the literature indicates that ICDs have benefited many patients who had them
implanted, they are not risk-free. Sometimes, the decision to implant an ICD leads to serious complications or side effects. For example, an ICD may falsely identify other, nonfatal
arrhythmias (e.g., atrial flutter or sinus tachycardia) as ventricular arrhythmias and inappropriately deliver painful shocks. Both appropriate and inappropriate shocks are common, though
their frequency varies depending on the patient’s baseline risk for ventricular arrhythmias.
Newer ICD programming has improved rhythm detection, but it has not entirely eliminated
this problem. In the AVID trial for secondary prevention, rates of ventricular arrhythmias
and subsequent shock therapy were 35 percent at three months, 53 percent at one year, and
68 percent at two years. Among 194,100 patients who received an ICD for either primary
or secondary prevention and were enrolled in a telemonitoring system by Boston Scientific,
23 percent received one or more inappropriate shocks, and 17 percent experienced one or more
inappropriate shocks within five years of implantation.38 The declining rate of delivered shocks
reported in recent studies probably reflects the greater proportion of implantations for primary
prevention of SCD (currently, about 80 percent).1 Heart failure patients have a higher risk of
cardiac arrhythmias than age-matched normal patients, but it is much lower than the risk
among patients who have had a prior SCD or bout of ventricular tachycardia.
Irrespective of whether a shock is appropriately delivered or not, it is painful and, for
some, terrifying. Patients frequently liken the sensation to “being kicked by a mule” or “an
earthquake.” Often, the recipient’s quality of life and health are negatively impacted.39 Compared with ICD patients who have never been shocked, those who experience one or more
ICD shocks have higher rates of anxiety and depression. Receiving ICD shocks is also associated with an increased risk of hospitalizations and subsequent mortality, most likely because
delivery of a shock is an indicator of more advanced disease. Speculation that shocks produce
myocardial injury has not been borne out.38
Surgical Complications
ICD implantation is associated with both periprocedural and long-term adverse events. During
hospitalization for ICD implantation, there is an approximate 3-percent risk of having a complication, the most common being lead dislodgement (1%), hematoma (0.9%), pneumothorax
(0.4%), and cardiac arrest (0.3%). The risk of these events is higher if a dual chamber or
a biventricular device is placed because these are more complicated procedures.40, 41 ICDs
implanted by non–electrophysiology-trained cardiologists and by thoracic surgeons are associated with higher complication rates than those implanted by electrophysiology specialists.1
Additionally, cases in which an ICD is placed for a non–evidence-based indication are also
76
Redirecting Innovation in U.S. Health Care: Case Studies
associated with higher complication rates.42 It is possible that these trends are collinear, but
this has not been proven.
Infection
Infection of the subcutaneous pocket that holds the ICD or the device’s intracardiac leads
(resulting in endocarditis) occurs in approximately 1–3 percent of implantations. Studies indicate that the rates of infection are rising. In 2008, the rate of infection was 2.8 percent, as
compared with 1.5 percent in 2003.42 The increase in infections is attributed to the expanded
use of ICDs in patients with multiple comorbidities because sicker patients are more prone to
infections. An infection of an ICD is costly, both to the patient and to the health care system.
Frequently, treatment requires protracted administration of powerful intravenous antibiotics
and perhaps surgery to replace and or remove infected leads and/or the pulse generator. The
societal costs of an iatrogenic infection are approximately $146,000 per case.43
Lead Failures
Fractures in an ICD lead, problems with lead insulation, and other device malfunctions can
result in undersensing of ventricular arrhythmias or detection of false signals, which trigger
inappropriate shocks. The true rate of lead failure is unknown, though it has been estimated
to occur at a rate of 0.5 percent per year. The Guidant Endotak DSP (model 0125) and the
Medtronic Transvene (model 6936) were found to have high lead failure rates. Two recent
recalls of the Medtronic Fidelis Lead (2007) and the St. Jude Riata lead (2011) have raised
awareness among both patients and clinicians of the problem of lead failures. In the case of
the Medtronic lead, there were two fracture sites along the conductor that were responsible
for high rates of inappropriate shocks and preventable death. The Fidelis lead was implanted
in 205,600 patients in the United States, less than a quarter of whom had the lead extracted.
This leaves approximately 150,000 Fidelis leads in active use.44 While these widely publicized
recalls have been implicated as a reason for the decline in ICD implants, research shows that
the trends were already occurring and that lead recall did not significantly impact utilization.45
Are ICDs Cost-Effective?
If, as is widely assumed, ICDs implanted for primary prevention produce the same survival
benefits observed in the published clinical trials, the estimated cost of implanting an ICD for
primary prevention of SCD is between $34,000 and $72,000 per quality-adjusted life year
(QALY) gained.46 While this cost is high, it is in the general ballpark with the cost of treating
end-stage renal disease, which Medicare has used as a reference point for comparisons of costeffectiveness. Using mortality results from the SCD-HeFT trial as a reference, another analysis
estimated the cost of an ICD to be $38,389 per life-year saved.47 An analysis among Medicare
beneficiaries found that treatment of ICD recipients cost $41,542 more in the first 30 days
following implantation than propensity-matched heart failure patients who were hospitalized
but did not receive an ICD.19 This cost difference persisted for up to a year, with relatively little
incremental costs after the initial hospitalization and the immediate 30-day follow-up period.
There are, however, limitations to the cost-effectiveness studies conducted to date. None
of these analyses has factored in the cost of potential complications or the impact of complications, such as infection or inappropriate shocks, on quality of life. One can easily imagine that
the cost of caring for a patient with multiple inappropriate shocks, especially if that person
Case Study 5: Implantable Cardioverter-Defibrillator
77
was nearing the end of life, would be much greater than that reflected in a simple cost-benefit
analysis.
A second concern is that the cost-effectiveness analyses produced to date were based on
the assumption that a single-chamber ICD device was used. However, in the United States
today, more than two-thirds of patients receive a far more expensive dual-chamber device,
which involves the placement of an extra atrial lead. Although the placement of an atrial lead
is useful in the relatively few patients who require atrial pacing, six out of ten patients who
receive a dual-chamber device do not need the extra lead or the device’s added complexity.
Advocates of dual-chambered devices say that they implant them based on theoretical benefits of improved rhythm interpretation, perhaps resulting in fewer inappropriate shocks and
enhanced efficiency, should there ever be a future need for pacing.
However, contrary to these beliefs, dual-chamber devices have not been shown to improve
outcomes. They are also associated with more complications than less expensive single-chamber
devices.40
“Treatment Creep”
ICDs represent the technical prowess of American biomedical engineering. They have benefited
thousands of Americans, many of whom might not be alive today without them. Rigorous,
prospective randomized trials involving carefully selected groups of patients have consistently
demonstrated a 30–40-percent relative risk reduction in mortality at a cost per QALY comparable to that of renal dialysis. This progress has come at a cost: Annual expenditures for ICD
therapy are estimated to be $4.6 billion per year, accounting for approximately 100,000 annual
ICD implantations. It has been estimated that if all of the 500,000 people who meet current
clinical guidelines for implantation of an ICD got one, annual expenditures would rise to
$15 billion.23, 48, 49
Unlike many therapies that slowly penetrate the marketplace, ICDs quickly worked their
way into clinical practice. The rapid diffusion of the technology was likely driven in part by
a series of positive clinical trials, in part by manufacturer encouragement, and, most importantly, by the willingness of CMS to progressively expand the population of patients who
qualify for coverage.
Professional societies and groups like the AHA have attempted to place reasonable boundaries on the use of ICDs. For example, they endorse consideration of ICD therapy for primary
prevention only when patients are receiving optimal medical therapy, have at least one year of
life expectancy, and are in good functional health. However, defining the life expectancy of a
patient with heart failure is notoriously difficult, particularly in older adults.
Unfortunately, these guidelines are often ignored. For example, experts recommend
against ICD therapy in patients who are undergoing bypass surgery or in the early period after
a myocardial infarction. They also advise against implanting ICDs in the first three months
following coronary revascularization, in heart failure patients with severe (i.e., NYHA class
IV) symptoms, and those with newly diagnosed heart failure. Nevertheless, a recently published study found that 22.5 percent of roughly 25,000 patients enrolled in the National Cardiovascular Data Registry-ICD Registry received an ICD for a non–evidence-based reason.50
Those who inappropriately received an ICD had a higher rate of in-hospital mortality and more
periprocedural complications than patients who received guideline-concordant treatment.
78
Redirecting Innovation in U.S. Health Care: Case Studies
Studies like this would not be possible without CMS’s “coverage with evidence development” policy that requires certain costly technologies, such as ICDs, to contribute data to
registries to monitor utilization, track outcomes, and identify safety concerns (such as device
malfunction). The policy should be used more often to monitor such costly technologies as
orthopedic implants and proton beam therapy.
The finding that many ICDs are inappropriately implanted also argues for active involvement of patients in decisions about the pros and cons of implanting an ICD.51 Early on, patient
information literature on the technology lacked balance.52, 53 Cardiologists who only follow
clinical guidelines, rather than considering patient preferences, could also push some patients
to accept a treatment they later regret. This may explain why many heart failure patients feel
inadequately informed about the prognosis of their disease and the ability of an ICD to extend
their life.54
With increasing data on the effectiveness of ICDs, cardiologists can provide moreindividualized assessments of likely benefits and risks. Matlock et al. are developing and testing
“decision aids” intended to help patients understand available data about benefits and tradeoffs and weigh their own preferences about ICD implantation.55 Respecting a patient’s values
and preferences regarding ICD implantation (and its subsequent deactivation) can improve
outcomes and reduce unwanted costs.56
More must be done to deter inappropriate implantations, as well as the inappropriate use
of more costly dual-chambered devices that offer no added benefit to patients and may even
increase the risk of harm. However, in carefully selected patients, implantation of an ICD saves
lives at a price per QALY in line with other costly but widely accepted life-sustaining technologies, such as renal dialysis. The challenge with ICDs and other high-cost, high-technology
devices like them is limiting their use to patients who are most likely to benefit.
References
1
Curtis JP, Luebbert JJ, Wang Y, et al., “Association of Physician Certification and Outcomes Among
Patients Receiving an Implantable Cardioverter-Defibrillator,” Journal of the American Medical Association,
Vol. 301, No. 16, April 22, 2009, pp. 1661–1670.
2 Zipes DP and Wellens HJJ, “Sudden Cardiac Death,” Circulation, Vol. 98, No. 21, November 24, 1998,
pp. 2334–2351.
3
Lown B, Amarasingham R, and Neuman J, “New Method for Terminating Cardiac Arrhythmias—Use of
Synchronized Capacitor Discharge, Journal of the American Medical Association, Vol. 182, November 3, 1962,
pp. 548–555.
4
Eisenberg M, “The Resuscitation Greats: Bernard Lown and Defibrillation,” Resuscitation, Vol. 69, No. 2,
May 2006, pp. 171–173.
5
“A Comparison of Antiarrhythmic-Drug Therapy with Implantable Defibrillators in Patients Resuscitated from Near-Fatal Ventricular Arrhythmias,” New England Journal of Medicine, Vol. 337, No. 22, 1997,
pp. 1576–1584.
6
Kuck KH, Cappato R, Siebels J, and Ruppel R, “Randomized Comparison of Antiarrhythmic Drug Therapy with Implantable Defibrillators in Patients Resuscitated from Cardiac Arrest: The Cardiac Arrest Study
Hamburg (CASH),” Circulation, Vol. 102, No. 7, August 15, 2000, pp. 748–754.
Case Study 5: Implantable Cardioverter-Defibrillator
79
7
Connolly SJ, Gent M, Roberts RS, et al., “Canadian Implantable Defibrillator Study (CIDS): A Randomized Trial of the Implantable Cardioverter Defibrillator Against Amiodarone,” Circulation, Vol. 101, No. 11,
March 21, 2000, pp. 1297–1302.
8
Members WC, Epstein AE, DiMarco JP, et al., “ACC/AHA/HRS 2008 Guidelines for Device-Based
Therapy of Cardiac Rhythm Abnormalities: Executive Summary—A Report of the American College of
Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise
the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia
Devices): Developed in Collaboration With the American Association for Thoracic Surgery and Society of
Thoracic Surgeons,” Circulation, Vol. 117, No. 21, May 27, 2008, pp. 2820–2840.
9
Moss AJ, Hall WJ, Cannom DS, et al., “Improved Survival with an Implanted Defibrillator in Patients
with Coronary Disease at High Risk for Ventricular Arrhythmia,” New England Journal of Medicine, Vol. 335,
No. 26, 1996, pp. 1933–1940.
10
Buxton AE, Lee KL, Fisher JD, Josephson ME, Prystowsky EN, and Hafley G, “A Randomized Study of
the Prevention of Sudden Death in Patients with Coronary Artery Disease,” New England Journal of Medicine,
Vol. 341, No. 25, 1999, pp. 1882–1890.
11
Moss AJ, Zareba W, Hall WJ, et al., “Prophylactic Implantation of a Defibrillator in Patients with Myocardial Infarction and Reduced Ejection Fraction,” New England Journal of Medicine, Vol. 346, No. 12, 2002,
pp. 877–883.
12
Bardy GH, Lee KL, Mark DB, et al., “Amiodarone or an Implantable Cardioverter-Defibrillator for Congestive Heart Failure,” New England Journal of Medicine, Vol. 352, No. 3, January 20, 2005, pp. 225–237.
13
Kadish A, Dyer A, Daubert JP, et al., “Prophylactic Defibrillator Implantation in Patients with Nonischemic Dilated Cardiomyopathy,” New England Journal of Medicine, Vol. 350, No. 21, May 20, 2004,
pp. 2151–2158.
14
Bigger JT, “Prophylactic Use of Implanted Cardiac Defibrillators in Patients at High Risk for Ventricular Arrhythmias After Coronary-Artery Bypass Graft Surgery,” New England Journal of Medicine, Vol. 337,
No. 22, 1997, pp. 1569–1575.
15
Hohnloser SH, Kuck KH, Dorian P, et al., “Prophylactic Use of an Implantable Cardioverter–Defibrillator After Acute Myocardial Infarction,” New England Journal of Medicine, Vol. 351, No. 24, 2004,
pp. 2481–2488.
16
Steinbeck G, Andresen D, Seidl K, et al., “Defibrillator Implantation Early After Myocardial Infarction,”
New England Journal of Medicine, Vol. 361, No. 15, 2009, pp. 1427–1436.
17
Masoudi FA, Go AS, Magid DJ, et al., “Longitudinal Study of Implantable Cardioverter-Defibrillators:
Methods and Clinical Characteristics of Patients Receiving Implantable Cardioverter-Defibrillators for Primary Prevention in Contemporary Practice,” Circulation; Cardiovascular Quality and Outcomes, Vol. 5, No. 6,
November 2012, pp. e78–e85.
18
Heiat A, Gross CP, and Krumholz HM, “Representation of the Elderly, Women, and Minorities in Heart
Failure Clinical Trials,” Archives of Internal Medicine, Vol. 162, No. 15, August 12–26, 2002, pp. 1682–1688.
19
Groeneveld PW, Farmer SA, Suh JJ, Matta MA, and Yang F, “Outcomes and Costs of Implantable
Cardioverter-Defibrillators for Primary Prevention of Sudden Cardiac Death Among the Elderly,” Heart
Rhythm: The Official Journal of the Heart Rhythm Society, Vol. 5, No. 5, May 2008, pp. 646–653.
20
Grady D, “Many Defibrillators Implanted Unnecessarily, Study Says,” New York Times, June 3, 2011.
21
Centers for Medicare & Medicaid Services, “Decision Memo for Implantable Cardioverter Defibrillators
(ICDs) (CAG-00157N),” 2003.
22
Centers for Medicare & Medicaid Services, “Decision Making for Implantable Defibrillators
(CAG-00157R3),” 2004.
80
Redirecting Innovation in U.S. Health Care: Case Studies
23 McClellan MB and Tunis SR, “Medicare Coverage of ICDs,” New England Journal of Medicine, Vol. 352,
No. 3, January 20, 2005, pp. 222–224.
24
Al-Khatib SM, Sanders GD, Mark DB, et al., “Implantable Cardioverter Defibrillators and Cardiac Resynchronization Therapy in Patients with Left Ventricular Dysfunction: Randomized Trial Evidence Through
2004,” American Heart Journal, Vol. 149, No. 6, June 2005, pp. 1020–1034.
25
Maisel WH, Moynahan M, Zuckerman BD, et al., “Pacemaker and ICD Generator Malfunctions: Analysis of Food and Drug Administration Annual Reports,” Journal of the American Medical Association, Vol. 295,
No. 16, April 26, 2006, pp. 1901–1906.
26 Perrone
M, “Boston Scientific Reports 25 Pct Fall in 3Q Profit,” October 2011. As of January 14, 2014:
http://www.boston.com/news/local/massachusetts/articles/2011/10/20/
boston_scientific_reports_25_pct_fall_in_3q_profit/
27
Hernandez AF, Fonarow GC, Liang L, et al., “Sex and Racial Differences in the Use of Implantable Cardioverter-Defibrillators Among Patients Hospitalized with Heart Failure,” Journal of the American Medical
Association, Vol. 298, No. 13, October 3, 2007, pp. 1525–1532.
