Call for Papers
State Health Initiatives
Chair: Jon Christianson, University of Minnesota
Sunday, June 25 • 8:30 am – 10:00 am
● Covering the Uninsured with Medicaid HIFA Waivers
Terri Coughlin, Sharon Long, John Graves
Presented By: Terri Coughlin, Principal Research Associate,
Health Policy Center, Urban Institute, 2100 M Street, NW,
Washington DC, DC 20037; Tel: 202.261.5639; Fax:
202.223.1149; Email: tcoughli@ui.urban.org
Research Objective: To examine how states used the
Medicaid Health Insurance Flexibility and Accountability
(HIFA) waiver authority to cover uninsured individuals, a major objective of the 2001 initiative. We also examine whether the waiver programs affected state uninsurance rates.
Study Design: Using case study methodology, we studied the design and implementation of ten state HIFA waiver demonstration projects. Case studies were carried out between January 2003 and April 2005. To determine if the demonstrations had affected the number of uninsured, we examined trends in uninsurance rates using data from the
1997 to 2004 Current Population Survey (CPS) in three of the ten study states.
Population Studied: Low-income individuals
Principal Findings: States adopted varied program designs, reflecting their particular goals and circumstances. They showed a strong interest in expanding coverage to groups that historically have been excluded from publicly sponsored health programs, such as childless adults, higher income parents and the working poor. We also found that states used a mix of coverage strategies,including direct Medicaid coverage, premium assistance programs as well as developing a new commerical insurance product expressly for HIFA demonstration enrollees. Despite initial fears, states on balance did not reduce benefits for current beneficiaries to help finance coverage expansions implemented as part of
HIFA. Nationally, we estimate that 300,000 individuals were covered under HIFA demonstrations as of spring 2005. Trend analysis using the CPS gave mixed results, with two states showing a slight drop in uninsurance rates after the demonstration was implemented, and one state showing a rise.
Conclusions: Although facing very tough fiscal climates, many states had the political will to expand coverage to lowincome individuals. Despite this, the HIFA initiative has fallen far short of its expected coverage goals.
Implications for Policy, Delivery, or Practice: The HIFA initiative has introduced considerable change in the health care landscape of many of the states we studied. The impact of these many important changes (such as imposing higher cost-sharing, adequacy of protection in premium assistance programs and the like) warrants rigorous study as well as close monitoring by policymakers. This is especially true given that several of the new generation of state Medicaid waivers
(for example, Florida and Vermont) build on several HIFA program features.
Primary Funding Source: CMS
● Using HMOs to Serve the Medicaid Population: What
Are the Effects on Expenditures and Does the Type of
HMO Matter?
Bradley Herring, Ph.D., E. Kathleen Adams, Ph.D.
Presented By: Bradley Herring, Ph.D., Assistant Professor,
Health Policy and Management, Emory University's Rollins
School of Public Health, 1518 Clifton Road NE, Atlanta, GA
30322; Tel: 404-727-3738; Fax: 404-727-9198;
Email: bradley.herring@emory.edu
Research Objective: States have increasingly used HMOs to provide medical services to the Medicaid population.
However, the effects of these initiatives on service utilization and total costs are unclear. Moreover, while some states have primarily used commercial HMOs which have both Medicaid and private enrollees, other states have relied on Medicaiddominated HMOs. We examine how changes in Medicaid
HMO penetration and its composition affected service use and expenses.
Study Design: We construct market measures of Medicaid
HMO penetration for both commercial and Medicaiddominated HMOs using data from CMS and InterStudy for years 1996, 1998, 2000, and 2002. We use these data with household survey data from the Community Tracking Study.
We test for the effect of HMO penetration on costs by generating “synthetic” expenditures for the CTS from “unit prices” based on MEPS data. We also analyze the effect on the mix of services by examining eight different measures of utilization (e.g., number of doctor visits, inpatient stays) and four measures of access. We use market-level fixed effects to focus on changes in penetration over time.
Population Studied: We use a sample of 8,425 non-elderly individuals covered by Medicaid, representative of the urban
United States. We also examine samples of children and adults separately.
Principal Findings: We find that total health expenditures for the Medicaid population are positively related to the penetration rate of Medicaid-dominated HMOs while unrelated to the penetration rate of commercial HMOs. We also find that the number of visits to a nurse practitioner, physician assistant, or midwife, the number of ER visits, and the number of inpatient nights are positively related to increases in the penetration of Medicaid-dominated HMOs, while the number of outpatient surgeries is negatively related to this measure. We also find that the odds of having a usual source of care other than the ER are negatively related to increases in Medicaid-dominated HMOs. The effects from
Medicaid-dominated HMOs are found largely for children.
