Writing Clinical
Research Protocols,
Part 2:
Writing and Editing
Clinical Research
Protocols
Jonathan McCall, BA
Editor
DCRI Communications
Writing Clinical Research Protocols
Writing Clinical Research Protocols
■ General Considerations
■ Parts of the Protocol
■ Good Writing Matters
■ Looking Ahead
■ Resources and Guidance
Who Reads Protocols?
■ Keep the “audience” in mind:
■ Other physicians
■ Nurses/CRAs
■ Patients and their families
■ Patient advocates
■ IRB members
■ Scientific reviewers
Using Templates
■ Many NIH programs encourage or require the
use of protocol templates, such as the ones
available here:
■ http://ctep.cancer.gov/guidelines/templates.html
■ Following template guidelines can help guide
authors, but not every part of a template will
necessarily apply in a given study.
Writing Clinical Research Protocols
■ General Considerations
■ Parts of the Protocol
■ Good Writing Matters
■ Looking Ahead
■ Resources and Guidance
Parts of the Protocol
■ Introduction
■ Safety/adverse events
■ Objectives (including
study schema)
■ Regulatory guidance
■ Eligibility criteria
■ Study design/methods
(including drug/device
info)
■ Statistical section
(including analysis and
monitoring)
■ Human subjects
protection/informed
consent
Adapted from:
Protomechanics: Chapter 1 (http://www.cc.nih.gov/ccc/protomechanics/),
CTEP Investigators’ Handbook, 2002
(http://ctep.cancer.gov/forms/Hndbk.pdf )
Background and Rationale
■ All protocols require a section detailing the
scientific rationale for a protocol and the
justification in medical and scientific literature
for the hypothesis being proposed.
■ Introductory section should be as succinct as
possible and should be organized in a logical,
sequential flow.
Background and Rationale
■ Double check all citations:
■ “Bibliographic inaccuracies harm the citing
author and may cast doubt on the quality of the
research being reported…”
Wyles DF, Behavioral and Social Sciences Librarian, 2004
■ “…[A]uthors should verify references against
the original document.”
Uniform Requirements for Manuscripts Submitted to Biomedical
Journals, ICMJE, 2005 (http://www.icmje.org/)
Objectives
■ Objectives should be stated clearly as
hypotheses to be tested.
■ Each objective should have a corresponding
discussion in the statistical section.
CTEP Investigators’ Handbook, 2002
(http://ctep.cancer.gov/handbook/index.html)
Writing Eligibility Criteria
STOP BEFORE YOU WRITE!
■ Eligibility criteria are the largest barrier to
accrual to clinical trials.1
■ Poorly written or poorly conceived criteria
may undermine a trial’s generalizability and
scientific validity.2
1Fuks
A, J Clin Epidemiol, 1998
2George SL, J Clin Oncol, 1996
Writing Eligibility Criteria
■ Eligibility criteria—stated as either exclusion
or inclusion criteria—define and limit the
kinds of patients that can participate in a
clinical trial.
■ Reasons for imposing eligibility criteria can
include scientific rationales, safety concerns,
regulatory issues, and practical
considerations.
George SL, J Clin Oncol, 1996
Writing Eligibility Criteria
■ Problems with restrictive criteria:
■ Limitations of generalizability
■ Failure to mimic clinical practice
■ Increased study complexity
■ Increased costs
■ Decreased patient accrual
George SL, J Clin Oncol, 1996
Writing Eligibility Criteria
■ Recommendations:
■ The number of eligibility criteria should be kept
to a minimum.
■ Criteria should include only those absolutely
necessary to ensure scientific validity and
patient safety.
■ Eligibility criteria should be clearly defined and
verifiable by an external auditor.
Adapted from:
George SL, J Clin Oncol, 1996
Fuks A, J Clin Epidemiol, 1998
Writing Eligibility Criteria
■ Eligibility criteria should be straightforward
and unambiguous. Which of these criteria
would you choose?
■ Pregnant and/or nursing women are not eligible.
■ All women of childbearing age are required to
have a serum pregnancy test.
■ Pregnant and/or nursing women are not eligible for
this study. All women of childbearing potential
(defined as…) must have a negative pregnancy
test (serum or urine) within 2 weeks of study
enrollment.
Writing Eligibility Criteria
■ However, be aware of the consequences of
highly specific criteria:
■ For example: consider the issues that will
follow from mandating a particular serum
concentration of some marker, rather than
building the definition around institutional
upper limits of normal.