28
Feder BJ, “Defibrillators Are Life Saver, but Risks Give Pause,” New York Times, September 13, 2008,
p. C:1.
29
Epstein AE, Kay GN, Plumb VJ, et al., “Implantable Cardioverter-Defibrillator Prescription in the Elderly,”
Heart Rhythm: The Official Journal of the Heart Rhythm Society, Vol. 6, No. 8, August 2009, pp. 1136–1143.
30
Simpson CS, Hoffmaster B, and Dorian P, “Downward Delegation of Implantable Cardioverter Defibrillator Decision-Making in a Restricted-Resource Environment: The Pitfalls of Bedside Rationing,” Canadian
Journal of Cardiology, Vol. 21, No. 7, May 15, 2005, pp. 595–599.
31
Al-Khatib SM, Hellkamp A, Bardy GH, et al., “Survival of Patients Receiving a Primary Prevention
Implantable Cardioverter-Defibrillator in Clinical Practice Vs Clinical Trials,” Journal of the American Medical
Association, Vol. 309, No. 1, January 2, 2013, pp. 55–62.
32
Chan PS, Nallamothu BK, Spertus JA, et al., “Impact of Age and Medical Comorbidity on the Effectiveness of Implantable Cardioverter-Defibrillators for Primary Prevention,” Circulation: Cardiovascular Quality
and Outcomes, Vol. 2, No. 1, January 1, 2009, pp. 16–24.
33 Heidenreich PA and Tsai V, “Is Anyone Too Old for an Implantable Cardioverter-Defibrillator?” Circulation: Cardiovascular Quality and Outcomes, Vol. 2, No. 1, January 1, 2009, pp. 6–8.
34 Mandrola
J, “Elderly Patients Considered for an ICD: Facing Facts, Mastering the Obvious, and Sharing
the Decision,” Trials and Fibrillations, theheart.org, 2013. As of January 16, 2014:
http://www.theheart.org/columns/trials-and-fibrillations-with-dr-john-mandrola/
elderly-patients-considered-for-an-icd-facing-facts-mastering-the-obvious-and-sharing-the-decision.do
35
Yung D, Birnie D, Dorian P, et al., “Survival After Implantable Cardioverter-Defibrillator Implantation in
the Elderly,” Circulation, Vol. 127, No. 24, June 18, 2013, pp. 2383–2392.
36
Chen CY, Stevenson LW, Stewart GC, et al., “Impact of Baseline Heart Failure Burden on Post-Implantable Cardioverter-Defibrillator Mortality Among Medicare Beneficiaries,” Journal of the American College of
Cardiology, Vol. 61, No. 21, May 28, 2013, pp. 2142–2150.
37
Connolly SJ, Hallstrom AP, Cappato R, et al., “Meta-Analysis of the Implantable Cardioverter Defibrillator
Secondary Prevention Trials: AVID, CASH and CIDS Studies; Antiarrhythmics Vs Implantable Defibrillator
Study; Cardiac Arrest Study Hamburg; Canadian Implantable Defibrillator Study,” European Heart Journal,
Vol. 21, No. 24, December 2000, pp. 2071–2078.
38
Saxon LA, Hayes DL, Gilliam FR, et al., “Long-Term Outcome After ICD and CRT Implantation and
Influence of Remote Device Follow-Up: The ALTITUDE Survival Study,” Circulation, Vol. 122, No. 23,
December 7, 2010, pp. 2359–2367.
Case Study 5: Implantable Cardioverter-Defibrillator
81
39
Gehi AK, Mehta D, and Gomes JA, “Evaluation and Management of Patients After Implantable Cardioverter-Defibrillator Shock,” Journal of the American Medical Association, Vol. 296, No. 23, December 20, 2006,
pp. 2839–2847.
40 Peterson
PN, Varosy PD, Heidenreich PA, et al., “Association of Single- Vs Dual-Chamber ICDs with Mortality, Readmissions, and Complications Among Patients Receiving an ICD for Primary Prevention,” Journal
of the American Medical Association, Vol. 309, No. 19, May 15, 2013, pp. 2025–2034.
41
Freeman JV, Wang Y, Curtis JP, Heidenreich PA, and Hlatky MA, “Physician Procedure Volume and
Complications of Cardioverter-Defibrillator Implantation,” Circulation, Vol. 125, No. 1, January 3, 2012,
pp. 57–64.
42 Greenspon
AJ, Patel JD, Lau E, et al., “16-Year Trends in the Infection Burden for Pacemakers and
Implantable Cardioverter-Defibrillators in the United States 1993 to 2008,” Journal of the American College of
Cardiology, Vol. 58, No. 10, August 30, 2011, pp. 1001–1006.
43 DiMarco JP, “Implantable Cardioverter-Defibrillators,” New England Journal of Medicine, Vol. 349, No. 19,
November 6, 2003, pp. 1836–1847.
44 Kalahasty
G and Ellenbogen KA, “Management of the Patient with Implantable Cardioverter-Defibrillator
Lead Failure,” Circulation, Vol. 123, No. 12, March 29, 2011, pp. 1352–1354.
45
Bhatt P, Curtis J, Wang Y, and Lampert R, “Impact of a Recent Lead Recall on Utilization of Implantable
Cardioverter Defibrillators: Data from the NCDR ICD Registry,” Journal of Cardiovascular Electrophysiology,
Vol. 23, No. 8, August 2012, pp. 861–865.
46 Sanders
GD, Hlatky MA, and Owens DK, “Cost-Effectiveness of Implantable Cardioverter–Defibrillators,”
New England Journal of Medicine, Vol. 353, No. 14, 2005, pp. 1471–1480.
47
Mark DB, Nelson CL, Anstrom KJ, et al., “Cost-effectiveness of Defibrillator Therapy or Amiodarone in
Chronic Stable Heart Failure: Results from the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT),”
Circulation, Vol. 114, No. 2, July 11, 2006, pp. 135–142.
48
Estes NA, “Implantable Cardioverter Defibrillator and Cardiac Resynchronization Therapy in Patients with
Left Ventricular Dysfunction: Evidence-Based Medicine, Economics, and Guidelines,” Medscape, 2006. As of
January 16, 2014:
http://www.medscape.org/viewarticle/528664_7
49
International Society for Pharmacoeconomics and Outcomes Research, “Medicare Could Save Money and
Provide More Defibrillators,” 2013. As of January 15, 2014:
http://www.ispor.org/pressrelease/june06/defibrillators.asp
50
Al-Khatib SM, Hellkamp A, Curtis J, et al., “Non-Evidence-Based ICD Implantations in the United
States,” Journal of the American Medical Association, Vol. 305, No. 1, January 5, 2011, pp. 43–49.
51
Caverly TJ, Al-Khatib SM, Kutner JS, Masoudi FA, and Matlock DD, “Patient Preference in the Decision to Place Implantable Cardioverter-Defibrillators,” Archives of Internal Medicine, Vol. 172, No. 14, July 23,
2012, pp. 1104–1105.
52
Reiffel JA and Dizon J, “Cardiology Patient Page—The Implantable Cardioverter-Defibrillator: Patient Perspective,” Circulation, Vol. 105, No. 9, March 5, 2002, pp. 1022–1024.
53
Sears SF, Woodrow L, Cutitta K, Ford J, Shea JB, and Cahill J, “A Patient’s Guide to Living Confidently
with Chronic Heart Failure,” Circulation, Vol. 127, No. 13, April 2, 2013, pp. e525–e528.
54
Stewart GC, Weintraub JR, Pratibhu PP, et al., “Patient Expectations from Implantable Defibrillators to
Prevent Death in Heart Failure,” Journal of Cardiac Failure, Vol. 16, No. 2, February 2010, pp. 106–113.
55
Matlock DD, “Development and Pilot of Three Patient Decision Aids for Implanted Defibrillators,” 2012.
As of March 15, 2014:
http://pfaawards.pcori.org/node/20/datavizwiz/detail/78089
82
Redirecting Innovation in U.S. Health Care: Case Studies
56
Dodson JA, Fried TR, Van Ness PH, Goldstein NE, and Lampert R, “Patient Preferences for Deactivation of Implantable Cardioverter-Defibrillators,” JAMA Internal Medicine, Vol. 173, No. 5, March 11, 2013,
pp. 377–379.
57 National
Heart, Lung, and Blood Institute, “What Is an Implantable Cardioverter Defibrillator?” revised
August 2009. As of March 25, 2014:
http://www.nhlbi.nih.gov/health//dci/images/scd_icd.jpg
Case Study 6
Prostate-Specific Antigen
Jonathan Bergman, M.D., M.P.H.
Assistant Professor of Urology and Family Medicine,
David Geffen School of Medicine, University of California, Los Angeles
Charles D. Scales, Jr., M.D.
Clinical Scholar, Assistant Professor of Surgery, Division of Urologic Surgery,
Duke University School of Medicine
The Technology
Prostate-specific antigen (PSA) is a protein secreted by the prostate gland. Low levels of PSA
are present in the serum of all men, but elevated serum PSA levels correlate with a higher likelihood of having prostate cancer.
Rationale
PSA was initially discovered in the course of forensic research on techniques to identify seminal fluid in cases of suspected rape. In the early 1980s, elevated serum levels of PSA were found
to correlate with prostate cancer, and PSA began to be used as a biomarker of prostate cancer.
In the late 1980s, the level of PSA was shown to be proportional to increasingly palpable prostate cancer on digital rectal examination.1 It was also shown to correlate with higher clinical
stages of prostate cancer.2 In 1994, the U.S. Food and Drug Administration (FDA) approved
PSA as a screening test for prostate cancer.
Although the PSA blood test is widely used today, it has drawbacks. PSA is sensitive
(meaning that it is almost always abnormal in prostate cancer), but it has very low specificity
(meaning that the test is often abnormal, even in men who do not have prostate cancer). Moreover, prostate cancer is ubiquitous, with a lifetime risk of 1 in 6.3 Cases of prostate cancer vary
widely in terms of growth rate and aggressiveness. In some men, the disease is so indolent that
it is incidentally identified after death from an unrelated cause.
Initially hailed as a highly useful cancer screening test, that value of PSA has recently
been reassessed as the costs and consequences of needless biopsies and treatment are tallied.
Nevertheless, 25 million PSA screening tests are ordered every year.4–11 Richard J. Albin, an
early PSA investigator, recently remarked, “I never dreamed that my discovery four decades
ago would lead to such a profit-driven public health disaster.”12
83
84
Redirecting Innovation in U.S. Health Care: Case Studies
Development
Development of the serum test for PSA can be traced to work conducted in the laboratory
of Professor T. Ming Chu at Roswell Park Memorial Institute (now Roswell Park Cancer
Institute) in the late 1970s. Financial support for this line of research was provided in part
by the National Cancer Institute as part of the National Prostate Cancer Project. In 1979,
investigators reported the discovery and purification of a protein isolated using rabbit-derived
antiserum, terming the substance prostate-specific antigen.13 Subsequently, Papsidero and colleagues, from the same research group, described experiments identifying PSA in the serum of
patients with metastatic prostate cancer.14 In 1981, the group followed with a report documenting the potential utility of using PSA to monitor treatment response in patients with prostate
cancer.15 These investigations and others resulted in a patent issued to Roswell Park and the
state of New York in 1984.
The technology was subsequently transferred to the biotechnology industry in order to
develop a commercially available serum test. In 1986, Hybritech received FDA approval for the
PSA Tandem-R assay for the monitoring of treatment and recurrence among patients already
diagnosed with prostate cancer. FDA approval of the PSA test for the purpose of screening
asymptomatic men for prostate cancer did not occur until 1994.
Early Adoption
PSA Testing First Used to Monitor Treatment
The first clinical application of the newly developed PSA test was for monitoring (not detection) of prostate cancer. The Roswell Park group contributed a number of studies that linked
changes of PSA levels with disease stage, treatment response, disease recurrence, and progression. In a landmark study, Stamey and colleagues reported the results of a study testing
2,200 serum samples from 699 patients, among whom 378 carried a diagnosis of prostate
cancer.16 They found that PSA levels were elevated in most men with previously undiagnosed
prostate cancer. In addition, PSA level correlated with cancer stage. Moreover, it dropped
to low or undetectable levels following surgical prostate removal. Finally, PSA was useful in
detecting residual as well as recurrent prostate cancer in men undergoing treatment for the disease. Based on these studies, the PSA test was applied in clinical practice.
FDA Approval for Screening Use Followed
Soon after the test’s adoption, other investigators focused on its potential use for detecting
prostate cancer in asymptomatic men. One of the leading proponents of PSA screening for
prostate cancer during this period was William J. Catalona. In a landmark 1991 paper partially
funded by Hybritech, a division of Eli Lilly & Company, Catalona’s group demonstrated that
PSA combined with digital rectal examination was a superior method of detecting otherwise
asymptomatic prostate cancer compared to examination alone.17
On the basis of these findings and those of other investigators, in 1994 the FDA approved
the PSA test for the purpose of screening asymptomatic men for prostate cancer. However,
even in these early days, concern existed in the medical community about the ethics, benefits,
Case Study 6: Prostate-Specific Antigen
85
and potential harms of widespread PSA screening.18–20 This presaged the current debate about
the utility of population-based screening.21–23
Subsequent Use
Even before FDA approval of PSA as a screening test, its use skyrocketed among cancer-free
men.24 Among men aged at least 65 years, PSA testing rose from less than 5 percent of white
men in 1989 to approximately 40 percent in 1994; in black men the rate rose from a similar
baseline to approximately 35 percent in 1994.24
By the early 1990s, the rapid growth and widespread use of PSA testing produced a sharp
spike in the prevalence of detected disease (Figure 6.1). This increase was not due to a rising
incidence of prostate cancer; rather, it reflected a sharp increase in the detection of previously
unknown and generally localized prostate cancers in otherwise asymptomatic men.
Professional Organizations Provide Strong Support
In the mid-1990s, momentum for the use of PSA as a prostate cancer screening test continued
to build. This was due, in large part, to the efforts of the American Urological Association and
the American Cancer Society, both of which recommended screening with PSA even before
FDA approval, and such organizations as the National Comprehensive Cancer Network and
National Cancer Institute, which recommended PSA screening after it was FDA-approved.
Stimulated by this support, legislation at the state and federal level enshrined receipt of an
annual PSA test as a legal right. For example, Section 4103 of the 1997 Balanced Budget Act
mandated federal coverage for prostate cancer screening tests (e.g., the PSA test). The resulting
Figure 6.1
Prostate Cancer Incidence Trends (1975–2006)
SOURCE: American Cancer Society, “Special Report on Prostate Cancer,” Cancer Facts and Figures 2010, 2010.
Used with permission.66
86
Redirecting Innovation in U.S. Health Care: Case Studies
national coverage determination from the Centers for Medicare and Medicaid Services took
effect on January 1, 2000.
Corporate and Nonprofit Lobbying Promotes Use
Industry involvement in promoting PSA testing drove uptake higher. Since 2000, the annual
frequency of PSA screening has nearly doubled.4 In 1992, even before the FDA had approved
PSA as a screening test, TAP Pharmaceuticals, which at the time was a manufacturer of a prostate cancer treatment, sponsored St. Louis Cardinals star Stan Musial, a three-time National
League MVP and first-ballot Hall of Fame Inductee, to promote “Prostate Cancer Awareness Week.” TAP and other pharmaceutical companies also contributed to such patient advocacy groups as “Us Too,” which lobbied the National Cancer Institute to promote PSA testing. Other celebrity advocates of PSA screening included Harry Belafonte, Arnold Palmer,
Joe Torre, and John Kerry. Gen. Colin Powell served as a spokesman for the Prostate Cancer
Education Council, which sponsors Prostate Cancer Awareness Week. In 2009, KimberlyClark started the “Depend Campaign to End Prostate Cancer,” which featured endorsements
from such athletes as Jim Kelly, Ozzie Smith, Mike Bossy, Rod Woodson, Len Dawson, and
Ken Griffey, Sr. It also sponsored “Men’s Health Month” to encourage prostate cancer screening. Depend, a product for adult incontinence, is one of the company’s leading sellers.
Cost and Health Impact
Prostate cancer is common, but the course of the disease is extremely variable. The lifetime risk
of developing prostate cancer is approximately 1 in 6; however, the lifetime risk of death from
prostate cancer is only 1 in 30.25 The reason for prostate cancer’s relatively low mortality rate
is that many cases involve low-level disease, or that the patient is so elderly at the time of diagnosis that he dies from an unrelated health problem, such as heart disease, before the cancer
spreads. An incidental finding of prostate cancer is made in up to 45 percent of men undergoing surgical removal of the bladder and prostate for bladder cancer.26–30 Since the introduction of PSA screening in the late 1980s, prostate cancer has been diagnosed at progressively
earlier stages and at less-advanced grades with increased frequency.31 Because low-risk prostate
cancer often follows an indolent clinical trajectory,32 the overall survival benefit of curative
management of prostate cancer as a whole is small and is limited to men with more-aggressive
disease.33–35
Although Watchful Waiting Is an Acceptable Option, Few Men Choose It
Men diagnosed with prostate cancer are typically offered one of three options: radical prostatectomy (surgery), radiation therapy (either via external beam or radioactive inserts), or active
surveillance (also called “watchful waiting”). Active surveillance entails periodic PSA testing,
digital rectal examinations, and prostate needle biopsies at predetermined intervals, although
the ideal interval is unknown.36–44
Despite the absence of demonstrated superiority of curative treatment over active
surveillance among men with low-risk prostate cancer, population-based studies suggest that
only one-tenth of men who are eligible for this option choose surveillance.45 In an observational
study of almost 2,000 men with clinically localized prostate cancer, 16 percent were deemed
Case Study 6: Prostate-Specific Antigen
87
candidates for active surveillance.46 Of the group of men who were good candidates, only
9 percent chose surveillance. The rest opted for aggressive treatment.