We find that all measures of utilization (except mental health visits) and access to care are unrelated to commercial HMOs.
Conclusions: Increases in Medicaid-dominated HMOs over time led to increases in total health expenditures and reductions in access. Increased costs, at least for children, are related in part to increased ER use. Conversely, increases in the use of commercial HMOs over time did not lead to either changes in health expenditures, changes in the mix of medical services, or changes in access to care.

Implications for Policy, Delivery, or Practice: States policymakers that have contracted, or plan to contract, with
HMOs to provide medical services to their Medicaid enrollees should be made aware of the effects that this change has had on total expenditures and access to care of the Medicaid population. Our results suggest that the use of Medicaiddominated HMOs has not met either the goal of increased access or decreased expenditures. It appears that states wanting to contract with HMOs can achieve relatively better outcomes for their Medicaid populations by using commercial
HMOs. This will likely require close attention to the level of payment rates in order to attract/retain commercial plans.
Primary Funding Source: RWJF
● Shaping High Performance Health Systems: Innovative
State Policies and Practices
Catherine Hess, M.S.W., Alan Weil, J.D., M.P.P., Sonya
Schwartz, J.D.
Presented By: Catherine Hess, MSW, Senior Program
Director, National Academy for State Health Policy, 1233 2oth
St. NW, Washington, DC 20036; Tel: 202-903-0101; Fax: 202-
903-2790; Email: chess@nashp.org
Research Objective: 1. To identify state policies and practices contributing to high performing health systems, consistent with the Commonwealth Fund's Commission on a High
Performance Health System. 2. To assess the extent to which states have adopted important policies and practices contributing to high performance health systems. 3. To describe innovative state policies and practices contributing to high performance health systems.
Study Design: Major domains and criteria for selecting state policies and practices for study were developed in relation to the work of the Commonwealth Fund's Commission on a
High Performance Health System. Literature reviews were conducted in fall 2005 to identify: similar relevant studies assessing state policy performance; evidence and recommendations for state health policy and practice in domains of interest; and sources of recent information collection for all states on relevant health policies and practices. Based on the literature review and advice from a subgroup of Commission members, domains were finalized and policies and practices for study selected. These are:
1. Coverage of Essential Benefits, with the goal that all people have equitable and affordable health coverage; includes state polices and practices addressing: • Public coverage beyond federal minimums • Private sector strategies such as market regulation • Enrollment and retention promoting continuous coverage 2. Quality, Efficiency and Value, with the goal that all people get the right care at reasonable cost, and get equitable care that is, safe, patient-centered, and coordinated; includes state policies and practices addressing: • Value purchasing, including performance incentives and disincentives • Public reporting of quality indicators
•Quality forums and agendas 3. Health Systems
Infrastructure, with all people having access to systems of personal health care and population-based public health services that promote long and healthy lives and that have capacity to improve; includes state policies and practices addressing: • Public health systems • Consumer health information and related technology • Provider capacity, including safety net services Information for the study is drawn from recent state policy surveys; new surveys of multiple state officials; and follow-up interviews on potentially innovative policies and practices. Results will be analyzed to sort state performance into categories. Results will be available well before the meeting in June.
Population Studied: This study focuses on health policies of the fifty United States and the District of Columbia.
Principal Findings: The study will compile, present, categorize and discuss state performance on multiple elements reflective of important state roles in shaping high performing health systems in the three domains identified above. The study also will identify and describe trendsetting, innovative approaches that states are taking to improve health systems performance.
Conclusions: This ambitious study addresses state roles in promoting high performance in health systems, roles that necessarily cross agency and program lines. The study’s conclusions will provide a “big picture” view of the impact states are having and the potential for states to do more to advance a high performance health systems agenda.
Implications for Policy, Delivery, or Practice: The study can be utilized by states and others to develop, refine and advance policy agendas and to implement practices to promote high performing health systems.
Primary Funding Source: CWF
● Enrollment Patterns, Service Use and Earnings Trends for Participants in the Medicaid Buy-In Program: A
Descriptive Profile from Linked Multi-Agency Data
Su Liu, Ph.D., Henry T. Ireys, Ph.D., Stephen F. Knapp
Presented By: Su Liu, Ph.D., Researcher, , Mathematica Policy
Research, Inc., 600 Maryland Ave., SW, Suite 550,
Washington, DC 20024; Tel: (202) 264-3499; Fax: (202) 863-
1763; Email: sliu@mathematica-mpr.com
Research Objective: Adults with disabilities face many employment barriers, one of which is the concern that working will result in the loss of public health insurance. The Medicaid
Buy-In program, offered by states as an option to the regular
Medicaid program since late 1990s, is designed to encourage individuals with disabilities to work by allowing them to “buy into” the Medicaid program when higher earnings would normally make them ineligible for Medicaid benefits. This study has two objectives. First, to provide a descriptive profile of enrollment patterns, service use and earnings trends of the individuals participated in 27 states’ Medicaid Buy-In programs. Second, to test the feasibility of linking administrative data from multiple federal agencies and building a comprehensive system that can be used to monitor health, health care and employment of individuals with disabilities overall.