Study Design
■ The study design section of the protocol
should contain a stepwise description of all
procedures required by the study.
■ A good study design section includes
sufficient information for the participating site
to develop a comprehensive clinical pathway
for study patients.
Study Design
■ Parts of the study design section may include:
■ Initial evaluations
■ Screening tests
■ Required lab tests
■ Details of treatment and ancillary procedures
■ Agent information or device specifications
■ Dose scheduling and modification
■ Calendars
Adapted from: Protomechanics: Chapter 1
(http://www.cc.nih.gov/ccc/protomechanics/)
Safety
■ The Safety (or Adverse Events) section should
include:
■ Detailed information for reporting adverse
events, including reporting to the FDA and/or
the sponsor
■ Unblinding processes (if applicable)
■ Lists of expected adverse events
Human Subjects Protection
■ This section includes discussion of:
■ Subject selection and exclusion
■ Proposed methods of patient recruitment
■ Minority representation
■ Recruitment (or exclusion) of special subjects,
including vulnerable subjects
■ Lists of potential risks and benefits, including
justification for risks
Adapted from: OHSR Information Sheet 5: Guidelines for Writing
Research Protocols (ohsr.od.nih.gov/info/sheet5.html )
Model Informed Consent
■ The Office of Human Subjects Research
recognizes 3 fundamental conditions for a
valid informed consent:
■ Disclosure of relevant information to
prospective research subjects
■ Comprehension of the information provided
■ Voluntary agreement of the subject, free from
coercion
Adapted from: OHSR Information Sheet 6: Guidelines for Writing
Informed Consents (ohsr.od.nih.gov/info/sheet6.html )
Model Informed Consent
■ To these ends, the protocol model informed
consent document must:
■ Be thorough and complete
■ Be written in simple, nontechnical language
■ Be carefully worded to avoid potentially
coercive phrasing
Model Informed Consent
■ OHSR recommends the following be included:
■ Statement that the study involves research
■ Purpose of the research and the length of the study
■ Description of risks and benefits
■ Discussion of alternative therapies
■ Confidentiality policy
■ Compensation for injury
■ Contact for further questions/information
■ Statement of voluntary participation
Adapted from: OHSR Information Sheet 6: Guidelines for Writing
Informed Consents (ohsr.od.nih.gov/info/sheet6.html )
Model Informed Consent
■ Writing a good MIC is a balancing act between
being thorough, being accurate, and being as
concise and simple as possible.
■ Patient advocates may offer invaluable
experience and insight during the drafting and
review phase of an MIC.
Correlative Studies
■ Integrating correlative studies/basic science
is a priority for the NIH:
■ “NIH Roadmap” initiative
■ Examples of correlative studies include:
■ Genomics
■ Proteomics
■ Pharmacokinetics
■ Decision-analysis studies
Correlative Studies
■ A “protocol within a protocol,” with a similar
structure listing rationale, methods, design,
and analysis plans? Or,
■ Integrated into the main body of the protocol?
The Statistical Section
■ Make sure that study objectives and study
design elements in the statistical section
mirror those in described in the Objectives
section!
■ If the study involves stopping rules, make
sure that descriptions and definitions of
toxicities in the statistical section match those
in the Safety/AE section.
Writing Clinical Research Protocols
■ General Considerations
■ Parts of the Protocol
■ Good Writing Matters
■ Looking Ahead
■ Resources and Guidance
Good Writing Matters
“…Many existing clinical trials contain
problems such as incompleteness, ambiguity,
and inconsistency. Most of the errors are
introduced during the protocol writing
process…”
Weng C, Medinfo, 2004
Good Writing Matters
■ Errors, omissions, or inaccurate or unclear
material in technical writing can potentially
result in:
■ Monetary loss
■ Damage to equipment
■ Loss of time
■ Loss of scientific value of research
■ Physical harm
Blake G and Bly R, “Fundamentals of Effective Technical Writing,”
The Elements of Technical Writing, 1993
Good Writing Matters
■ Costs of a badly written protocol?
■ Poorly written inclusion criteria have resulted in
a number of ineligible and inevaluable patients
being enrolled to a study.
NOW WHAT?
Good Writing Matters
■ To fix this problem, the protocol has to be
amended. Remember:
“Any change to the protocol document or
Informed Consent that affects the scientific
intent, study design, patient safety, or
human subject protection is considered an
amendment, and therefore must be approved
by CTEP and your IRB…”
NCI CTEP Amendment Request Submission Policy, Version Date
May 14, 2004 (http://ctep.cancer.gov/guidelines/templates.html)
Good Writing Matters
■ What are some potential costs in this
scenario?