For Most Patients, PSA Screening Does More Harm Than Good
One reason aggressive treatment is chosen so often is that patients are terrified by the diagnosis
of cancer and want to be told they have been “cured.” Although this gives patients emotional
relief, the claim is overstated. Many of these men would have done well with active surveillance. Moreover, the risks and long-term side effects of treatment can be very consequential.
These statements are based on several prospective, randomized studies. The Scandinavian
Prostate Cancer Group Study Number 4 (SPCG-4) trial randomized 695 men diagnosed with
localized prostate cancer to radical prostatectomy or watchful waiting. A small but durable survival benefit was shown for men treated with radical prostatectomy. Not surprisingly, healthrelated quality of life (HRQOL) was lower in individuals diagnosed with prostate cancer than
in population-based, age-matched controls, regardless of the severity of disease. This suggests
that simply being diagnosed with prostate cancer, whether treated or not, evokes enough fear
and concern to be deleterious to men’s overall HRQOL.47
A second prospective, controlled trial from Europe randomized 182,000 men to PSA
screening or control. It also showed a small reduction in prostate cancer–specific death, but
a significant risk of overdiagnosis and treatment.48 The American Prostate, Lung, Colorectal,
and Ovarian (PLCO) Cancer Screening Trial found no evidence of a mortality benefit for
organized annual screening compared with opportunistic screening, but it did not compare the
health outcomes of men who were screened with those who were not screened.49
Most notably, a recent meta-analysis of data from 341,342 participants in several highquality studies examined the value of PSA screening for prostate cancer.50 It found no reduction in prostate cancer–specific mortality (risk ratio [RR] 1.00, 95% confidence interval [CI]
0.70–1.30)50 but an increased incidence of harms, mainly due to overdiagnosis and complications of treating what would have otherwise been indolent disease.
Although watchful waiting and active surveillance spare men the worst side effects of
aggressive treatment, it does not relieve them of the anxiety of living with cancer. As a result,
many live in fear that the disease may someday spread. It is not clear whether the subsequent
deterioration of HRQOL that is often seen in this group is due to progressive distress over the
diagnosis (a direct consequence of screening), the emergence of cancer-related disease effects,
or simply the progressive effects of aging.
The immediate health effects on those who opt for curative treatment are much more
dramatic. Many suffer significant sexual dysfunction, urinary symptoms, and other forms of
distress. In the only large, prospective, randomized, controlled trial published to date in which
the HRQOL of men receiving curative treatment was compared with those who opted for
watchful waiting, five-year outcomes revealed significantly more erectile dysfunction (80 percent versus 45 percent) and urinary leakage (49 percent versus 21 percent) in men undergoing
prostatectomy, but less urinary obstruction (28 percent versus 44 percent).51 Other measures,
including bowel function, anxiety, depression, well-being, and overall HRQOL, were similar
between the two groups after five years. By six to eight years after diagnosis, other psychological domains of HRQOL, such as anxiety and depression, grew worse in those who chose
watchful waiting, perhaps in anticipation of the recurring tests, exams, and biopsies required
to monitor the condition.52
88
Redirecting Innovation in U.S. Health Care: Case Studies
A study at the University of California, San Francisco, Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE), has enriched our understanding of temporal
changes in HRQOL of men treated for prostate cancer.54 Enrollees undergoing radical prostatectomy have immediate postoperative declines in disease-specific and general HRQOL, with
some recovery in all domains during the first year after treatment. Sexual function continues
to improve from the immediate postoperative nadir in the second year, but it rarely returns
to the baseline presurgical state. Compared with men treated with external beam radiation
therapy, primary androgen deprivation, or active surveillance, sexual function declined most
precipitously for those treated with radical prostatectomy, but the prostatectomy group enjoyed
the greatest degree of subsequent recovery as well. After all treatment options, recovery of
urinary and sexual function generally occurred within two years, with smaller gains in year
three.55 For men treated with radical prostatectomy, urinary function improved over the first
year and remained stable in year two. For men treated with radiation therapy, urinary function
remained stable, but urinary bother was worse.56
As a Result of This Evidence, PSA Screening Is No Longer Recommended by the United
States Preventive Services Task Force
In 2012, after reviewing these studies and others, the United States Preventive Services Task
Force (USPSTF) recommended against routine PSA screening for prostate cancer.59–60 The task
force based its position on the determination that more men are physically or psychologically
harmed by screening than the small number who potentially benefit.
Despite the USPSTF’s Recommendation, Medicare Is Legally Obliged to Pay for PSA
Screening
When the USPSTF released its recommendation against screening, the American Urological Association pushed back, as did its allies in Congress. Rep. Jon Runyan (R-N.J.) and Rep.
Joe Baca (D-Calif.), co-chairs of the Congressional Prostate Cancer Task Force, sent a letter
to the Secretary of Health and Human Services opposing the USPSTF’s recommendations.
Public reactions, including messages posted on social media, were generally critical of the
USPSTF recommendations and supported continued PSA cancer screening.61 Reacting to the
criticism, USPSTF co-chair Michael LeFevre qualified that if an individual being screened
understands the risks and benefits of PSA screening and makes a personal decision that the
small possibility of benefit outweighs the known risk of harm, PSA screening is acceptable.
In May 2013, the American College of Physicians60 issued two guidance statements on
screening for prostate cancer, based on a critical review of existing guidelines:
1. Clinicians should “inform men between the age of 50 and 69 years about the limited
potential benefits and substantial harms of screening for prostate cancer.” The guideline
further specifies that the decision to use PSA screening should be based on a discussion
of the benefits and harms of screening, a patient’s risk for prostate cancer and general
health, and patient preferences (“clinicians should not screen for prostate cancer using
the prostate-specific antigen test in patients who do not express a clear preference for
screening”).
2. Clinicians “should not screen for prostate cancer using the PSA test in average-risk men
under the age of 50 years, men over the age of 69 years, or men with a life expectancy
of less than 10 to 15 years.”
Case Study 6: Prostate-Specific Antigen
89
No organization recommends PSA screening for men older than 75.62 However, many
doctors ignore the guidelines, and screening rates remain high.62–63 A recent study showed
that 40 percent of men age 75 and older were screened; whether or not a man was screened
depended heavily on which primary care physician he saw.64
Despite clinical guidelines and the recommendations of the USPSTF, federal law will
continue to require reimbursement for PSA screening until the law is changed. The same is true
of similar state laws. In all likelihood, Medicare and other insurers will also continue to pay
for the large number of needless biopsies, surgical and nonsurgical extirpations, and the costly
side effects and complications they often cause.
Unintended Consequences
PSA screening for prostate cancer was envisioned as a low-cost, accurate way to detect a
common and potentially deadly form of cancer at a point when definitive treatment could be
curative. The rapid uptake and dissemination of the technology triggered a marked increase in
the diagnosis and treatment of prostate cancers, many of which would have remained clinically
occult or produced indolent disease that would have never have harmed the patient.
The surge in needless biopsies and aggressive treatments that followed undoubtedly cured
some cancers that would have otherwise progressed. According to a recent analysis from the
USPSTF, if 1,000 men are screened for prostate cancer, one man, at most, will avoid death
from prostate cancer (Figure 6.2).59 Among these same 1,000 screened men, the number who
experience harms from treatment is much larger: 30 to 40 will lose the ability to control urine
leakage or to achieve an erection; two will experience a significant cardiovascular event (i.e.,
a heart attack); one will have a potentially life-threatening blood clot.59 For every 3,000 men
screened for prostate cancer, one will die as a result of surgical treatment.59 The procedures
inflicted many cases of psychological and physical harm, including impotence and incontinence. The test also triggered a marked increase in health care spending on prostate cancer
testing, treatment, and follow-up. All of these outcomes were unintended consequences of a
well-intended test.
PSA cancer screening is costly, and its benefits are scant. The treatments it triggers are
even more costly; the diagnosis of cancer is frightening—so much so that the mere knowledge that an individual has the disease degrades his quality of life. The aggressive surgical and
radiological treatments used to “cure” cancer patients work in many cases, but they also produce long-term problems. In the years since PSA screening became commonplace, large-scale
European and U.S. survival studies have determined that large-scale screening and aggressive
treatment of prostate cancer may (or may not) slightly reduce cancer-specific mortality, but it
does so at the cost of significant negative effects on the physical and emotional quality of life
of large numbers of men. Many would have been spared the costs, discomfort, and complications of aggressive treatment, or the lingering anxiety of living with a diagnosis of cancer, had
they not been tested.
For all of these reasons, PSA testing offers an example of a seemingly useful and low-cost
test that has been found to produce little more than higher rates of anxiety, treatment-related
complications, and health care spending.
90
Redirecting Innovation in U.S. Health Care: Case Studies
Figure 6.2
Benefits and Harms of PSA Screening for Prostate Cancer
SOURCE: National Cancer Institute, “Infographic: Benefits and
Harms of PSA Screening for Prostate Cancer,” NCI Cancer Bulletin,
Special Issue: The Science Behind Cancer Screening, Vol. 9, No. 23,
November 27, 2012.65
Case Study 6: Prostate-Specific Antigen
91
References
1
Stamey TA, Yang N, Hay AR, McNeal JE, Freiha FS, and Redwine E, “Prostate-Specific Antigen as a
Serum Marker for Adenocarcinoma of the Prostate,” New England Journal of Medicine, Vol. 317, No. 15, 1987,
p. 909.
2
Stamey T, Kabalin J, McNeal J, et al., “Prostate Specific Antigen in the Diagnosis and Treatment of Adenocarcinoma of the Prostate, II. Radical Prostatectomy Treated Patients, Journal of Urology, Vol. 141, No. 5,
1989, p. 1076.
3
Sakr W, Grignon D, Haas G, Heilbrun L, Pontes JE, and Crissman J, “Age and Racial Distribution of
Prostatic Intraepithelial Neoplasia,” European Urology, Vol. 30, No. 2, 1996, pp. 138–144.
4
Wever EM, Draisma G, Heijnsdijk EA, et al., “Prostate-Specific Antigen Screening in the United States
Versus in the European Randomized Study of Screening for Prostate Cancer–Rotterdam,” Effects of Prostate
Cancer Screening and Treatment, Vol. 102, No. 5, 2012, p. 41.
5
Farwell WR, Linder JA, and Jha AK, “Trends in Prostate-Specific Antigen Testing from 1995 Through
2004,” Archives of Internal Medicine, Vol. 167, No. 22, 2007, p. 2497.
6
Oliver S, Gunnell D, and Donovan J, “Comparison of Trends in Prostate-Cancer Mortality in England and
Wales and the USA,” The Lancet, Vol. 355, No. 9217, 2000, pp. 1788–1789.
7
Jemal A, Bray F, Center MM, Ferlay J, Ward E, and Forman D, “Global Cancer Statistics,” CA: A Cancer
Journal for Clinicians, Vol. 61, No. 2, 2011, pp. 69–90.
8
Goodwin JS, Tan A, Jaramillo E, and Kuo Y-F, “Prostate-Specific Antigen Testing in Men Aged 40–64
Years: Impact of Publication of Clinical Trials,” Journal of the National Cancer Institute, Vol. 105, No. 10,
2013, pp. 743–745.
9
Schroeck FR, Kaufman SR, Jacobs BL, et al., “Technology Diffusion and Diagnostic Testing for Prostate
Cancer,” Journal of Urology, 2013.
10
Manvar AM, Pruthi RS, Wallen EM, and Nielsen ME, “Epidemiology of Prostate Cancer,” Prostate Cancer:
A Comprehensive Perspective, Springer, 2013, pp. 285–299.
11
Etzioni R, Tsodikov A, Mariotto A, et al., “Quantifying the Role of PSA Screening in the US Prostate
Cancer Mortality Decline,” Cancer Causes & Control, Vol. 19, No. 2, 2008, pp. 175–181.
12
Ablin RJ, “The Great Prostate Mistake,” New York Times, March 10, 2010. As of January 16, 2014:
http://www.nytimes.com/2010/03/10/opinion/10Ablin.html?emc=eta1
13
Wang MC, Valenzuela LA, Murphy GP, and Chu TM, “Purification of a Human Prostate Specific Antigen,” Investigative Urology, Vol. 17, No. 2, September 1979, pp. 159–163.
14
Papsidero LD, Wang MC, Valenzuela LA, Murphy GP, and Chu TM, “A Prostate Antigen in Sera of Prostatic Cancer Patients,” Cancer Research, Vol. 40, No. 7, July 1980, pp. 2428–2432.
15
Kuriyama M, Wang MC, Lee CI, et al., “Use of Human Prostate-Specific Antigen in Monitoring Prostate
Cancer,” Cancer Research, Vol. 41, No. 10, October 1981, pp. 3874–3876.
16
Stamey TA, Yang N, Hay AR, McNeal JE, Freiha FS, and Redwine E, “Prostate-Specific Antigen as a
Serum Marker for Adenocarcinoma of the Prostate,” New England Journal of Medicine, Vol. 317, No. 15, October 8, 1987, pp. 909–916.
17
Catalona WJ, Smith DS, Ratliff TL, et al., “Measurement of Prostate-Specific Antigen in Serum as a
Screening Test for Prostate Cancer,” New England Journal of Medicine, Vol. 324, No. 17, April 25, 1991,
pp. 1156–1161.
18
Adami HO, Baron JA, and Rothman KJ, “Ethics of a Prostate Cancer Screening Trial,” Lancet, Vol. 343,
No. 8903, April 16, 1994, pp. 958–960.
92
Redirecting Innovation in U.S. Health Care: Case Studies
19
Woolf SH, “Screening for Prostate Cancer with Prostate-Specific Antigen: An Examination of the Evidence,” New England Journal of Medicine, Vol. 333, No. 21, November 23, 1995, pp. 1401–1405.
20
Kramer BS, Brown ML, Prorok PC, Potosky AL, and Gohagan JK, “Prostate Cancer Screening: What
We Know and What We Need to Know,” Annals of Internal Medicine, Vol. 119, No. 9, November 1, 1993,
pp. 914–923.
21
Moyer VA, “Screening for Prostate Cancer: U.S. Preventive Services Task Force Recommendation Statement,” Annals of Internal Medicine, Vol. 157, No. 2, July 17, 2012, pp. 120–134.
22
Brawley OW, “Prostate Cancer Screening: What We Know, Don’t Know, and Believe,” Annals of Internal
Medicine, Vol. 157, No. 2, July 17, 2012, pp. 135–136.
23
Catalona WJ, D’Amico AV, Fitzgibbons WF, et al., “What the U.S. Preventive Services Task Force Missed
in Its Prostate Cancer Screening Recommendation,” Annals of Internal Medicine, Vol. 157, No. 2, July 17,
2012, pp. 137–138.
24
Legler JM, Feuer EJ, Potosky AL, Merrill RM, and Kramer BS, “The Role of Prostate-Specific Antigen
(PSA) Testing Patterns in the Recent Prostate Cancer Incidence Decline in the United States,” Cancer Causes
Control, Vol. 9, No. 5, October 1998, pp. 519–527.
25
Siegel R, Ward E, Brawley O, and Jemal A, “Cancer Statistics, 2011: The Impact of Eliminating Socioeconomic and Racial Disparities on Premature Cancer Deaths,” CA: A Cancer Journal for Clinicians, Vol. 61,
No. 4, July–August 2011, pp. 212–236.
26 Revelo
MP, Cookson MS, Chang SS, Shook MF, Smith JA, Jr., and Shappell SB, “Incidence and Location
of Prostate and Urothelial Carcinoma in Prostates from Cystoprostatectomies: Implications for Possible Apical
Sparing Surgery,” Journal of Urology, Vol. 171, No. 2, Part 1, February 2004, pp. 646–651.
27
Moutzouris G, Barbatis C, Plastiras D, et al., “Incidence and Histological Findings of Unsuspected Prostatic Adenocarcinoma in Radical Cystoprostatectomy for Transitional Cell Carcinoma of the Bladder,” Scandinavian Journal of Urology and Nephrology, Vol. 33, No. 1, February 1999, pp. 27–30.
28
Abbas F, Hochberg D, Civantos F, and Soloway M, “Incidental Prostatic Adenocarcinoma in Patients
Undergoing Radical Cystoprostatectomy for Bladder Cancer,” European Urology, Vol. 30, No. 3, 1996,
pp. 322–326.
29
Sakr WA, Grignon DJ, Crissman JD, et al., “High Grade Prostatic Intraepithelial Neoplasia (HGPIN)
and Prostatic Adenocarcinoma Between the Ages of 20-69: An Autopsy Study of 249 Cases,” In Vivo, Vol. 8,
No. 3, May–June 1994, pp. 439–443.