Study Design: We used individual-level Buy-In enrollment data submitted by states to link Medicare and Medicaid eligibility/claims data from CMS with annual earnings and
SSDI/SSI benefits data from SSA, under an inter-agency data user agreement. The linked data included information previously unavailable with a single data source, such as service use in both Medicare and Medicaid program (many
Buy-In participants are dual eligibles), type of impairment, receipt of SSDI/SSI benefits, and earnings. Descriptive statistics of variables in the linked data were provided, as well as cross-tabulations of multiple variables. Comparisons of key

variables, such as earnings, were done for pre- and postenrollment periods. Analyses were conducted for all or subgroups of Buy-In participants, and at both national and state level.
Population Studied: 126,606 individuals with disabilities who participated in the Medicaid Buy-In programs between 2000 and 2004 in 27 states (out of 32 states that had a Buy-In program as of December 2004). The sample for a specific analysis may be restricted to individuals we were able to match to one or more of the data sources used. Subgroups of
Buy-In participants may be selected to answer a specific question, for example, what were the characteristics of Buy-In participants who had the highest earnings in a year?
Principal Findings: Considerable variations across states were found throughout the study, likely the product of interactions between state program features and contextual factors that are sometimes subtle and often complex. On average, about 85 percent of Buy-In participants receive SSDI benefits, and slightly more than one-third have some form of mental illness or other mental disorder. But more importantly, our findings provide new insights into trends in the earnings of Buy-In participants and patterns of medical expenditures for different subgroups of participants—insights that would have been impossible without having combined data from several sources. For example, we now know that about one-third of those who enrolled in the Medicaid Buy-In program in 2001 or 2002 had higher earnings after enrollment than before. Participants with mental retardation had the highest per member per month (PMPM) Medicaid expenditure but the lowest PMPM Medicare expenditures compared with participants who had other types of impairments. This study also proves the feasibility and value of linked multi-agency data.
Conclusions: People do work despite disabilities. The
Medicaid Buy-In program may help bring down the barrier for individuals with disabilities to sustain employment without worrying about losing health care coverage, but participants with certain characteristics may have benefited more from the program than others. Further analysis is needed to understand the specific medical services (for example, personal care services or psychiatric medications) participants are using, and assess the program’s impact.
Implications for Policy, Delivery, or Practice: States may consider alternative program designs to better serve the
Medicaid Buy-In participants. The process used to generate the data used for this study could be adapted for a broader population, and could serve potentially as a mechanism for tracking disability, health care, and employment for other working-age adults with disabilities. Such a database would provide researchers with a useful tool for examining a range of critical policy questions, and support state efforts to better serve the persons with disabilities.
Primary Funding Source: CMS
● Patient Experiences and Utilization in Medicaid Provider-
Sponsored Organizations
K. Cameron Schiller, MS, Allyson Hall, Ph.D., Robert Weech-
Maldonado, Ph.D.
Presented By: K. Cameron Schiller, MS, Research Assistant,
Health Services Research, Management, and Policy, University of Florida, P. O. Box 100195, Gainesville, FL 32610-0195; Tel:
352-273-5932; Email: schiller@ufl.edu
Research Objective: Florida Medicaid is moving away from primary care case management (PCCM) in favor of Provider
Sponsored-Organizations (PSOs) to provide care to enrollees.
Using the Andersen Behavioral Model, this study evaluated patient experiences and utilization in three PSO pilot projects in Florida Medicaid: an administrative services organization
(ASO), a management services organization (MSO), and a provider sponsored network (PSN).
Study Design: A quasi-experimental design was used to compare differences in Consumer Assessments of Healthcare
Providers and Systems (CAHPS) reports, emergency room
(ER) visits, and physician visits in three PSOs to PCCM.
Ordinary Least Squares regression adjusting for case mix variables was used to test differences in reports and utilization.
Population Studied: A random sample of adult Medicaid participants was contacted in 2005 using computer assisted telephone interview to complete a survey that included the
CAHPS instrument. There were 1882 surveys completed: 374 in the ASO, 376 in the MSO, 767 in PCCM, and 365 in the
PSN. Most respondents were female (70%). Median age was
45 – 54 years old and median education was high school.