■ Patient distress
■ Study design compromised
■ Accrual delay (possibly measured in months)
■ Administrative costs of submitting and
approving amendment (coordinating center
and sites)
■ Site goodwill
Tools for Better Writing
■ Appoint a dedicated editor to oversee the
drafting and compilation of a protocol.
■ Adopt standard references for your protocol.
■ Many eyes make accurate work: review both
among and outside of the team can make the
difference.
Tools for Better Writing:
Style Sheets
■ Establishing a list of common style elements
can help maintain uniformity and consistency
in protocols:
■ Preferred punctuation
■ Preferred citation style
■ Layout and organization of protocols
■ Templates for calendars and schedules
■ Common data elements
Tools for Better Writing: Proofreading
■ Make time for proofreading
■ Don’t make last-minute changes and submit a
document for review or approval without
subjecting it to a formal proofreading.
■ Once is not enough
■ Remember: protocols are lengthy, complex
documents. You will need more than single
review to catch all mistakes.
Tools for Better Writing: Proofreading
■ Fresh eyes
■ Working too long on a protocol may habituate
eyes and brains to mistakes, simply because
they’ve been there all along. Get an “outside”
reviewer!
■ Spell-checkers, etc.
■ A document that has been “checked” by
automatic software has NOT been proofread.
Avoiding Pitfalls: Inconsistency
■ Adopt and apply standards for:
■ Procedure names
■ Agents and devices (Remember: avoid trade
names!)
■ Abbreviations
■ Units
Avoiding Pitfalls: Ambiguity
■ Use precise language.
■ Are spans of time inclusive or exclusive?
■ Identify procedures as specifically as possible
(eg, “balloon angioplasty” rather than “heart
surgery”).
Writing Clinical Research Protocols
■ General Considerations
■ Parts of the Protocol
■ Good Writing Matters
■ Looking Ahead
■ Resources and Guidance
Looking Ahead
■ Even the best protocols may have to be
amended, due to:
■ Changes in the scientific or clinical landscape
■ Regulatory changes, including updated
procedures
■ Unforeseen clinical developments that may
affect patient safety or study validity
Looking Ahead
■ Avoid being “reactive” when amending
protocols
■ Seize the opportunity to review the entire
protocol; this “continuing review” may uncover
errors or omissions that passed through earlier
reviews.
■ Don’t just make a “point” change to a protocol;
check the entire document for ramifications.
Looking Ahead
■ Remember: maintain version control!
■ Archive EVERYTHING!
■ Control access to materials:
■ Have a defined procedure for creating or
modifying files
■ Limit write-access to files, where appropriate
Writing Clinical Research Protocols
■ General Considerations
■ Parts of the Protocol
■ Good Writing Matters
■ Looking Ahead
■ Resources and Guidance
NIH Guidance on Protocol Writing
■ Protomechanics:
http://www.cc.nih.gov/ccc/protomechanics/
■ The Office of Human Subjects Research:
http://ohsr.od.nih.gov/info/info.html
■ The NCI Investigators’ Handbook:
http://ctep.cancer.gov/handbook/index.html
Templates
■ The NIH phase III template:
http://www.ninds.nih.gov/funding/research/
clinical_research/ProtocolTemplate.htm
■ The Cancer Therapy Evaluation Program
(CTEP) Templates (phases I–III; based on NIH
model):
http://ctep.cancer.gov/guidelines/templates.html
Informed Consents
■ Office of Human Subjects Research
http://ohsr.od.nih.gov/info/info.html
■ The Office for Human Research Protections
(OHRP):
http://www.hhs.gov/ohrp/policy/index.html#informed
Writing Resources
■ The International Committee of Medical
Journal Editors (ICJME) Uniform
Requirements for Manuscripts Submitted to
Biomedical Journals
http://www.icmje.org/
Writing Resources
■ The American Medical Association style
manual may provide a useful “default”
reference for protocol writing:
Cheryl Iverson, Chair. The American Medical
Association Manual of Style. 9th ed. Lippincott
Williams and Wilkins; 1998.
■ The Duke University Medical Center Library
maintains a select bibliography of style
guides at:
http://www.mclibrary.duke.edu/subject/style
Summary
■ Writing a clinical protocol is an elaborate,
collaborative task that will require the talents
of many contributors to ensure a result that is
scientifically sound and logistically
practicable.
■ The quality of writing in a protocol and the
quality of the protocol itself are inseparable.
Questions?