30
Guileyardo JM, Johnson WD, Welsh RA, Akazaki K, and Correa P, “Prevalence of Latent Prostate Carcinoma in Two U.S. Populations,” Journal of the National Cancer Institute, Vol. 65, No. 2, August 1980,
pp. 311–316.
31
Cooperberg MR, Lubeck DP, Meng MV, Mehta SS, and Carroll PR, “The Changing Face of Low-Risk
Prostate Cancer: Trends in Clinical Presentation and Primary Management,” Journal of Clinical Oncology,
Vol. 22, No. 11, June 1, 2004, pp. 2141–2149.
32
Albertsen PC, Hanley JA, Gleason DF, and Barry MJ, “Competing Risk Analysis of Men Aged 55 to 74
Years at Diagnosis Managed Conservatively for Clinically Localized Prostate Cancer,” Journal of the American
Medical Association, Vol. 280, No. 11, September 16, 1998, pp. 975–980.
33
Albertsen PC, Hanley JA, and Fine J, “20-Year Outcomes Following Conservative Management of Clinically Localized Prostate Cancer,” Journal of the American Medical Association, Vol. 293, No. 17, May 4, 2005,
pp. 2095–2101.
34 Chodak
GW, Thisted RA, Gerber GS, et al., “Results of Conservative Management of Clinically Localized
Prostate Cancer,” New England Journal of Medicine, Vol. 330, No. 4, January 27, 1994, pp. 242–248.
35
Schroder FH, “Stratifying Risk—The U.S. Preventive Services Task Force and Prostate-Cancer Screening,”
New England Journal of Medicine, Vol. 365, No. 21, November 24, 2011, pp. 1953–1955.
Case Study 6: Prostate-Specific Antigen
93
36 Heidenreich A, Aus G, Bolla M, et al., “EAU Guidelines on Prostate Cancer,” European Urology, Vol. 53,
No. 1, January 2008, pp. 68–80.
37
Carter HB, Kettermann A, Warlick C, et al., “Expectant Management of Prostate Cancer with Curative
Intent: An Update of the Johns Hopkins Experience,” Journal of Urology, Vol. 178, No. 6, December 2007,
pp. 2359–2364, discussion pp. 2364–2355.
38
Klotz L, “Active Surveillance for Prostate Cancer: For Whom?” Journal of Clinical Oncology, Vol. 23,
No. 32, November 10, 2005, pp. 8165–8169.
39
van As NJ, Norman AR, Thomas K, et al., “Predicting the Probability of Deferred Radical Treatment
for Localised Prostate Cancer Managed by Active Surveillance,” European Urology, Vol. 54, No. 6, December 2008, pp. 1297–1305.
40 Dall’Era
MA, Konety BR, Cowan JE, et al., “Active Surveillance for the Management of Prostate Cancer in
a Contemporary Cohort,” Cancer, Vol. 112, No. 12, June 15, 2008, pp. 2664–2670.
41
Warlick C, Trock BJ, Landis P, Epstein JI, and Carter HB, “Delayed Versus Immediate Surgical Intervention and Prostate Cancer Outcome,” Journal of the National Cancer Institute, Vol. 98, No. 5, March 1, 2006,
pp. 355–357.
42 Patel
MI, DeConcini DT, Lopez-Corona E, Ohori M, Wheeler T, and Scardino PT, “An Analysis of Men
with Clinically Localized Prostate Cancer Who Deferred Definitive Therapy,” Journal of Urology, Vol. 171,
No. 4, April 2004, pp. 1520–1524.
43
Soloway MS, Soloway CT, Williams S, Ayyathurai R, Kava B, and Manoharan M, “Active Surveillance; A
Reasonable Management Alternative for Patients with Prostate Cancer: The Miami Experience,” BJU International, Vol. 101, No. 2, January 2008, pp. 165–169.
44 Hardie
C, Parker C, Norman A, et al., “Early Outcomes of Active Surveillance for Localized Prostate
Cancer,” BJU International, Vol. 95, No. 7, May 2005, pp. 956–960.
45
Harlan SR, Cooperberg MR, Elkin EP, et al., “Time Trends and Characteristics of Men Choosing Watchful Waiting for Initial Treatment of Localized Prostate Cancer: Results from CaPSURE,” Journal of Urology,
Vol. 170, No. 5, November 2003, pp. 1804–1807.
46 Barocas
DA, Cowan JE, Smith JA, Jr., and Carroll PR, “What Percentage of Patients with Newly Diagnosed Carcinoma of the Prostate Are Candidates for Surveillance? An Analysis of the CaPSURE Database,”
Journal of Urology, Vol. 180, No. 4, October 2008, pp. 1330–1334, discussion pp. 1334–1335.
47
Johansson E, Steineck G, Holmberg L, et al., “Long-Term Quality-of-Life Outcomes After Radical Prostatectomy or Watchful Waiting: The Scandinavian Prostate Cancer Group-4 Randomised Trial,” Lancet Oncology, Vol. 12, No. 9, 2011, pp. 891–899.
48
Schröder FH, Hugosson J, Roobol MJ, et al., “Screening and Prostate-Cancer Mortality in a Randomized
European Study,” New England Journal of Medicine, Vol. 360, No. 13, 2009, pp. 1320–1328.
49
Andriole GL, Crawford ED, Grubb RL, et al., “Prostate Cancer Screening in the Randomized Prostate,
Lung, Colorectal, and Ovarian Cancer Screening Trial: Mortality Results After 13 Years of Follow-Up,” Journal of the National Cancer Institute, Vol. 104, No. 2, 2012, pp. 125–132.
50
Ilic D, Neuberger MM, Djulbegovic M, and Dahm P, “Screening for Prostate Cancer,” Cochrane Database
of Systematic Reviews, Vol. 1, 2013, CD004720.
51
Steineck G, Helgesen F, Adolfsson J, et al., “Quality of Life After Radical Prostatectomy or Watchful Waiting,” New England Journal of Medicine, Vol. 347, No. 11, September 12, 2002, pp. 790–796.
52
Johansson E, Bill-Axelson A, Holmberg L, Onelov E, Johansson JE, and Steineck G, “Time, Symptom
Burden, Androgen Deprivation, and Self-Assessed Quality of Life After Radical Prostatectomy or Watchful
Waiting: The Randomized Scandinavian Prostate Cancer Group Study Number 4 (SPCG-4) Clinical Trial,”
European Urology, Vol. 55, No. 2, February 2009, pp. 422–430.
94
Redirecting Innovation in U.S. Health Care: Case Studies
53
Johansson E, Steineck G, Holmberg L, et al., “Long-Term Quality-of-Life Outcomes After Radical Prostatectomy or Watchful Waiting: The Scandinavian Prostate Cancer Group-4 Randomised Trial,” Lancet Oncology, Vol. 12, No. 9, September 2011, pp. 891–899.
54
Porten SP, Cooperberg MR, Konety BR, and Carroll PR, “The Example of CaPSURE: Lessons Learned
from a National Disease Registry,” World Journal of Urology, Vol. 29, No. 3, June 2011, pp. 265–271.
55
Huang GJ, Sadetsky N, and Penson DF, “Health Related Quality of Life for Men Treated for Localized
Prostate Cancer with Long-Term Followup,” Journal of Urology, Vol. 183, No. 6, June 2010, pp. 2206–2212.
56
Litwin MS, Pasta DJ, Yu J, Stoddard ML, and Flanders SC, “Urinary Function and Bother After Radical
Prostatectomy or Radiation for Prostate Cancer: A Longitudinal, Multivariate Quality of Life Analysis from
the Cancer of the Prostate Strategic Urologic Research Endeavor,” Journal of Urology, Vol. 164, No. 6, December 2000, pp. 1973–1977.
57
Wu AK, Cooperberg MR, Sadetsky N, and Carroll PR, “Health Related Quality of Life in Patients
Treated with Multimodal Therapy for Prostate Cancer,” Journal of Urology, Vol. 180, No. 6, December 2008,
pp. 2415–2422, discussion p. 2422.
58
Litwin MS, Flanders SC, Pasta DJ, Stoddard ML, Lubeck DP, and Henning JM, “Sexual Function and
Bother After Radical Prostatectomy or Radiation for Prostate Cancer: Multivariate Quality-of-Life Analysis
from CaPSURE: Cancer of the Prostate Strategic Urologic Research Endeavor,” Urology, Vol. 54, No. 3, September 1999, pp. 503–508.
59
Moyer VA, “Screening for Prostate Cancer: US Preventive Services Task Force Recommendation Statement,” Annals of Internal Medicine, Vol. 157, No. 2, 2012, pp. 120–134.
60 Qaseem
A, Barry MJ, Denberg TD, Owens DK, and Shekelle P, “Screening for Prostate Cancer: A Guidance Statement from the Clinical Guidelines Committee of the American College of Physicians,” Annals of
Internal Medicine, Vol. 158, No. 10, 2013, pp. 761–769.
61
Prabhu V, “Using Social Media to Gauge Reaction to the USPSTF Report on Prostate Cancer Screening: Twitter as an Investigative Tool,” paper presented at the 34th Annual Meeting of the Society for Medical
Decision Making, 2012.
62
Walter LC, Bertenthal D, Lindquist K, and Konety BR, “PSA Screening Among Elderly Men with Limited
Life Expectancies,” Journal of the American Medical Association, Vol. 296, No. 19, 2006, pp. 2336–2342.
63
Goodwin JS, Tan A, Jaramillo E, and Kuo YF, “Prostate-Specific Antigen Testing in Men Aged 40–64
Years: Impact of Publication of Clinical Trials,” Journal of the National Cancer Institute, Vol. 105, No. 10,
2013, pp. 743–745.
64 Jaramillo
E, Tan A, Yang L, Kuo YF, and Goodwin JS, “Variation Among Primary Care Physicians in
Prostate-Specific Antigen Screening of Older Men,” Journal of the American Medical Association, Vol. 310,
No. 15, 2013, pp. 1622–1624.
65. National
Cancer Institute, “Infographic: Benefits and Harms of PSA Screening for Prostate Cancer,” NCI
Cancer Bulletin, Special Issue: The Science Behind Cancer Screening, Vol. 9, No. 23, November 27, 2012. As of
November 23, 2013:
http://www.cancer.gov/ncicancerbulletin/112712/page12
66 American
Cancer Society, “Special Report on Prostate Cancer,” Cancer Facts and Figures 2010, 2010. Used
with permission. As of March 25, 2014:
http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/documents/document/acspc-026238.pdf
Case Study 7
Robotic Surgery
Charles D. Scales, Jr., M.D.
Clinical Scholar, Robert Wood Johnson Foundation Clinical Scholars,
David Geffen School of Medicine, University of California, Los Angeles
Jonathan Bergman, M.D., M.P.H.
Assistant Professor of Urology and Family Medicine (Joint),
David Geffen School of Medicine, University of California, Los Angeles
The Technology
Robotically assisted laparoscopic surgery differs from traditional (i.e., manual) laparoscopic
surgery through the coupling of the surgeon’s movements to those of a remote-controlled
robot. This eliminates tremor. Robotically assisted surgery was pioneered in the 1980s and
spread widely over the next two decades into surgical procedures in cardiology, cardiothoracic
surgery, general surgery, urology, gynecology, pediatrics, vascular surgery, neurosurgery, and
orthopedics.1–3
Rationale
The development of minimally invasive surgery via laparoscopy produced several advantages
over open surgery: better visualization (for certain procedures), shorter hospital stays, less postoperative pain, improved cosmesis, lower rates of blood transfusion, and a decreased risk of
perioperative infection.4–7 One technical rationale for introduction of the robot was to facilitate
challenging surgical tasks, such as suturing with laparoscopic instruments.
Genesis
Initially, robots were utilized to carry out a single task: The PUMA 560 was used to guide a
needle for brain biopsy in 1985, and the PROBOT, representing the next generation, was first
used in 1988.8–11 In the late 1990s, successful operations for coronary artery bypass grafting
and hybrid revascularization were performed in Europe and the United States.12–14 The da Vinci
Surgical System first received FDA approval for general intra-abdominal surgery in 2000. The
first robotically assisted laparoscopic radical prostatectomies were performed that same year;
first in Germany, then France, then the United States.15, 16 A successful cholecystectomy was
performed in 2001. Over the rest of the decade, a host of major procedures followed.17–20 The
first robotically assisted cadaveric kidney transplant was performed in 2009.21
95
96
Redirecting Innovation in U.S. Health Care: Case Studies
The development work leading to the da Vinci Surgical System was performed at
SRI International, a nonprofit research institute. The work was initially funded by the National
Institutes of Health; subsequently, the Defense Advanced Research Projects Agency (DARPA)
supported work on the technology.22 DARPA’s interest centered on the opportunity to use
telesurgery and robotics for surgical interventions on the battlefield. The founders of Intuitive Surgical licensed SRI’s technology and went on to further develop prototypes, leading to
today’s da Vinci robot.23
The Da Vinci Surgical System Dominates the Current Market
The da Vinci Surgical System was not the first approved in the marketplace, but it grew rapidly by buying competitor products. The PUMA 560, PROBOT, and ROBODOC systems
predated the development of the da Vinci, and the Computer Motion with AESOP and ZEUS
systems rose at roughly the same time as da Vinci. Intuitive Surgical, which makes the da
Vinci, came to dominate the market by buying Computer Motion in 2003 and aggressively
marketing the da Vinci product thereafter. The merger had the added value of resolving patent
infringement lawsuits between the two companies. The deal was financed by equity; Intuitive
Surgical issued shares, and final ownership of the merged entity was divided with 68 percent
going to Intuitive Surgical shareholders (68-percent equity) and the remainder going to Computer Motion shareholders (32 percent).24 The deal worked; today, Intuitive Surgical has a near
monopoly in the field of robotic surgery.25–27
Early Adoption
Robotic Surgery Expanded Rapidly
Over the past five to seven years, performance of robotic procedures, and the diffusion of
robotic technology, has expanded exponentially in both Europe and in the United States. In
2007, approximately 80,000 robotically assisted procedures were performed globally. Three
years later, more than 200,000 procedures were performed.28–30 Intuitive Surgical’s da Vinci
robot is the dominant technology in use today. According to the company, its worldwide
installed base as of September 30, 2012, encompasses 2,462 robotic systems. This includes
1,789 systems in the United States, 400 in Europe, and 273 in the rest of the world.31–33
Robotic Assistance Increases the Cost of a Typical Surgical Procedure
The proliferation of this technology has come at substantial cost. The price of a single robotic
unit ranges from $1.2 to $2.3 million. Single-use consumables cost $1,300–$2,200 per procedure.31, 34 Longer operative times involved in using a surgical robot also increase the associated
costs per procedure for operating room staff, including an anesthesiologist or nurse anesthetist, scrub nurses, circulating nurses, and technicians.35, 36 Theoretically, these expenses may be
offset to some degree by reduced convalescence, earlier return to work, reductions in hospital
length of stay, and minimized analgesic utilization, but the financial benefits of these offsets
have not been documented.37, 38 Overall, robotic assistance increases the cost of a typical surgery by approximately 13 percent (down to about 6 percent if the amortized cost of the robot
is excluded).28
Interestingly, hospitals do not directly receive increased reimbursement to offset the
added cost for robotic surgery versus traditional laparoscopic procedures. Intuitive Surgical
Case Study 7: Robotic Surgery
97
notes this issue in regulatory filings: “A surgical procedure, completed with or without robotic
assistance, continues to be assigned to the same MS-DRG . . . both hospitals and physicians
receive the same reimbursement amount for their respective services regardless of the actual
costs incurred . . . and [reimbursement] is unrelated to the specific products used in that procedure.”39 In addition, observers have noted that neither Medicare nor many private insurers
pay additional fees for robot-assisted interventions.40 However, hospitals can increase charges
for procedures or the diagnoses associated with the robotic system; whether the fees are paid
or not is considered proprietary information by hospitals and health plans alike. Also, because
Medicare and private insurer reimbursement rates are set based on hospital charges, hospitals
can, over time, increase their revenue indirectly.40 The most tangible benefit of acquiring a
robotic system, given the fact that hospitals function in a high fixed-cost environment, is the
additional surgeons—and surgical volume—that they attract.41, 42
Economics aside, hospitals in competitive markets often acquire cutting-edge technology
to differentiate themselves, attract surgeons and surgical case loads, and signal their advanced
competence.40, 42, 43 At least one hospital executive has noted that competitive pressure played
in his decision to acquire a robotic surgical system.44
The Impact of Adoption Is Most Dramatic in Prostatectomy Procedures
The transition from open surgery to laparoscopy to robotic-assisted laparoscopic procedures
has been most striking for surgical removal of the prostate. Following the first robotically
assisted radical prostatectomy in 2000, uptake was slow until about 2005, when adaptation
accelerated at a rapid rate.45, 46 Although the technology was introduced into clinical practice
only 12 years ago (FDA regulatory clearance for use in prostatectomy procedures was granted
in May 2001)39 and did not begin to widely proliferate before 2005, it has transformed how
radical prostatectomies are performed. Today, approximately 80 percent of all radical prostatectomies in the United States are performed with robotic assistance.47
As robotic capabilities increased, not only did the percentage of surgical candidates
undergoing prostatectomy by robotic procedure increase; so did the overall number of surgical
prostatectomies. Since 2005, the number of prostatectomies among both older (>65 years old)
and younger (<65 years old) U.S. men has increased by 60 percent.45, 48, 49
Growing Reliance on Surgical Treatment of Prostate Cancer Is Linked to the Growing Use
of Robotic Technology
Today, far more men diagnosed with prostate cancer opt to undergo prostatectomy, compared
with a few years ago. Although some may view this development as a sign of progress, it is not
necessarily positive, since there is ample evidence that other approaches, including radiation
therapy and active surveillance (“watchful waiting”), may yield equivalent outcomes at lower
cost (and, in the case of active surveillance, fewer side effects).50–52
The striking increase in operative treatment is not explained by changes in the underlying
incidence or severity of disease. Instead, it appears to be related to the growing availability of
robotic surgical technology. Studies of hospital and local health care market practices strongly
suggest that the recent increase in surgical treatment of prostate cancer is both geographically and temporally related to acquisition of robotic surgical systems (see “Subsequent Use,”
below).53–55
More-aggressive treatment is not necessarily better for patients. Recently, an analysis
of linked Surveillance, Epidemiology, and End Results—Medicare (SEER-Medicare) claims
98
Redirecting Innovation in U.S. Health Care: Case Studies
demonstrated that use of advanced technologies, such as robots, has increased dramatically,
even among those men who are not good candidates for surgery, such as those with low-risk
prostate cancer, and men at high risk of noncancer mortality.56 In these groups, prostate surgery is generally regarded as equivocal (i.e., low-risk prostate cancer) or inappropriate (i.e., men
at high risk of surgical complications or noncancer mortality). In these groups, current guidelines advise against screening for and treatment of prostate cancer).57, 58
Subsequent Use
Acquisition of Robots Has Boosted Radical Prostatectomy Rates in Hospitals That Get Them
Data from the Healthcare Cost and Utilization Project (a family of databases comprising multistate, inpatient, and outpatient discharge records on insured and uninsured patients) suggest
that diffusion of the technology produced dramatic changes in hospital markets. Early in the
robotic diffusion period (2001–2005), hospitals that acquired a surgical robot increased their
prostatectomy case volume by an average of 30.1 cases annually (95% confidence interval [CI]
23.37, 36.4), compared with a change in volume of –4.9 cases annually (–7.2, –2.7) in hospitals
not acquiring a robot during the same time period.53 At the level of hospital referral regions
(HRRs), radical prostatectomy rates increased in regions that acquired robots, compared to
HRRs where robotic system were not acquired (Figure 7.1). For example, an HRR gaining
Figure 7.1
Mean Change in HRR Number of Radical Prostatectomy Procedures Following Robot Acquisition,
Controlling for Baseline HRR Volume and Herfindahl Index
SOURCE: Makarov DV, Yu JB, Desai RA, Penson DF, and Gross CP, “The Association Between Diffusion of the
Surgical Robot and Radical Prostatectomy Rates,” Medical Care, Vol. 49, No. 4, April 2011, pp. 333–339. 53
Case Study 7: Robotic Surgery
99
three robots between 2001 and 2005 saw an estimated annual increase of 106.6 radical prostatectomy cases per year, controlling for baseline HRR volume and Herfindahl index (a measure
of market concentration used by the U.S. Department of Justice when considering horizontal
mergers).59
Hospitals with Robots Attract Cases
Early on, it is likely that most of the growth in case volume of HRRs with robots was drawn
from HRRs where no robot was installed. The latter saw a decrease in case volume by an estimated 41.2 cases annually.53 Therefore, during the early period of robotic technology adoption,
the overall rate of surgical intervention for radical prostatectomy remained relatively stable.