Principal Findings: Preliminary results show that patients in the ASO reported significantly better provider communication
(b=2.38, p=0.047) but significantly worse customer service
(b=-8.63, p=0.038) than PCCM. Patients in the PSN reported significantly lower scores for getting care quickly than those in
PCCM (b=-6.79, p=0.007). There were no significant differences in CAHPS reports between the MSO and the
PCCM. Compared to PCCM, patients in the ASO (b=-0.34, p=0.028) had significantly fewer physician visits but patients in the MSO (b=-0.18, p=0.232) and PSN (b=-0.28, p=0.053) did not. There were no significant differences between the
PSOs and PCCM in terms of ER visits.
Conclusions: Results suggest that patient experiences and utilization vary by PSO model. In the ASO, provider communication is better while physician use is lower than in the PCCM. In the PSN, a hospital-based organization, patients have lower reports for timeliness of care than in the PCCM.
Patient experiences and utilization were no different from
PCCM in the MSO. Finally, compared to the PCCM, the PSOs did not affect ER utilization.
Implications for Policy, Delivery, or Practice: The results of this study will give insights into the impact of these changes on consumer experiences and help inform current Florida
Medicaid reform. Further research is needed to examine the organizational factors that may account for differences in patient experiences and utilization among the PSO models.
Primary Funding Source: Florida Agency for Health Care
Administration

State Health Initiatives
Poster Session A
Sunday, June 25 • 2:00 pm – 3:30 pm
● Identifying Outling Prescription Drug Use for Quality and
Cost Containment
Charles Betley, MA, David Dorsk, Lorin Ranbom
Presented By: Charles Betley, MA, Health Policy Supervisor,
Bureau of Health Plan Policy, Ohio Health Plans, 30 E Broad
Street 27th Floor, Columbus, OH 43215; Tel: 614 466 6420;
Email: betlec@odjfs.state.oh.us
Research Objective: To develop a model to predict prescription drug spending by non-institutional and institutional Medicaid participants, using a two-part regression model to adjust for health status, demographic, and program characteristics, and identify participants who exceed expected spending by standard outlier thresholds for further interventions.
Study Design: A two-step Heckman selection process (1st part probit, 2nd part Heteroskedasticity-robust ordinary least squares) was used to model prescription drug spending, in order to account for the skewed distribution of the dependent variable and latent selection processes. Independent variables included standard demographic characteristics (age, gender, ethnicity, program eligibility) along with measures of health status (episodes of chronic or acute conditions, specific clinical conditions such as diabetes and hypertension, and
Diagnostic Cost Group risk adjustment scores).
Persons whose actual spending exceeded the 95 percent confidence interval of spending predicted by the model were identified for further investigation of their drug spending patterns, and the findings were used to refine the prediction model.
Population Studied: The research team drew a 3-percent random sample of non-institutional Medicaid recipients and a
10-percent sample of institutionalized recipients from administrative databases, and examined drug claims along with clinical services use.
Principal Findings: Variables measuring health status availble through administrative data significantly improved the predictive power of models to identify likely levels of prescription drug spending. Age, gender, race and ethnicity, and urban and rural location also explained differences in spending despite health status adjustments, and may reflect differences in recipients' use and access to care under the
Medicaid system.
Conclusions: Standard statistical techniques explaining variation in drug spending can be used to identify a manageable number of program participants whose outlying behavior could reflect quality concerns, possible fraudulent use, and potential gaps in program performance.
Implications for Policy, Delivery, or Practice: Standard, broadly accepted health services research techniques can be used by providers and insuring organizations to create focused interventions for anomolous health care use.
Primary Funding Source: No Funding
● Lessons Learned: Economic Evaluation of Home Visiting
(HV) Programs for High Risk, Pregnant, First Time
Mothers and Babies
Jane Bolin, Ph.D, J.D., B.S.N., Linnae Hutchison, M.B.A.
Presented By: Jane Bolin, Ph.D, J.D., B.S.N., Assistant
Professor, Health Policy & Management, Texas A&M HSC
School of Rural Public Health, TAMU Box 1266, College
Station, TX 77842; Tel: (979) 862-4238; Fax: (979) 862-8371;
Email: jbolin@srph.tamhsc.edu
Research Objective: Home visitation (HV) services have been an effective and valuable public health strategy for helping improve health status and reducing maternal and child morbidity and mortality with first time, high risk mothers. However, where multiple programs are operating, cost effectiveness assessment may provide valuable information for funders. Our goal is to conduct a costeffectiveness assessment of three HV programs serving highrisk, low income mothers and babies.
Study Design: The study was a multi-phase evaluation of HV programs, including on-site interviews, a survey of staff, and data analyses of outcomes in order to conduct a costeffectiveness analysis and assess the relative “success” of the three independent HV programs.
Population Studied: High risk, low-income, first time mothers and babies.
Principal Findings: Differences in HV program approaches to providing HV services created differences in health inputs and cost inputs, and limited capacity to conduct program model comparisons. While each HV program model demonstrated varying success in achieving goals, some program models appeared able to do so more cost-efficiently. Differences in program data gathering and differences in variable parameters made comparisons problematic.
Conclusions: Economic evaluations of public health interventions may be impeded by programmatic differences in data gathering as well as differences in implementation of program objectives.
Implications for Policy, Delivery, or Practice: We make recommendations for state agencies wishing to conduct economic evaluations and cost comparisons of public health programs in order to conduct more reliable cost comparisons.
These recommendations include baseline standards for outcomes and data gathering by key stakeholders.
Primary Funding Source: Missouri DHSS
● Publicly Reporting Hospital Outcome Measures: The
Oregon Experience
Tina Edlund, MS, Janne Boone, M.P.H.
Presented By: Tina Edlund, MS, Research and Data Manager,
Office for Oregon Health Policy and Research, 255 Capitol
Street NE, 5th floor, Salem, OR 97310; Tel: 503-378-2422;
Fax: 503-378-5511; Email: tina.d.edlund@state.or.us
Research Objective: A critical issue in the drive to improve healthcare quality has been a move to increase the transparency of hospital outcomes. The handful of states that have made efforts to publish outcome data have often found their efforts thwarted or data obfuscated. In contrast, after extensive efforts to create buy-in among the hospital and

physician communities, the state of Oregon recently began publishing 12 risk-adjusted outcome measures for all hospitals in the state. The objective of this study was to qualitatively document the process of creating a mechanism to publicly report hospital-level outcome measures and to report outcomes for conditions and procedures where quality has been associated with variation in mortality rates.
Study Design: This study used mixed (quantitative and qualitative) methods.
Population Studied: We used focus groups with hospital executives and physician groups to identify concerns around publishing outcome measures. These focus groups were followed up with key informant interviews. We used the
Oregon 2004 inpatient hospital discharge data (n=365,000).
Principal Findings: Initial opposition to publishing measures was gradually overcome through a series of focus groups held by state officials. Four themes emerged as powerful influences. First, opposition was softened when it was clear that high-ranking individuals in state government had resolved to publish outcome measures, but that they were inviting input from the medical community. Second, an open and public process engaging stakeholders had the effect of motivating key players from initial opposition to strategic cooperation. Third, the use of evidence-based indicators as developed by the Agency for Healthcare Quality and Research
(AHRQ) created further buy-in. Fourth, the resolution to publish the data was coupled with a stance of humility and acknowledgement of the shortcomings of the data and risk adjustment methods, along with hospital-submitted details about special circumstances and other issues regarding specific outcomes. Once published, hospital leadership acknowledged efforts on two fronts: first, in establishing systematic review of their coding practices in areas where the hospital showed significant differences from state averages, and second, in initiating focused quality improvement activities in specific areas.
Conclusions: Successful public reporting of comparative hospital quality data requires committed leadership, a strong evidence-based process and a collaborative, public partnership with the stakeholder community.
Implications for Policy, Delivery, or Practice: Only a handful of states currently publish comparative hospital quality data, partly because of long-standing opposition from the hospital community and partly because of the expense of data collection and technical statistical issues in developing and reporting the data. The Oregon experience shows that hospital concerns can be mitigated through public collaboration. As hospitals drill down into the drivers of their results, efforts to improve coding of hospital discharge data and specific quality improvement activities are instigated.
Primary Funding Source: Other Government
● An Asessment of Acute Stroke Management in Nebraksa
Hospitals
Katherine Jones, P.T., Ph.D., Michelle Mason, M.A., Liyan Xu,
M.S., Michael Shambaugh-Miller, Ph.D.
Presented By: Katherine Jones, P.T., Ph.D., Assistant
Professor, Preventive and Societal Medicine, Universityof
Nebraska Medical Center, 984350 Nebraska Medical Center,
Omaha, NE 68198-4350; Tel: (402) 559-8913; Fax: (402) 559-
7259; Email: kjonesj@unmc.edu
Research Objective: To inform state policymakers about the structures and processes (infrastructure) within Nebraska hospitals used to treat acute stroke and the extent to which they are consistent with the guidelines developed by the Brain
Attack Coalition and the American Stroke Association.