The primary effect of robots appeared to be attracting surgical cases to the hospitals that
acquired them.
After 2005, Robots Triggered Market Consolidation and Intensification
Analyses conducted in the post-2005 era, when use of robotic surgery surged, paint a different
picture. At this phase of diffusion, the effects of the technology on health care practice—and,
therefore, spending—grew more profound. For example, a population-based analysis of radical
prostatectomy procedures in New York, New Jersey, and Pennsylvania demonstrated that the
total number of radical prostatectomy procedures grew substantially, from 8,115 in 2000 to
10,241 in 2009, a 26-percent increase.54 Most of the increase was noted at high-volume hospitals that acquired a robotic surgical system. Interestingly, while the total number of cases grew
by over 2,000 procedures annually, the number of hospitals performing this surgical procedure
fell by 37 percent, from 390 hospitals in 2000 to 244 in 2009 (Figure 7.2). By 2009, slightly
more than 35 percent of hospitals—those with a surgical robot—performed 85 percent of all
radical prostatectomies, indicating that robots were triggering both market consolidation and
intensification.54
Robots Also Influenced Surgical Training and Practice
Based on case log data for initial board certification and recertification, a growing number of
urologists stopped doing open procedures and only performed robot-assisted radical prostatectomies. In 2004, 13 out of 427 urologists (3%) performed only robotically assisted radical prostatectomies; by 2010, there was a sixfold increase (95 out of 500 [19%]).55 Similarly,
among urologists undergoing initial certification or recertification, the proportion performing
both open and robotic procedures increased dramatically, from 18 out of 427 (4%) in 2004 to
153 out of 500 in 2010 (31%). By 2010, half of practicing urologists used the robotic technique
in either all or some of their cases; the other half continued to perform only open prostatectomies. The increased reliance on robots has been noted among urologists undergoing initial
board certification, as well as those undergoing recertification.55
Robot-Assisted Prostatectomy Affects Diagnostic Procedures, Disparities, and
Inappropriate Use
Other effects on patterns of care—some concerning—were noted during this transition. For
example, after controlling for preoperative cancer risk group, an analysis of SEER data found
that men undergoing robot-assisted prostatectomy were less likely to undergo pelvic lymph
node dissection.60 Node sampling is important because it provides prognostic information in
men with intermediate or high-risk cancer. Other investigators have noted economic dispari-
100
Redirecting Innovation in U.S. Health Care: Case Studies
Figure 7.2
Trends in Radical Prostatectomy Procedure Volume and Number of Hospitals
Performing the Procedure in New York, New Jersey, and Pennsylvania54
SOURCE: Stitzenberg KB, Wong YN, Nielsen ME, Egleston BL, and Uzzo RG, “Trends in Radical
Prostatectomy: Centralization, Robotics, and Access to Urologic Cancer Care,” Cancer, Vol. 118,
No. 1, January 1, 2012, pp. 54–62. 54
Case Study 7: Robotic Surgery
101
ties in the location of hospitals acquiring robotic systems. Hospitals situated in areas associated with lower-income and higher-minority populations are less likely to acquire robotic systems.61 Thus, while these hospitals may be offering comparable quality of care, they may be
losing revenue-generating urological cases to hospitals that can afford to purchase a surgical
robot. Consistent with the market consolidation hypothesis, the distance traveled by patients
undergoing robot-assisted radical prostatectomy has recently increased.54 Finally, a more recent
analysis of SEER cancer registry data linked to Medicare claims suggests that use of robotic
technology is growing rapidly among patient groups that are least likely to benefit from the
technology and may even be harmed: men with low-risk prostate cancer and those at high risk
of noncancer mortality.56
Impact on Health and Costs
Despite the quick uptake and rapid proliferation of surgical robots, there is no evidence, based
on properly designed randomized controlled trials, that robotically assisted radical prostatectomies produce better outcomes or reduced side effects relative to manual (open) radical prostatectomy. Several prospective, nonrandomized trials have been performed; generally, these studies
have found that patients who undergo robotically assisted surgery have a somewhat more-rapid
convalescence, shorter hospital stays, and less blood loss than conventional cases.62–68 Other
studies, which produced less-reliable evidence because they used retrospectively compared
case series69, 70 or historical series as controls, have reported similar findings.71, 72 Recently, the
Health Technology Assessment Program of the Washington State Healthcare Authority published an assessment of the robotic technology. After analyzing the available data, the program
concluded that the level of evidence supporting use of the technology was moderate and, therefore, that the service would be covered. However, it did so with the proviso that no additional
payments be made beyond what the state would pay for the traditional procedure.73
The most robust comparison of outcomes between robotic and open radical prostatectomy performed to date was a SEER-Medicare analysis that used propensity score adjustment
before comparing the outcomes of the group receiving robotic surgery versus those treated with
open (traditional surgery) extirpation. As was noted by earlier assessments, use of the robotic
approach was associated with shorter lengths of stay, lower rates of blood transfusion, and
lower postoperative respiratory complications.74 However, the robotic approach was also associated with more-frequent surgical complications, including urinary incontinence, genitourinary
complications, and erectile dysfunction. Given adequate knowledge of these trade-offs, many
men might not opt for robotic surgery or might select a nonsurgical option, such as watchful
waiting, radiation therapy, or implants.
Hospitals Cite Robotic Surgery in Marketing
Despite the lack of evidence that robotic surgery is superior to open surgical approaches, hospitals frequently tout the advantages of robotic surgery. Hospital websites often use images and
text taken directly from the manufacturers of robotic systems and promise improved perioperative outcomes, decreased complications, reduced operative time, shorter recovery, less scarring, reduced risk for infection, and less pain.75 Some of these advertisements use high-tech
or emotional language, such as “state-of-the-art,” “cutting edge,” “you owe it to yourself,” and
“loved one.”75–77 Marketing claims include higher lymph node yield, improved cancer out-
102
Redirecting Innovation in U.S. Health Care: Case Studies
comes, and “ideal treatment.”77, 78 None of these claims is substantiated by rigorous evidence;
some, such as node sampling, are contradicted by the available evidence.60
The training given to surgeons who are beginning to use a robot has not been studied
intensively, but it appears to be highly variable. Recently, Intuitive Surgical was the defendant
in a case in Washington State (Estate of Fred E. Taylor v. Intuitive Surgical Inc., 09-2-03136-5).
The plaintiffs focused on Intuitive’s recommendation that a one-day session at the company
headquarters, followed by two supervised robotic surgeries, is sufficient to prepare surgeons to
perform robotic procedures. The jury found for Intuitive Surgical, so no claims were paid.79
An important question in evaluating any new technology is whether or not it produces
better outcomes than the technologies it is replacing. Therefore, it is reasonable to ask whether
the evidence compiled to date indicates that robotic surgery is superior to open surgery for
treating cancer of the prostate. Unfortunately, no rigorously randomized controlled trials comparing open and robot-assisted radical prostatectomy have been performed. Most studies that
have examined the effectiveness of robotic surgery employ observational designs that are prone
to bias (i.e., patients are selected to one group or the other based on characteristics that may
independently influence their outcome). Therefore, these studies are generally considered to
represent a lower level of evidence.
A Medical “Arms Race”
Adoption and diffusion of robotic surgical technology has been rapid and extensive. Although
there is scant evidence that the technology improves short-term surgical outcomes—and some
evidence that it may create problems as well—the technology has transformed the landscape
of prostate cancer treatment in the United States. When early adopters of robotic surgical
systems began advertising that they had the technology and started drawing cases from nonadopting hospitals, it set off a “medical arms race” between competing hospitals and health
care systems.43 Soon, hospitals scrambled to acquire the technology in order to retain or attract
urologic surgeons and the cases they perform. Although third-party payments for the technique are generally similar to those associated with conventional open surgery, hospitals that
adopted the technology have dramatically boosted their surgical caseloads relative to hospitals
that did not. Thus, the technology has driven health care spending higher without making an
appreciable impact on the health or long-term prognosis of patients treated with robotically
assisted surgery.
Although there is no evidence that robotically assisted surgery is superior to alternative
approaches, including conventional open surgery, radiation therapy, and watchful waiting,
rates of surgical prostatectomy are up, and the large majority of prostate removal surgeries use
robotic technology. More follow-up data are needed to fully understand how this technology
affects the treatment and long-term outcomes of men with prostate cancer.
Robotic surgery is only one example of the medical arms race. Other costly technologies,
such as the even costlier proton beam therapy, have also emerged in recent years. Hospitals
that acquire these extraordinarily expensive machines (more than $180 million each) hope that
the allure of the technology and the prestige it confers will attract patients to their facility.83
Similar to claims made in support of robotic surgery, proponents of proton beam therapy claim
that the technology offers greater precision, improved efficacy, less damage to healthy tissue,
and fewer side effects than older or less-costly techniques.80 However, there is no evidence that
Case Study 7: Robotic Surgery
103
proton beam therapy offers victims of prostate cancer (or many other cancers) better outcomes
than that provided by less expensive conventional (photon) radiation therapy.84
Medical “arms race” technologies such as these have done little or nothing to improve the
health of men with prostate cancer. They have, however, increased the intensity of treatment
offered to middle-aged and elderly men with this common, frequently indolent, and, by many
measures, already overtreated condition. And because there is substantial evidence that aggressive treatment may not only be costly but also harmful, it may offer little value to those whom
these technologies were designed to help.81, 82
References
1
Glassman E, Hanson WA, Kazanzides P, et al., “Image-Directed Robotic System for Precise Robotic Surgery Including Redundant Consistency Checking,” U.S. Patent US5086401 A, February 4, 1992.
2
Camarillo DB, Krummel TM, and Salisbury JK Jr, “Robotic Technology in Surgery: Past, Present, and
Future,” American Journal of Surgery, Vol. 188, 2004, p. 2S.
3 Mack MJ, “Minimally Invasive and Robotic Surgery,” Journal of the American Medical Association, Vol. 285,
2001, pp. 568–572.
4
Garrard C, Clements R, Nanney L, Davidson J, and Richards W, “Adhesion Formation Is Reduced After
Laparoscopic Surgery,” Surgical Endoscopy, Vol. 13, 1999, pp. 10–13.
5
Young-Fadok T, Long KH, McConnell E, Rey GG, and Cabanela R, “Advantages of Laparoscopic Resection for Ileocolic Crohn’s Disease,” Surgical Endoscopy, Vol. 15, 2001, pp. 450–454.
6
Spaw AT, Reddick EJ, and Olsen DO, “Laparoscopic Laser Cholecystectomy: Analysis of 500 Procedures,”
Surgical Laparoscopy & Endoscopy, Vol. 1, 1991, p. 2.
7
Milsom JW, Böhm B, Hammerhofer KA, Fazio V, Steiger E, and Elson P, “A Prospective, Randomized
Trial Comparing Laparoscopic Versus Conventional Techniques in Colorectal Cancer Surgery: A Preliminary
Report,” Journal of the American College of Surgeons, Vol. 187, 1998, pp. 46–54.
8
Armstrong B, Khatib O, and Burdick J, “The Explicit Dynamic Model and Inertial Parameters of the
PUMA 560 Arm,” Proceedings of the 1986 IEEE International Conference on Robotics and Automation Proceedings, 1986, pp. 510–518.
9
Corke PI and Armstrong-Helouvry B, “A Search for Consensus Among Model Parameters Reported for the
PUMA 560 Robot,” 1994 Proceedings, 1994 IEEE International Conference on Robotics and Automation, 1994,
pp. 1608–1613.
10
Harris S, Arambula-Cosio F, Mei Q, et al., “The Probot—An Active Robot for Prostate Resection,” Proceedings of the Institution of Mechanical Engineers, “Part H: Journal of Engineering in Medicine,” Vol. 211,
1997, pp. 317–325.
11
Mei Q, Harris S, Arambula-Cosio F, et al., “PROBOT—A Computer Integrated Prostatectomy System,”
Visualization in Biomedical Computing, Springer, 1996, pp. 581–590.
12
Garcia-Ruiz A, Smedira NG, Loop FD, et al., “Robotic Surgical Instruments for Dexterity Enhancement
in Thoracoscopic Coronary Artery Bypass Graft,” Journal of Laparoendoscopic & Advanced Surgical Techniques,
Vol. 7, 1997, pp. 277–283.
13
Reichenspurner H, Boehm DH, Welz A, et al., “Minimally Invasive Coronary Artery Bypass Grafting:
Port-Access Approach Versus Off-Pump Techniques,” Annals of Thoracic Surgery, Vol. 66, 1998, p. 1036.
14
Shennib H, Bastawisy A, McLoughlin J, and Moll F, “Robotic Computer-Assisted Telemanipulation Enhances Coronary Artery Bypass,” Journal of Thoracic and Cardiovascular Surgery, Vol. 117, 1999,
pp. 310–313.
104
Redirecting Innovation in U.S. Health Care: Case Studies
15
Bentas W, Wolfram M, Jones J, Brautigam R, Kramer W, and Binder J, “Robotic Technology and the
Translation of Open Radical Prostatectomy to Laparoscopy: The Early Frankfurt Experience with Robotic
Radical Prostatectomy and One Year Follow-Up,” European Urology, Vol. 44, 2003, pp. 175–181.
16
Binder J and Kramer W, “Robotically-Assisted Laparoscopic Radical Prostatectomy,” BJU International,
Vol. 87, 2001, pp. 408–410.
17
Vidovszky TJ, Smith W, Ghosh J, and Ali MR, “Robotic Cholecystectomy: Learning Curve, Advantages,
and Limitations,” Journal of Surgical Research, Vol. 136, 2006, pp. 172–178.