Study Design: We used the Dillman four contact method to conduct a mailed survey of those providers most knowledgeable about the process of acute stroke treatment within each hospital. We reviewed the literature and recruited an advisory panel from seven representative hospitals and interested state agencies to create an instrument organized in domains reflecting the elements of a primary stroke center as described by the Brain Attack Coalition. We collected additional information about risk reduction activities, perceived barriers to delivering thrombolytic therapy, and attitudes towards a statewide stroke system.
Population Studied: We surveyed the 83 hospitals in
Nebraska that provide treatment for acute stroke: 60 Critical
Access Hospitals (CAHs) with 25 or fewer beds, 7 small rural hospitals (SRHs) with 26 to 49 beds, and 16 larger hospitals with 94 or more beds.
Principal Findings: Of the 83 hospitals, two larger hospitals reported having a formal acute stroke team. Approximately
15% of CAHs and SRHs compared to 40% of larger hospitals reported having the infrastructure required to diagnose ischemic vs. hemorrhagic stroke in a timely manner. One-third of CAHs, 43% of SRHs, and 100% of larger hospitals reported giving thrombolytic therapy at least once in the previous year.
Of these hospitals, one-half to two-thirds consistently followed the guidelines for this therapy. Considerable proportions of all hospitals reported inadequate treatment protocols, lack of neurological diagnostic support, and apprehension about adverse outcomes and litigation as barriers to thrombolytic therapy—CAHs were more likely to report absence of a CT scanner and lack of radiology support as barriers. Over threefourths of all hospitals agreed that public education of stroke as a medical emergency, EMS education regarding signs and symptoms of stroke, transport protocols, and accessibility to diagnostic infrastructure should be coordinated at the state level.
Conclusions: Nebraskans served by CAHs are at greatest risk of having limited access to thrombolytic therapy for stroke due to lack of written treatment protocols, timely access to a CT scanner, and review of findings by a radiologist and neurologist. Hospitals of all sizes can benefit from education regarding appropriate use of thrombolytic therapy, written treatment protocols, resources for public education regarding stroke as a medical emergency, and utilization of Nebraska’s telehealth network for diagnostic support.
Implications for Policy, Delivery, or Practice: Policymakers in rural states must assess the infrastructure for acute stroke treatment to identify disparities in access to adequate care.

This assessment must include the locations of facilities capable of treating stroke according to evidence-based guidelines. The goal of the assessment should be to maximize access to appropriate stroke care through the establishment of systems of care coordinated at the state and local level and including the public, EMS providers, and hospitals. Integrated systems of care for the treatment of stroke are necessary in rural states where large proportions of the population are served by Critical Access Hospitals.
Primary Funding Source: State of Nebraska Department of
Health and Human Services System
● Asthma Measurement over Three State Collaboratives
Constance Martin, MS, M.P.H., Jane Deane Clark, Ph.D., Purvi
Kobawala Smith, MS, M.P.H., Karl Weimer, M.P.H., Pat Barta,
M.P.H., Joanie Reller, M.P.A.
Presented By: Constance Martin, MS, M.P.H., Program
Officer, Measurement and Evaluation, Center for Health Care
Strategies, PO Box 3469, Princeton, NJ 08543; Tel: 609-895-
8101; Fax: 609-895-9648; Email: cmartin@chcs.org
Research Objective: Collecting and disseminating comparative data on the impact of tested clinical and administrative practices is a critical step in establishing evidence based practices in asthma care for Medicaid enrollees. At present, there is only one HEDIS measure related to asthma and not all Medicaid managed care organizations collect and report on this measure. Through the establishment of a uniform measurement strategy in three regional State Collaborative Workgroups, The Center for
Health Care Strategies (CHCS) seeks to create, collect and disseminate much needed comparative data on an expanded and more comprehensive set of Common Measures for asthma care and asthma related health outcomes.
Study Design: CHCS has convened three regional State
Collaborative Workgroups in California, New York and
Indiana, with the overall goal of developing, piloting, refining and sustaining clinical and administrative practices to improve asthma care and health outcomes for Medicaid enrollees. MCO participants in the first of the three State
Collaboratives underwent a rigorous process to establish a set of Common Measures. These measures were created to provide comparative data for each of the teams; by team and in the aggregate. At each step in the process, ideas were solicited and consensus was reached. The end result included measures for: • Identification of members with asthma
• Inpatient admissions and days • Emergency department visits • Appropriate use of controller medications
In California, 10 of 11 participating health plans submitted
Common Measures data for three consecutive years. These data were analyzed to evaluate the overall impact on asthma care and health outcomes. In New York, with the aid of the state Medicaid agency, Common Measures from all 12 health plans were submitted to CHCS and are currently being analyzed. In Indiana, 5 health plans have agreed to collect and share the same set of Common Measures.