18 Goh P, Lomanto D, and So J, “Robotic-Assisted Laparoscopic Cholecystectomy,” Surgical Endoscopy,
Vol. 16, 2002, p. 216.
19
Beecken W, Wolfram M, Engl T, et al., “Robotic-Assisted Laparoscopic Radical Cystectomy and IntraAbdominal Formation of an Orthotopic Ileal Neobladder,” European Urology, Vol. 44, 2003, p. 337.
20
Murphy DG, Challacombe BJ, Elhage O, et al., “Robotic-Assisted Laparoscopic Radical Cystectomy with
Extracorporeal Urinary Diversion: Initial Experience,” European Urology, Vol. 54, 2008, pp. 570–580.
21
Kessaris N, Papalois V, Canelo R, and Hakim N, “Live Kidney Transplantation,” Living Related Transplantation, 2010, p. 39.
22
International Directory of Company Histories: Cengage Gale, “Intuitive Surgical,” undated. As of January 16, 2014:
http://www.answers.com/topic/intuitive-surgical-inc
23
Intuitive Surgical, “Form 10-K,” 2005. As of January 16, 2014:
http://www.sec.gov/Archives/edgar/data/1035267/000119312505052547/d10k.htm
24
Intuitive Surgical, “Form S-4,” 2003. As of January 16, 2014:
http://www.sec.gov/Archives/edgar/data/1035267/000089161803001503/f88583orsv4.htm#007
25
Boehm DH, Reichenspurner H, Gulbins H, et al., “Early Experience with Robotic Technology for Coronary Artery Surgery,” Annals of Thoracic Surgery, Vol. 68, 1999, pp. 1542–1546.
26 Prasad
SM, Maniar HS, Soper NJ, Damiano RJ, Jr., and Klingensmith ME, “The Effect of Robotic
Assistance on Learning Curves for Basic Laparoscopic Skills,” American Journal of Surgery, Vol. 183, 2002,
pp. 702–707.
27
Bargar WL, Bauer A, and Borner M, “Primary and Revision Total Hip Replacement Using the Robodoc
System,” Clinical Orthopedics and Related Research, 1998, pp. 82–91.
28
Barbash GI and Glied SA, “New Technology and Health Care Costs—The Case of Robot-Assisted Surgery,” New England Journal of Medicine, Vol. 363, 2010, pp. 701–704.
29
Zhao D and Liu D, “Clinical Training Technology for Vascular Interventional Surgery Robot System
Based on Master-Slave Expansion,” 2012 International Conference on Mechatronics and Automation (ICMA),
IEEE, 2012, pp. 604–610.
30
Matthews CA, “The Promise of Robotics in Urogynecology,” International Urogynecology Journal, 2012,
pp. 1–2.
31
Intuitive Surgical, “Investor Presentation Q4 2012,” 2012. As of January 20, 2014:
http://phx.corporate-ir.net/phoenix.zhtml?c=122359&p=irol-IRHome
32
Finnegan KT, Meraney AM, Staff I, and Shichman SJ, “da Vinci Skills Simulator Construct Validation
Study: Correlation of Prior Robotic Experience with Overall Score and Time Score Simulator Performance,”
Urology, 2012.
33
Ishikawa N, Watanabe G, Inaki N, Moriyama H, Shimada M, and Kawaguchi M, “The da Vinci Surgical
System Versus the Radius Surgical System,” Surgical Science, Vol. 3, 2012, pp. 358–361.
Case Study 7: Robotic Surgery
105
34 Lotan
Y, Cadeddu JA, and Gettman MT, “The New Economics of Radical Prostatectomy: Cost Comparison of Open, Laparoscopic and Robot Assisted Techniques,” Journal of Urology, Vol. 172, 2004,
pp. 1431–1435.
35
Ramsay C, Pickard R, Robertson C, et al., “Systematic Review and Economic Modelling of the Relative
Clinical Benefit and Cost-Effectiveness of Laparoscopic Surgery and Robotic Surgery for Removal of the Prostate in Men with Localised Prostate Cancer,” Health Technology Assessment (Winchester, England), Vol. 16,
2012, p. 1.
36
Amodeo A, Linares QA, Joseph J, Belgrano E, and Patel H, “Robotic Laparoscopic Surgery: Cost and
Training,” Minerva Urologica e Nefrologica [Italian Journal of Urology and Nephrology], Vol. 61, 2009, p. 121.
37
Weinstein GS, O’Malley Jr BW, Desai SC, and Quon H, “Transoral Robotic Surgery: Does the Ends [sic]
Justify the Means?” Current Opinion in Otolaryngology & Head and Neck Surgery, Vol. 17, 2009, pp. 126–131.
38
Wexner SD, Bergamaschi R, Lacy A, et al., “The Current Status of Robotic Pelvic Surgery: Results of a
Multinational Interdisciplinary Consensus Conference,” Surgical Endoscopy, Vol. 23, 2009, pp. 438–443.
39
Intuitive Surgical, “Annual Report 2011, Form 10K,” 2012. As of January 19, 2014:
http://www.sec.gov/Archives/edgar/data/1035267/000119312512039226/d263902d10k.htm
40 Barbash
GI and Glied SA, “New Technology and Health Care Costs—The Case of Robot-Assisted Surgery,” New England Journal of Medicine, Vol. 363, 2010, pp. 701–704.
41
Garber AM, “The Price of Growth in the Medical-Device Industry,” New England Journal of Medicine,
Vol. 355, 2006, pp. 337–339.
42 Leff B and Finucane TE, “Gizmo Idolatry,” Journal of the American Medical Association, Vol. 299, 2008,
pp. 1830–1832.
43
Trinh HQ, Begun JW, and Luke RD, “Hospital Service Duplication: Evidence on the Medical Arms Race,”
Health Care Management Review, Vol. 33, 2008, pp. 192–202.
44 Levy
P, “Uncle!” Not Running a Hospital Blog, 2008. As of January 23, 2014:
http://runningahospital.blogspot.com/2008/11/uncle.html
45
Steinberg PL, Merguerian PA, Bihrle W, and Seigne JD, “The Cost of Learning Robotic-Assisted Prostatectomy,” Urology, Vol. 72, 2008, pp. 1068–1072.
46 Patel
VR, Tully A, Holmes R, and Lindsay J, “Robotic Radical Prostatectomy in the Community Setting—
The Learning Curve and Beyond: Initial 200 Cases,” Journal of Urology, Vol. 174, 2005, pp. 269–272.
47
Su LM, “Robot-Assisted Radical Prostatectomy: Advances Since 2005,” Current Opinion in Urology,
Vol. 20, 2010, pp. 130–135.
48
Burgess SV, Atug F, Castle EP, Davis R, and Thomas R, “Cost Analysis of Radical Retropubic, Perineal,
and Robotic Prostatectomy,” Journal of Endourology, Vol. 20, 2006, pp. 827–830.
49
Tewari A, El-Hakim A, and Leung RA, “Robotic Prostatectomy: A Pooled Analysis of Published Literature,” Expert Review of Anticancer Therapy, Vol. 6, 2006, pp. 11–20.
50
Hugosson J, Carlsson S, Aus G, et al., “Mortality Results from the Göteborg Randomised PopulationBased Prostate-Cancer Screening Trial,” Lancet Oncology, Vol. 11, 2010, pp. 725–732.
51
Andriole GL, Crawford ED, Grubb III RL, et al., “Mortality Results from a Randomized Prostate-Cancer
Screening Trial,” New England Journal of Medicine, Vol. 360, 2009, pp. 1310–1319.
52
Andriole GL, Crawford ED, Grubb RL, et al., “Prostate Cancer Screening in the Randomized Prostate,
Lung, Colorectal, and Ovarian Cancer Screening Trial: Mortality Results After 13 Years of Follow-Up,” Journal of the National Cancer Institute, Vol. 104, 2012, pp. 125–132.
53
Makarov DV, Yu JB, Desai RA, Penson DF, and Gross CP, “The Association Between Diffusion of the Surgical Robot and Radical Prostatectomy Rates,” Medical Care, Vol. 49, 2011, pp. 333–339.
106
Redirecting Innovation in U.S. Health Care: Case Studies
54
Stitzenberg KB, Wong YN, Nielsen ME, Egleston BL, and Uzzo RG, “Trends in Radical Prostatectomy:
Centralization, Robotics, and Access to Urologic Cancer Care,” Cancer, Vol. 118, 2012, pp. 54–62.
55
Lowrance WT, Eastham JA, Savage C, et al., “Contemporary Open and Robotic Radical Prostatectomy
Practice Patterns Among Urologists in the United States,” Journal of Urology, Vol. 187, 2012, pp. 2087–2092.
56
Jacobs BL, Zhang Y, Schroeck FR, et al., “Use of Advanced Treatment Technologies Among Men at
Low Risk of Dying from Prostate Cancer,” Journal of the American Medical Association, Vol. 309, 2013,
pp. 2587–2595.
57
Thompson I, Thrasher JB, Aus G, et al., “Guideline for the Management of Clinically Localized Prostate
Cancer: 2007 Update,” Journal of Urology, Vol. 177, 2007, pp. 2106–2131.
58
Carter HB, Albertsen PC, Barry MJ, et al., “Early Detection of Prostate Cancer: AUA Guideline,” Journal
of Urology, Vol. 190, 2013, pp. 419–426.
59 U.S. Department of Justice and the Federal Trade Commission, Horizontal Merger Guidelines, 2010. As of
January 23, 2014:
http://www.justice.gov/atr/public/guidelines/hmg-2010.html
60 Feifer AH, Elkin EB, Lowrance WT, et al., “Temporal Trends and Predictors of Pelvic Lymph Node Dissection in Open or Minimally Invasive Radical Prostatectomy,” Cancer, 2011.
61
Kim SP, Boorjian SA, Shah ND, et al., “Disparities in Access to Hospitals with Robotic Surgery for Patients
with Prostate Cancer Undergoing Radical Prostatectomy,” Journal of Urology, Vol. 189, 2013, pp. 514–520.
62
Ficarra V, Novara G, Fracalanza S, et al., “A Prospective, Non-Randomized Trial Comparing RobotAssisted Laparoscopic and Retropubic Radical Prostatectomy in One European Institution,” BJU International, Vol. 104, 2009, pp. 534–539.
63
Nelson B, Kaufman M, Broughton G, et al., “Comparison of Length of Hospital Stay Between Radical
Retropubic Prostatectomy and Robotic Assisted Laparoscopic Prostatectomy,” Journal of Urology, Vol. 177,
2007, pp. 929–931.
64 Farnham
SB, Webster TM, Herrell SD, and Smith JA, Jr., “Intraoperative Blood Loss and Transfusion
Requirements for Robotic-Assisted Radical Prostatectomy Versus Radical Retropubic Prostatectomy,” Urology,
Vol. 67, 2006, pp. 360–363.
65
Miller J, Smith A, Kouba E, Wallen E, and Pruthi RS, “Prospective Evaluation of Short-Term Impact and
Recovery of Health Related Quality of Life in Men Undergoing Robotic Assisted Laparoscopic Radical Prostatectomy Versus Open Radical Prostatectomy,” Journal of Urology, Vol. 178, 2007, pp. 854–858, discussion
p. 859.
66 Tewari
A, Srivasatava A, and Menon M, “A Prospective Comparison of Radical Retropubic and RobotAssisted Prostatectomy: Experience in One Institution,” BJU International, Vol. 92, 2003, pp. 205–210.
67
Tewari A, Peabody JO, Fischer M, et al., “An Operative and Anatomic Study to Help in Nerve Sparing
During Laparoscopic and Robotic Radical Prostatectomy,” European Urology, Vol. 43, 2003, pp. 444–454.
68
Menon M, Tewari A, Baize B, Guillonneau B, and Vallancien G, “Prospective Comparison of Radical Retropubic Prostatectomy and Robot-Assisted Anatomic Prostatectomy: The Vattikuti Urology Institute Experience,” Urology, Vol. 60, 2002, pp. 864–868.
69
Smith JA, Jr., Chan RC, Chang SS, et al., “A Comparison of the Incidence and Location of Positive Surgical Margins in Robotic Assisted Laparoscopic Radical Prostatectomy and Open Retropubic Radical Prostatectomy,” Journal of Urology, Vol. 178, 2007, pp. 2385–2389, discussion pp. 2389–2390.
70
Krambeck AE, DiMarco DS, Rangel LJ, et al., “Radical Prostatectomy for Prostatic Adenocarcinoma: A
Matched Comparison of Open Retropubic and Robot-Assisted Techniques,” BJU International, Vol. 103,
2009, pp. 448–453.
71
Ahlering TE, Eichel L, and Skarecky D, “Evaluation of Long-Term Thermal Injury Using Cautery During
Nerve Sparing Robotic Prostatectomy,” Urology, Vol. 72, 2008, pp. 1371–1374.
Case Study 7: Robotic Surgery
107
72
Ahlering TE, Eichel L, Edwards RA, Lee DI, and Skarecky DW, “Robotic Radical Prostatectomy: A Technique to Reduce PT2 Positive Margins,” Urology, Vol. 64, 2004, pp. 1224–1228.
73
Gleitsmann K, Bunker K, Kriz H, et al., Robotic Assisted Surgery, Portland, Oreg.: Center for EvidenceBased Policy, Oregon Health and Science University, 2012.
74
Hu JC, Gu X, Lipsitz SR, et al., “Comparative Effectiveness of Minimally Invasive Vs Open Radical Prostatectomy,” Journal of the American Medical Association, Vol. 302, 2009, pp. 1557–1564.
75
Schiavone MB, Kuo EC, Naumann RW, et al., “The Commercialization of Robotic Surgery: Unsubstantiated Marketing of Gynecologic Surgery by Hospitals,” American Journal of Obstetrics and Gynecology, 2012.
76
Jin LX, Ibrahim AM, Newman NA, Makarov DV, Pronovost PJ, and Makary MA, “Robotic Surgery
Claims on United States Hospital Websites,” Journal for Healthcare Quality, Vol. 33, 2011, pp. 48–52.
77 Reed
MP, Saliba S, and White V, “Marketing of a Robotic Surgical Program,” Robotic Urologic Surgery,
2012, pp. 509–527.
78 Estey EP, “Robotic Prostatectomy: The New Standard of Care or a Marketing Success?” Canadian Urological Association Journal, Vol. 3, 2009, p. 488.
79
Guthrie P and Rosenblatt J, “Intuitive Wins Trial, Defeats Negligent Claims,” Bloomberg, May 23, 2013.
As of January 24, 2014:
http://www.bloomberg.com/news/2013-05-23/intuitive-wins-trial-defeats-negligent-training-claims.html
80 Nguyen
PL, Gu X, Lipsitz SR, et al, “Cost Implications of the Rapid Adoption of Newer Technologies for
Treating Prostate Cancer,” Journal of Clinical Oncology, Vol. 29, 2011, pp. 1517–1524.
81
Daskivich TJ, Chamie K, Kwan L, et al., “Overtreatment of Men with Low-Risk Prostate Cancer and Significant Comorbidity,” Cancer, Vol. 117, 2011, pp. 2058–2066.
82
Chamie K, Daskivich TJ, Kwan L, et al., “Comorbidities, Treatment and Ensuing Survival in Men with
Prostate Cancer,” Journal of General Internal Medicine, Vol. 27, 2012, pp. 492–499.
83 Emanuel EJ and Person SD, “It Costs More, But Is It Worth More?” New York Times, January 2, 2012. As
of January 30, 2014:
http://opinionator.blogs.nytimes.com/2012/01/02/it-costs-more-but-is-it-worth-more/?_r=0
84 Allen
AM, Pawlicki T, Dong L, et al., “An Evidence Based Review of Proton Beam Therapy: The Report
of ASTRO’s Emerging Technology Committee,” Radiotherapy and Oncology, Vol. 103, No. 1, April 2012,
pp. 8–11. As of January 30, 2014:
http://www.ncbi.nlm.nih.gov/pubmed/22405807
Case Study 8
Telemedicine
Hiu-Fai Fong, M.D.1
Fellow, Robert Wood Johnson Clinical Scholars Program,
University of Pennsylvania Fellow, Child Abuse Pediatrics, Children’s Hospital of Philadelphia
The Concept
Telemedicine has its roots in Europe. In 1905, Dutch physician and inventor Willem Einthoven
(1860–1927) demonstrated the feasibility of transmitting and receiving cardiograms over telephone lines. He referred to this as the “tele-cardiogram.” In the United States and Canada,
several projects with different applications were conducted between 1948 and the early 1970s,
involving psychiatry and single-specialty and multispecialty clinics, using various terms with
the prefix tele- (meaning “distance”). Kenneth Bird popularized the term “telemedicine” in the
1970s.1
Telemedicine represents a modality in health care delivery that relies on information and
communication technology (ICT). Types of health care services include primary and specialty
medical care, patient monitoring, consumer health information, and medical education.2 The
way the services are delivered generally entails one of two basic modes that largely differ in
terms of the timing of information acquisition, transmission, and response between the parties
that send and receive the information.
Store-and-Forward (Asynchronous) Mode
This mode involves collection of data (including presenting symptoms, medical history, lab
tests, and images), transmission, retrieval, and response. Clinical data are acquired from the
patient at the remote site and transmitted to the consultant (typically in a medical center),
who reviews and interprets the data at a time convenient to her or him and subsequently
renders an opinion about diagnosis and/or treatment. The two-way transmission of information is “asynchronous” because it does not require simultaneous availability of the patient and
health care consultant during the encounter.3, 4 It is often used in radiology, dermatology, and
pathology—specialties that rely on images for diagnosis and treatment. In the case of radiology
and pathology, contact with the patient is not normally required. A somewhat similar application is observed in the remote monitoring and management of chronic illnesses, such as heart
disease, diabetes, and chronic obstructive pulmonary disease. In this case, patients transmit
hemodynamic, anthropomorphic, or laboratory data (e.g., blood pressure, heart rate, weight,
blood glucose measurements), collected manually or electronically, to a clinic or call center
staffed with nurses and specialists. Feedback may or may not be provided in real time.4
1 Acknowledgments:
To Rashid L. Bashshur, Ph.D., for his editorial support.
109
110
Redirecting Innovation in U.S. Health Care: Case Studies
Real-Time (Synchronous) Mode
In this mode, patient and provider interact in real time via ICT. Hence, the encounter is synchronous (occurring at the same time but at different locations). The processing and interpretation of data occur in real time and require the simultaneous availability of the patient (often
attended to by the local health care provider) and the remote consultant (typically located at a
medical center), who provides diagnosis and treatment recommendations. This mode is used
most commonly to deliver specialty care in various areas, as well as primary and critical care.1, 4
Depending on the type of service being delivered, telemedicine has been used to connect such parties as medical centers, clinics, independent providers, monitoring centers, and
patients. It employs a range of technologies, including videoconferencing, Internet links, and
mobile devices for information exchange between provider and client.1
Rationale
The goal of telemedicine is to extend the reach of health care providers to serve patients beyond
the constraints of geography and time. Beneficiaries have included remote, confined, or underserved populations. Primary objectives include the following:1–3, 5
1. to promote equity of access to care for those who are geographically, financially, logistically, and/or culturally challenged in the receipt of care
2. to enhance the efficiency and effectiveness of health care systems by relying on rational
triage of patients at the point of need, providing care at the appropriate place and time,
and relying on established protocols
3. to reduce opportunity cost for patients, as well as itinerant providers, by obviating the
need for unnecessary travel
4. to contain health care costs by reducing unnecessary duplication of diagnostic testing;
reducing unnecessary hospitalization and emergency room visits; and fostering information-sharing among health care providers, institutions, and patients.
The case for telemedicine is even more compelling in the context of health care reform.6
Bashshur and Shannon, writing on behalf of a diverse group of health care academicians,
providers, researchers, and industry representatives across the United States with expertise in
telemedicine, provide a strong case for the necessity of telemedicine in effective health care
reform.5
Development
Although the concept of remotely guided health care is not new, it has dramatically evolved
with advances in technology since the 19th century. In History of Telemedicine: Evolution, Context, and Transformation (2009), Bashshur and Shannon deliver a comprehensive account of
the history of telemedicine, from its genesis in ancient societies to its adoption by the modern
health care system.5
Telegraphy, an early form of telemedicine, was used to transmit casualty lists and facilitate ordering medical supplies on the battlefield during the Civil War. The invention of the
Case Study 8: Telemedicine
111
telephone in 1876 and the subsequent development of telephonic networks would later enable
more-rapid, bi-directional voice and data transmission for medical diagnosis and treatment,
such as amplified stethoscope sounds, electrocardiograms, and electroencephalograms. At the
end of the 19th century, radio technology was used to extend medical care to seafarers.1, 3
Dutch physician and physiologist Willem Einthoven, considered to be the first authentic telemedicine pioneer, used the newly developed galvanometer to measure electrical charges from
the heart and transmit them via telephone wire as the first “tele-cardiogram” in 1905.1
Organized telemedicine systems started in the 1960s and 1970s and relied on closed circuit television, microwave transmission, and satellite communications:
• In 1964, the U.S. National Institute for Mental Health (NIMH) funded the creation of
a two-way closed-circuit television system between the Nebraska Psychiatric Institute in
Omaha and the Norfolk State mental hospital 112 miles away. The system was used to
provide interactive medical consultations and education between specialists and general
practitioners.7
• In 1967, Kenneth T. Bird created an interactive television system that linked Boston’s
Logan Airport and Massachusetts General Hospital. The system used direct microwave
transmission to provide 24-hour medical consultation to over 1,000 airline travelers and
employees.8
The National Aeronautics and Space Administration (NASA), which had been using
telemedicine to meet the health care needs of astronauts on space flight missions, played an
instrumental role in the development and application of these technologies. NASA’s early telemedicine systems were developed to collect basic physiologic data to study the effects of space
travel on the human body. In the 1970s, NASA began applying its resources and experience to
several terrestrial health care projects:9
• In 1971, NASA’s Applied Technology Satellite-1 (ATS-1) was used by the U.S. National
Library of Medicine’s Lister Hill National Center for Biomedical Communication to
study the reliability of providing satellite-mediated video consultation to 26 Alaskan village sites.10 This was the first part of a multi-stage project that ultimately linked remote
villages, field service unit hospitals, and medical centers via a satellite voice network that
was credited with improving the efficiency of medical care and saving lives in the region.
Because of the success of this project, Alaska dedicated $5 million toward the purchase
and installation of satellite earth stations in 200 remote villages, linking each of them
with regional hospitals and the Alaska Native Medical Center.1
• In 1972, NASA began the Space Technology Applied to Rural Papago Advanced Health
Care (STARPAHC) program to test a terrestrial-based telemedicine system for its manned
space program. The test site was the Papago (now Tohono O’odham) reservation. In addition to fixed locations on the reservation near Tucson and Phoenix, Arizona, a specially
designed and fully equipped mobile health unit (MHU) traveled on a scheduled route to
serve sparsely populated destinations on the reservation. The MHU was linked via twoway microwave transmission (including redundant narrow- and broadband) to physician
consultants at the small Indian Health Service (IHS) hospital in Sells on the reservation
and a referral IHS diagnostic center in Phoenix. Fully sponsored by NASA, the project
was implemented jointly by NASA, the Indian Health Service, and Lockheed Missile and
112
Redirecting Innovation in U.S. Health Care: Case Studies
Space Company. The STARPAHC project was operational for three to four years and was
gradually phased out by NASA.1, 11
• In 1977, satellite technology was adopted internationally when Canada’s Memorial University of Newfoundland employed the joint Canadian/U.S. Hermes satellite to deliver
distance education and medical care across the province.12
The 1960s and 1970s laid the conceptual foundations for modern telemedicine. Much
of the efforts were supported by federal agencies, including NASA; the U.S. Department
of Health, Education, and Welfare (DHEW); and the U.S. National Science Foundation,
especially the seven exploratory projects funded by the Health Care Technology Division of
DHEW. There were several limitations in these early telemedicine projects, including expensive, bulky, and unreliable equipment; restrictions in licensed public use frequencies; limited
acceptance by the medical profession; design flaws; and a lack of business plans that would
secure long-term sustainability. The issue of reimbursement for telemedicine providers was not
relevant at this time; demonstration projects were terminated when public funding expired.
Telemedicine did not take off until the 1990s and 2000s, when advances in computers and
telecommunications and the dropping cost of information technology enabled its adoption in
a growing number of clinical practices. The recognition that telemedicine had the potential to
help address problems in health care delivery (e.g., limited access, rising costs, and variations
in quality) also encouraged its adoption.
Early Adoption
In the 1980s, most of the activity was focused on teleradiology, the quintessential form of
“store-and-forward” telemedicine. The first teleradiology systems were characterized by inefficiency, poor quality, limited scalability, and high cost. Photos or videos of hard copy radiology
films were digitized for image transfer, enabling radiologists outside the hospital to provide
remote and after-hours readings. Despite the potential benefits of these systems, the secondary
digitization process was cumbersome and time-consuming because images had to be handled
one at a time. The degradation of image quality during transmission was another concern.
Other obstacles included the expense of the equipment (e.g., computer, digital imaging and
data transmission systems) and the proprietary nature of teleradiography software.9
By the 1990s and 2000s, the rise of the Internet and the rapid incorporation of picture archiving and communications systems (PACS) into radiology transformed the field.9
PACS are medical imaging systems that allow storage, retrieval, transmission, and archiving
of digital images. They consist of an imaging modality, a secure network for data transmission, computer-based workstations for image interpretation, and archives for image storage and
retrieval.13 The digital image communication in medicine (DICOM) standard was established
by the American College of Radiology and the National Electrical Manufacturing Association,
helping to ensure the quality of images transferred. For an in-depth review of the history and
key issues pertaining to teleradiology, see Joseph Gitlin’s chapter in Telemedicine: Theory and
Practice.14
Some of the first PACS research and development projects were funded by the U.S. Army
in the early to mid-1980s. One of these was the Installation Site for Digital Imaging Network
and Picture Archiving and Communication System (DIN/PACS) project, which was led by
Case Study 8: Telemedicine
113
the MITRE Corporation in collaboration with Philips Medical Systems and AT&T and rolled
out at two academic universities.15 Over the next decade, the National Institutes of Health
funded several large-scale PACS research projects at the University of California, Los Angeles
(UCLA), which ultimately led to introduction of PACS into clinical care at that institution.
Expertise gained in the UCLA projects was used to develop a hospital-integrated PACS at the
University of California, San Francisco (UCSF), Medical Center in 1996. Funding for the
latter project was provided by the National Library of Medicine, the Army Medical Research
and Materiel Command, the National Cancer Institute, the California Cancer Research Program, Pacific Bell, and UCSF.16
The development of compression algorithms opened the door for the transmission of
large amounts of data over limited bandwidth, which improved digital image capture and data
transmission. PACS systems produced rapid improvements in teleradiology’s efficiency, quality, scalability, and cost. Simultaneously, the development of newer digital imaging techniques
(e.g., computed tomography and ultrasound) increased the opportunities for teleradiology. The
incorporation of PACS into clinical practice enabled the existing supply of radiologists to meet
this demand by providing interpretations remotely without the earlier concerns about excessive
cost and the degradation of image quality.9
In addition to the store-and-forward systems used by teleradiology, real-time telemedicine
applications began to diffuse rather rapidly. For example, telepsychiatry—the delivery of psychiatric services using live interactive videoconferencing—emerged as an alternative to faceto-face treatment.17 The development of computer-based systems, videoconferencing, mobile
technology, and digital communications expanded opportunities for patients to receive mental
health treatment from providers located elsewhere. Similar technology is being employed by
neurologists to manage both acute and chronic neurologic conditions, such as stroke, epilepsy,
and movement disorders, particularly in patients who live in rural areas and have limited ability to travel.18
Subsequent Use
Expansion to Clinical and Academic Environments
Once the pioneer projects of the 1960s showed that the concept was feasible and advances
in technology made it practical, telemedicine began to expand in both clinical and academic
arenas. A 2002 survey of teleconsultation activity (excluding teleradiology) in the United States
reported that over 85,000 teleconsultations were performed by more than 200 programs in
30 specialties.19 The most represented medical fields were mental health, pediatrics, dermatology, cardiology, and orthopedics.19 The literature on telemedicine grew as well. Between 1964
and 2003, the number of papers on telemedicine indexed on MEDLINE increased from a few
each year to 700–800 annually, with the greatest output occurring during the 1990s.20 Two
peer-reviewed journals dedicated to telemedicine were launched: Telemedicine and E-Health
and the Journal of Telemedicine and Telecare.21 The American Telemedicine Association (ATA)
was founded in 1993 as a nonprofit organization composed of individuals, health care institutions, companies, and other organizations dedicated to promoting telemedicine throughout
the world.22
114
Redirecting Innovation in U.S. Health Care: Case Studies
The Affordable Care Act Encourages Telemedicine to Expand Access
More recently, telemedicine was embraced by the drafters of the Affordable Care Act of 2009
(ACA) as a promising means of expanding access to health care and care coordination across
settings. The ACA directs the Center for Medicare and Medicaid Innovation (CMMI) to study
ways to improve the use of telemedicine to treat behavioral problems and stroke in the IHS as
part of a broader goal to promote cost-effective care. CMMI is also encouraged to incorporate
such technologies as patient-based remote monitoring systems into delivery models to improve
care coordination.6, 18
Advances in Telecommunications and Computer Technology Facilitate Adoption
By far the most important factor facilitating the adoption and dissemination of telemedicine has been the rapid advances in telecommunications and computer technology that have
occurred over the last 10–20 years. In particular, the switch from analog to digital technology and the decreasing cost of information transmission have played a major role in fueling
the expansion of telemedicine. In addition, the creation of the Internet, computer networks,
and web-based applications have enabled a variety of real-time telemedicine applications that
would have been technically impossible a generation ago.3, 9 However, as the barriers of technology have fallen by the wayside, new barriers have emerged to limit the widespread adoption
and dissemination of telemedicine.
Barriers
Regulatory Barriers
Perhaps the most formidable barrier to expanding telemedicine today is the current system of
medical licensure regulation. In our federal system of government, the states have the authority
to ensure the quality of medical care provided to their residents. One means by which they do
this is by regulating medical licensure. State-level regulation worked well for many years when
medical care was almost exclusively delivered on a face-to-face basis. However, it presents challenges to physicians who want to practice telemedicine across state borders.
Currently, all state medical licensing boards require that a physician engaging in telemedicine hold a medical license in the state where the patient is located.23 There is no reciprocity
between states to allow a physician licensed in one state to practice telemedicine in another,
despite the existence of national board examinations, national specialty certification, a national
malpractice data bank, and the similar medical licensing requirements between the states.24
In 1994, the American College of Radiology (ACR) published the ACR Standards for
Teleradiology. The standards mirror existing state laws by asserting that physicians providing
official interpretations through teleradiology should maintain licensure in both the initiating and receiving states. Furthermore, the ACR standards stipulated that physicians also hold
staff hospital staff credentials at the site of the patient encounter.9 Some states go even further,
requiring the practitioner to obtain written consent from the patient to participate in telemedicine, to have a medical staff person present with the patient, and to incorporate face-to-face
visits into the treatment plan.17 Presumably, the rationale for these policies is protection of
patient safety. But they are sufficiently burdensome to make telemedicine infeasible in most
instances. Also, there are no data to support the need for these requirements or to prove that
their addition provides the desired margin of safety.
Case Study 8: Telemedicine
115
The need for physicians to obtain licensure across multiple states—and even credentials
at each individual hospital—serves as a major barrier to the expansion of telemedicine. As a
result, few physicians are legally credentialed to provide telemedical care to patients in remote
or underserved areas. Until a solution to this problem is found, the potential benefits of telemedicine will not be fully realized, despite needy patients, willing providers, and the requisite
enabling technology. The consequences of limiting expansion of telemedicine will become
increasingly apparent when ACA implementation increases the number of insured patients in
2014 and thereafter.
In 2011, the ATA proposed four potential strategies to reduce the regulatory barriers that
impede the practice and expansion of telemedicine:24
1. Create a federal medical licensure and regulation system (modeled after that used by the
Federal Aviation Administration for licensure of civilian airline pilots).
2. Maintain a state-level medical licensure and regulation system without the requirement
that practitioners of telemedicine be licensed in the state in which the patient resides.
Instead, the state in which the provider is located would have regulatory authority over
the licensing and activities of the provider.
3. Grant functional licensure, which would supersede state requirements, in defined circumstances. An example of this is the Veterans E-Health and Telemedicine Support
(VETS) Act (H.R. 2001), a bill introduced in Congress in May 2013. This act would
provide an exemption from the need to obtain multiple state licenses to U.S. Department of Veterans Affairs health care providers.25
4. Create a system of mutual recognition and portability of medical licenses between states
(modeled after that used for issuing drivers licenses).
Legislation designed to foster a voluntary national medical licensure system is currently
being drafted by Sen. Tom Udall of New Mexico, who represents a state with substantial rural
and frontier areas. The proposed system would provide a uniform set of standards and application process for national medical licensure and create a comprehensive data exchange system
for primary source verification of credentials. Physicians holding a valid state medical license
would be eligible to apply for a national license. The combination of the two would enable a
physician to practice telemedicine across state lines.26
Financial Barriers
Expansion of telemedicine is hindered by financial barriers as well. Chief among these is the
still-substantial cost of the technology and insufficient reimbursement for telemedicine services. For example, restrictive Medicare reimbursement policies leave substantial gaps in coverage. Medicare only reimburses for telemedicine services provided to patients living in a Health
Provider Shortage Area or a county outside of a metropolitan statistical area (MSA). Given that
nearly 80 percent of Medicare beneficiaries live within one of the 1,092 MSAs (yet still have
poor access to care), this policy significantly limits the number of patients who might otherwise benefit from telemedicine services.