Population Studied: Medicaid managed care organizations, their provider networks, and the Medicaid enrollees they serve.
Principal Findings: Common Measures allow for evaluation of the Collaborative as a whole and for comparability among and across participating health plans and states. Having access to these data also allow health plans to develop informed strategies to target resources to improve quality of asthma care. The findings from each State Collaborative will be shared unblinded within each collaborative and blinded among other collaborative participants. Furthermore, CHCS will share these data more widely through development of a
Toolkit and other publications for national dissemination.
Conclusions: By collecting Common Measures before and after efforts made by these State Collaborative Workgroups,
CHCS is positioned to generate and disseminate unprecedented comparative data on the impact of quality improvement initiatives in Medicaid managed care across the country. Lessons from the three State Collaborative
Workgroups in California, New York and Indiana will be broadly applicable to other quality improvement initiatives, and other states.
Implications for Policy, Delivery, or Practice: •Enhancing asthma-related member benefits and quality of life for people with asthma. • Standardizing promotion of uniform practice guidelines in competing/regional marketplace. • States working with health plans to standardize asthma guidelines.
Primary Funding Source: RWJF
● Outcomes of Quality care in Free Medical Clinic
Martha Morrow, Ph.D., FNP-C
Presented By: Martha Morrow, Ph.D., FNP-C, Associate
Professor, Nursing, Shenandoah University, 1775 N. Sector Ct,
Winchester, VA 22601; Tel: 540-665-5508; Fax: 540-665-5519;
Email: mmorrow@su.edu
Research Objective: Identify quality outcomes of providing consistent care in a Free Medical Clinic with nurse practitioners
Study Design: Retrospective chart review of sample patients at two Free Medical Clinics.Charts were selected according to the defined inclusion criteria. They were manually reviewed and the following information was documented on the initial and current dates of service by the nurse practitioner; systolic and diastolic blood pressure, weight, tobacco use, lipid panel, risk for cardiac disease, antihypertensive medication classifications, use of lipid lowering drugs, HgA1c, antidiabetic medication. In addition, demographic information on age and gender were recorded. In addition, a list of participants who met the inclusion criteria for the study was compiled and a computerized retrospective chart review was performed in medical records at the regional medical center. Data was collected for a time period of two years prior to the patient seeking services of the FC up to the time of the study. Data included number of hospital admissions, hospitalized days, and emergency room visits. The length of time selected was done to calculate an annual average number of visits and days to be used in data analysis instead of actual visits and days in a given year.
Population Studied: The population consisted of participants in two free medical clinics serving the medically indigent.
These people were working poor at or below 150% of the povery income level who did not qualify for or offered medical insurance. The location of the free clinics was a rural community served by a single regional medical center and one small local community hospital. Patients were selected for inclusion in the study who met the following criteria: current patients of the FC, who were being seen consistently by the
Nurse Practitioner for at least the previous year. “Seen

consistently” was defined as being seen at least three time per year at the FC for chronic disease. In addition, the FC Nurse
Practitioner was also designated as the patient’s primary care provider and the patients were not being followed by an outside primary care provider.
Principal Findings: The study found significant differences in both community indicators and individual outcomes after this population began receiving care at the Free Clinic where their primary practitioner was a nurse practitioner. Hospital admissions, hospital days, and emergency room visits were reduced. There were significant decreases in systolic blood pressure, diastolic blood pressure, HgA1c, LDL levels, and risk for heart disease. Other documented decreases included cigarette smoking, weight, total cholesterol, and triglycerides, and an increase in HDL levels.
Conclusions: This study indicates that consistent high quality care provides benefits to community as well as control of disease and health maintenance.
Implications for Policy, Delivery, or Practice: Implications for practice- consistent quality care provides improved control of disease, improved health maintenance, decrease inappropriate use of medical resources, and a quality health care safety net for medically indigent.
Primary Funding Source: No Funding
● A Critical Case Study of Program Fidelity in TennCare
Carole Myers, Ph.D., MSN, BS
Presented By: Carole Myers, Ph.D., MSN, BS, Assistant
Professor, College of Nursing, University of Tennessee, 1200
Volunteer Blvd., Knoxville, TN 337996; Tel: 865-414-7218; Fax:
865-974-3569; Email: cmyers9@utk.edu
Research Objective: The purpose of this study was to evaluate the fidelity of the design of Tennessee’s Medicaid managed care program in comparison to the actual program operation.