Furthermore, reimbursement is contingent upon the patient being physically present in a
“medical facility” (e.g., a provider’s office, hospital, or rural health clinic). Home consultations
do not qualify. This provision preempts one of the principal benefits of telemedicine: providing
expert care to mobility-impaired patients and those who live in remote settings.27
116
Redirecting Innovation in U.S. Health Care: Case Studies
Medicaid imposes additional restrictions. Because Medicaid is a joint federal and state
program, the states can decide whether to cover telemedicine services. States may choose which
services to cover, where in the state telemedicine can be provided, what types of providers are
covered, how much providers will be reimbursed, and other terms.28 To date, only ten states
have legislation authorizing Medicaid coverage for telemedicine, mostly interactive video
consultations.29
Private health plans have their own restrictions. Currently, there is no widely accepted
standard among private plans for reimbursing telemedicine providers, so their approaches are
highly variable.27 Only 19 states have legislation mandating coverage of telemedicine services
by private payers.29 Taken together, the restrictive reimbursement policies by government and
commercial payers sharply limit the viability of telemedicine services, inhibiting the pace and
extent of adoption.30
The one exception to the slow pace of adoption is teleradiology. Because remote viewing
of medical images—a core aspect of teleradiology—is not considered “telemedicine” by the
Centers for Medicare & Medicaid Services (CMS), teleradiology services are quickly and consistently reimbursed.9 Thus, the use of PACS and other aspects of teleradiology is widespread.
Because reimbursement is so variable and is often lower than care provided in a typical
face-to-face encounter, the cost of acquiring the technology remains a barrier as well. Prices are
much lower than a decade ago, but they are still substantial. Miller et al. reported that the costs
for purchasing a basic electronic medical record system—a prerequisite to an efficient telemedicine capability—are approximately $44,000 per physician for initial startup and $8,500 per
physician per year for maintenance, respectively.31 Regions of the United States that lack the
infrastructure or funding to acquire this technology face particular challenges, leading to geographical and socioeconomic disparities in access to telemedicine services.21
Provider Reluctance
The third major barrier is the lack of interest on the part of many health care providers to
incorporate telemedicine into their practice. Providers may be reluctant to invest the time and
resources into learning and staying updated on new technology and telemedicine systems.21
Others have cited concerns about disrupting the traditional physician-patient relationship
through the interposition of technology.18
Medico-Legal Concerns
Telemedicine services, which may consist of consultations through video, email, or other electronic media, create scenarios that challenge existing patient privacy and malpractice laws.
In telepsychiatry, psychiatrists providing mental health assessments and treatment to patients
through videoconferencing systems must ensure that their systems meet the requirements of
the Health Insurance Portability and Accountability Act (HIPAA) and federal and state privacy statutes.17 The ways in which current legal criteria apply to telemedicine consultations
using electronic media are unclear, including professional expectations, practice standards, and
potential liabilities associated with use of the technology.21
Uncertain Benefits
Whether policymakers, private investors, and health care providers will push for the expansion
of telemedicine depends in large part on the compilation of evidence of its clinical and economic value.21 This is discussed more fully in the next section.
Case Study 8: Telemedicine
117
Cost and Health Impact
Telemedicine was created to expand access to care, improve quality of services provided to
rural and underserved populations, promote health care equity, increase work efficiency, reduce
health care costs, and improve inter-professional communications and information-sharing
within the health care system. However, the extent to which these benefits have been documented in the medical literature is limited. One reason is that the adoption of telemedicine
into clinical practice—and, therefore, the study of its clinical and economic outcomes—is relatively new. This creates a catch-22: Although telemedicine is technically feasible, its impact on
costs and quality of care has not been demonstrated in a compelling way. But without adequate
dissemination to conduct the evaluation, it is impossible to prove its potential impact.
Second, the heterogeneity of the interventions makes it challenging to draw general conclusions about the overall effectiveness of telemedicine. Individual programs have demonstrated positive outcomes among patients, families, health care providers, and the health care
system, including increased access to health care services, cost-effectiveness, enhanced educational opportunities, improved health outcomes, better quality of care, and enhanced social
support.32, 33 However, when these studies are merged in systematic reviews and meta-analyses,
the results are mixed.
For example, Currell et al. conducted a systematic review of studies comparing telemedicine programs with face-to-face patient care.34 Seven studies involving 800 patients were
included. Included telemedicine programs used at least two communication media interactively, and the majority involved home care or patient self-monitoring for chronic disease.
The review demonstrated feasibility and acceptability of the programs; however, the impact
of telemedicine on clinical outcomes, including safety, was unclear. Furthermore, there was
insufficient information to determine the cost-effectiveness of the interventions.34 In another
systematic review, Whitten et al. concluded that there was little published evidence to confirm
whether telemedicine is a cost-effective alternative to standard health care delivery.35
Notwithstanding these studies, there is also evidence of the effectiveness of telemedicine
in the medical literature. As one example, McLean et al. conducted a meta-analysis of 21 randomized controlled trials of telemedicine interventions for pediatric and adult asthma management. Telemedicine asthma interventions used a range of technologies, including telephone,
video conferencing, Internet, text messaging, and other networked systems. They found no
difference in patient-reported quality of life and number of emergency department visits when
comparing patients who were served by telemedicine and those who were not. However, there
was a reduction in the rate of hospitalizations among patients who received the intervention.36
A second example is a longitudinal analysis of the business value of implementing and
integrating electronic medical records, radiology information systems, and PACS at a regional
medical center. Outcomes of financial revenues, operating lead times, and customer satisfaction were examined. In the 12 months after information technology implementation, there
was a 20- to 40-percent increase in revenues related to improvements in billing, an 80-percent
reduction in turnaround time for generating final radiology reports (a marker of work efficiency), and a statistically significant increase in satisfaction among referring physicians.37
It would be premature to make sweeping generalizations about the utility of telemedicine
from these reviews. Craig and Patterson suggest that the outcomes of telemedicine programs
should be evaluated on an individual basis.3 They argue that because each telemedicine program is designed to serve a specific purpose within a defined context, the relative importance
118
Redirecting Innovation in U.S. Health Care: Case Studies
of desired outcomes (feasibility, acceptability, cost, effectiveness, safety, and sustainability) will
differ from program to program. In addition, they suggest that studies comparing telemedicine interventions with face-to-face patient care may not be relevant or appropriate if the
intent of the intervention is not to replace face-to-face encounters or specialist referrals, but
rather to enable provision of care in settings where conventional services are highly limited or
nonexistent.3
It is ironic that telemedicine programs are expected to show superiority over face-to-face
encounters with clinicians, while the much larger and more powerful pharmaceutical industry
need only show that a new drug is relatively safe and better than placebo to gain regulatory
approval. If new drugs were required to show superiority over existing products on the market,
few could meet that standard.
Prospective research on the impact of telemedicine, particularly in austere settings that
currently have inadequate access to care, could resolve this question. Performing such studies
would fill a major gap in the telemedicine literature and encourage greater stakeholder support
for this promising technology.21 Until such evidence is generated, the expansion of telemedicine is likely to be slow.
In 1996, the Institute of Medicine (IOM) released the report Telemedicine: A Guide to
Assessing Telecommunications for Health Care.38 In that report, the IOM Committee on Evaluating Clinical Applications of Telemedicine offered the following observation (p. 208):
Telemedicine is similar in most respects to other technologies for which better evidence of effectiveness is also being demanded. Telemedicine, however, has some special
characteristics—shared with information technologies generally—that warrant particular
notice from evaluators and decision makers. Most notably, telemedicine is not a single technology or a discrete set of related technologies; it is, rather, a large and very heterogeneous
collection of clinical practices, technologies, and organizational arrangements. In addition,
widespread adoption of effective telemedicine applications depends on a complex, broadly
distributed technical and human infrastructure that is only partly in place and is being
profoundly affected by rapid changes in health care, information, and communications
system.
Sixteen years later, the IOM revisited telemedicine and “telehealth” in a major workshop.39 Sponsored by the Health Resources and Services Administration (HRSA), the workshop examined how telehealth could serve geographically isolated individuals and extend the
reach of scarce health care providers and services. Workshop sessions covered a wide range of
issues, including the operational difficulties of providing telehealth in rural communities, the
challenge of securing payment, use by providers in different settings across the health care
continuum, and the strengths and limitations of past evaluations. In the closing session, one
attendee reminded participants that, ultimately, the technology is less important than those it
serves (Nina M. Antoniotti, p. 128):
Telehealth is about the people, not the process. Public policy should not place barriers based
on assumptions. For example, we are mired in discussions of presuming to know what
patients want. Furthermore, we need to establish the methodology that HHS [the U.S.
Department of Health and Human Services] would use to translate consumer momentum
around mobile devices. Consumers push us to do things differently in health care, and we
do it because it makes sense, has good outcomes, and engages patients.
Case Study 8: Telemedicine
119
Adoption Block
Telemedicine is an example of adoption block: Although the technology is feasible and highly
promising, it cannot gain traction in the marketplace because its adoption threatens the
interests of powerful institutions and groups, such as local health care practitioners and state
licensing boards. Where parochial concerns do not exist and reimbursement is assured—for
instance, in teleradiology—the technology has flourished. But in other settings, adoption of
telemedicine has been severely constrained.
Telemedicine has the potential to increase provider efficiency, promote health care equity,
and enhance access to care in geographically remote and underserved communities. But until
sufficient will is mustered to break down political and regulatory barriers that are impeding its
adoption, its promise will be largely unfulfilled.
References
1 Bashshur
RL and Shannon GW, History of Telemedicine: Evolution, Context, and Transformation, New
Rochelle, N.Y.: Mary Ann Liebert, Inc., 2009.
2 American
Telemedicine Association, “What is Telemedicine?” 2013. As of June 27, 2013:
http://www.americantelemed.org/learn/what-is-telemedicine
3 Craig J and Patterson V, “Introduction to the Practice of Telemedicine,” Journal of Telemedicine and Telecare,
Vol. 11, No. 1, 2005, pp. 3–9.
4 Anker
SD, Koehler F, and Abraham WT, “Telemedicine and Remote Management of Patients with Heart
Failure,” Lancet, Vol. 378, No. 9792, August 20, 2011, pp. 731–739.
5 Bashshur RL and Shannon GW, “National Telemedicine Initiatives: Essential to Healthcare Reform,” Telemedicine and e-Health, Vol. 15, No. 6, 2009, pp. 600–610.
6 The
Patient Protection and Affordable Care Act, Public Law 111-148, 2010.
7 Benschoter
RA, Wittson CL, and Ingham CG, “Teaching and Consultation by Television: I. Closed-Circuit
Collaboration,” Mental Hospitals, Vol. 16, March 1965, pp. 99–100.
8 Murphy
RL, Jr., and Bird KT, “Telediagnosis: A New Community Health Resource; Observations on the
Feasibility of Telediagnosis Based on 1000 Patient Transactions,” American Journal of Public Health, Vol. 64,
No. 2, February 1974, pp. 113–119.
9 Thrall
JH, “Teleradiology, Part I. History and Clinical Applications,” Radiology, Vol. 243, No. 3, June 2007,
pp. 613–617.
10 Foote
DR, “Satellite Communication for Rural Health Care in Alaska,” Journal of Communication, Vol. 27,
No. 4, Autumn 1977, pp. 173–182.
11 Freiburger G, Holcomb M, and Piper D, “The STARPAHC Collection: Part of an Archive of the History of
Telemedicine,” Journal of Telemedicine and Telecare, Vol. 13, No. 5, 2007, pp. 221–223.
12 Elford
R, “Telemedicine Activities at Memorial University of Newfoundland: A Historical Review, 1975–
1997,” Telemedicine Journal: The Official Journal of the American Telemedicine Association, Vol. 4, No. 3,
Fall 1998, pp. 207–224.
13 Choplin RH, Boehme JM, 2nd, and Maynard CD, “Picture Archiving and Communication Systems: An
Overview,” Radiographics: A Review Publication of the Radiological Society of North America, Inc., Vol. 12,
No. 1, January 1992, pp. 127–129.
120
Redirecting Innovation in U.S. Health Care: Case Studies
14 Bashshur RL, Sanders JH, and Shannon GW, Telemedicine: Theory and Practice, Springfield, Ill.: Charles C.
Thomas, 1996.
15 Tong
C and Wong E, “Picture Archiving and Communication System in Health Care,” in Pagani M, ed.,
The Encyclopedia of Multimedia Technology and Networking, Hershey, Pa.: Idea Group Inc., 2005, p. 821.
16 Huang H, introduction to Huang HK et al., “Digital Radiology at the University of California, Los Angeles: A Feasibility Study,” Journal of Digital Imaging, Vol. 16, No. 1, 2003, p. 69.
17 Shore
JH, “Telepsychiatry: Videoconferencing in the Delivery of Psychiatric Care,” American Journal of Psychiatry, Vol. 170, No. 3, March 1, 2013, pp. 256–262.
18 Wechsler LR, Tsao JW, Levine SR, et al., “Teleneurology Applications: Report of the Telemedicine Work
Group of the American Academy of Neurology,” Neurology, Vol. 80, No. 7, February 12, 2013, pp. 670–676.
19 Grigsby
B, TRC Report on US Telemedicine Activity, Kingston, N.J.: Civic Research Institute, 2004.
20 Moser PL, Hauffe H, Lorenz IH, et al., “Publication Output in Telemedicine During the Period January 1964 to July 2003,” Journal of Telemedicine and Telecare, Vol. 10, No. 2, 2004, pp. 72–77.
21 Miller
EA, “Solving the Disjuncture Between Research and Practice: Telehealth Trends in the 21st Century,” Health Policy, Vol. 82, No. 2, July 2007, pp. 133–141.
22 American Telemedicine Association, “Who is ATA?” 2012. As of June 27, 2013:
http://www.americantelemed.org/about-ata/who-is-ata
23 Federation
of State Medical Boards, “Telemedicine Overview—Board by Board Approach,” last updated
August 2012. As of June 27, 2013:
http://www.fsmb.org/pdf/grpol_telemedicine_licensure.pdf
24 American
Telemedicine Association, “Policy on State Medical Licensure and Practice Requirements,”
June 2011. As of June 27, 2013:
http://www.americantelemed.org/docs/default-source/policy/
ata-policy-on-state-medical-licensure-and-practice-requirements.pdf
25 Veterans
E-Health and Telemedicine Support Act, 2013.
26 Fern
Goodhart, Office of U.S. Senator Tom Udall (New Mexico), personal communication, “Draft Bill:
Increasing Telehealth and Access to Medical Care Through Streamlining Licensing and Credentialing Act,”
2013.
27 U.S. Department of Health and Human Services, Health Resources and Services Administration, “What
Are the Reiumbursement Issues for Telehealth?” undated. As of June 27, 2013:
http://www.hrsa.gov/healthit/toolbox/RuralHealthITtoolbox/Telehealth/whatarethereimbursement.html
28 Medicaid.gov, “Telemedicine,” undated. As of June 27, 2013:
http://medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Delivery-Systems/Telemedicine.html
29 American
Telemedicine Association, “2013 State Telemedicine Legislation Tracking (as of 6/21/13),”
June 21, 2013. As of June 27, 2013:
http://www.americantelemed.org/docs/default-source/policy/
state-telemedicine-legislation-matrix.pdf?sfvrsn=36
30 American
Telemedicine Association, “2013 Telemedicine Policy Priorities,” undated. As of June 27, 2013:
http://www.americantelemed.org/docs/default-source/policy/
ata-federal-policy-priorities-for-2012.pdf?sfvrsn=32
31 Miller
RH, West C, Brown TM, Sim I, and Ganchoff C, “The Value of Electronic Health Records in Solo
or Small Group Practices,” Health Affairs, Vol. 24, No. 5, September–October 2005, pp. 1127–1137.
32 Jennett
PA, Affleck Hall L, Hailey D, et al., “The Socio-Economic Impact of Telehealth: A Systematic
Review,” Journal of Telemedicine and Telecare, Vol. 9, No. 6, 2003, pp. 311–320.
Case Study 8: Telemedicine
121
33 American
Telemedicine Association, “Telemedicine’s Impact on Healthcare Cost and Quality,” April 2013.
As of June 27, 2013:
http://www.americantelemed.org/docs/default-source/policy/
examples-of-research-outcomes---telemedicine’s-impact-on-healthcare-cost-and-quality.pdf?sfvrsn=4
34 Currell
R, Urquhart C, Wainwright P, and Lewis R, “Telemedicine Versus Face to Face Patient Care:
Effects on Professional Practice and Health Care Outcomes,” Cochrane Database of Systematic Reviews, Vol. 1,
2010.
35 Whitten
PS, Mair FS, Haycox A, May CR, Williams TL, and Hellmich S, “Systematic Review of Cost
Effectiveness Studies of Telemedicine Interventions,” BMJ, Vol. 324, No. 7351, June 15, 2002, pp. 1434–1437.
36 McLean
S, Chandler D, Nurmatov U, et al., “Telehealthcare for Asthma: A Cochrane Review,” CMAJ:
Canadian Medical Association Journal [Journal de l’Association Medicale Canadienne,] Vol. 183, No. 11,
August 9, 2011, pp. E733–E742.
37 Ayal M and Seidman A, “An Empirical Investigation of the Value of Integrating Enterprise Information Systems: The Case of Medical Imaging Informatics,” Journal of Management Information Systems, Vol. 26, 2009,
pp. 43–68.
38 Institute of Medicine, Telemedicine: A Guide to Assessing Telecommunications for Health Care, Washington,
D.C.: National Academies Press, 1996.
39 Institute
of Medicine, The Role of Telehealth in an Evolving Health Care Environment: Workshop Summary,
Washington, D.C.: National Academies Press, 2012.