Study Design: The retrospective recollections of 26 informants, including two former Governors of Tennessee, a former HCFA Administrator, state government and managed care executives, advocates and provider representatives, captured in recorded interviews, were triangulated with timestable documents in this single case study design to reveal an intricate web of people and circumstances that have shaped the program.
Population Studied: The TennCare program was the focus of this study.
Principal Findings: Thematic analysis illuminated both promises and failures of TennCare. Three themes were prominent: authority, management and fragmentation.
Governor McWherter, TennCare’s creator, established strong executive authority to model and implement TennCare; a void was created when he left office. Subsequent administrations have not transitioned to a more balanced and inclusive authoritative structure, nor have they developed an adequate oversight model. Continued mismanagement of the administration of benefits and failure to meet established care standards set the stage for the imposition of federal judiciary authority. Management of TennCare operations has been reactionary and politicized and, in many instances, inappropriately abdicated or conferred upon the wrong or unprepared people or entities. Turmoil in state government hindered stabilization of the program. Evolution of the marketplace has not been broadly realized; the state has retreated from basic managed care principles. The state failed to integrate the management of TennCare carve-outs. This fragmentation resulted in diffuse management accountability, duplication of services, gaps in patient care and patient hassle. The web of connectivity between themes changed over time, with themes presenting as causes, catalysts or consequences of the others at different times. A cardinal example of this is how the mismanagement of TennCare program led to the breakdown in key alliances and the imposition of consent decrees resulting in reactive and disjointed management which significantly contributed to the gap between what was envisioned for TennCare and what actually resulted.
Conclusions: Although TennCare has been successful in increasing access to health care coverage, this success has been overshadowed by pervasive problems, a failure to implement basic building blocks of managed care, such as risk-sharing and competition, and effectively manage employed vendors.
Implications for Policy, Delivery, or Practice: In the span of only 15 years, the introduction of managed care and other market-based strategies from the private sector precipitated a transformation of the delivery of Medicaid services in the
United States. These monumental changes remain poorly understood. The implementation of managed care in
Tennessee’s Medicaid program is an excellent public policy exemplar because of the far-reaching scope of the program and the ongoing development of the program. The themes illuminated by this study will be informative to planners of similar state initiatives.
Primary Funding Source: Sigma Theta Tau
● Massachusetts Uncompensated Care Pool High-cost
Users
Beth Perry, M.P.H.
Presented By: Beth Perry, M.P.H., Data Analyst,
Massachusetts Division of Health Care Finance and Policy, 2
Boylston Street, Boston, MA 02116; Tel: (617) 988-3265;
Email: beth.perry@hcf.state.ma.us
Research Objective: The Massachusetts Uncompensated
Care Pool (UCP) makes payments to hospitals and community health centers for some of the costs of the care for the uninsured and underinsured in the Commonwealth. Over the past several years, the Massachusetts Division of Health
Care Finance and Policy (DHCFP) has developed a sophisticated patient-level database that contains demographic, clinical, and charge information for individuals whose medical care is reimbursed by the UCP. One component of DHCFP’s ongoing analysis of UCP utilization has been to focus on those with very high yearly costs to the
Pool.
Study Design: Quantitative and qualitative analyses were performed to describe the characteristics of the high-cost UCP users, those with yearly costs in the top 1% of Pool users.
Eligibility data was analyzed for demographic traits of these users, including age, gender, and income. Claims data was analyzed for specific utilization of inpatient and outpatient hospital services. Cost data was used to analyze the impact of these users on the overall cost of the UCP. This UCP data was

then compared with other published data on high-cost users of medical services.
Population Studied: Pool Fiscal Year 2005 (PFY05; October
2004 – September 2005) top 1% high-cost UCP users (4,515 individuals).
Principal Findings: In PFY05, 4,515 high-cost UCP users generated over $162 million in costs. This represents 25% of total yearly UCP costs. Inpatient services for the high-cost users accounted for 51% of all PFY05 inpatient costs to the
Pool; 71% of high-cost users’ costs were spent on inpatient services, while only 29% of costs were reimbursements for outpatient services. This skew in costs is more dramatic than in other populations. Demographically, high-cost users were more likely to be older than other Pool users, but under the age of 65. Clinically, the 4,515 high-cost users had 7,060 hospital discharges in PFY05; representing 17% of all PFY05 inpatient UCP discharges. High-cost users had longer lengths of stay, received more significant procedures, and had more complex diagnoses (cardiac, oncology) than other UCP users.
However, utilization for mental health and substance abuse inpatient services was similar between high-cost and non high-cost users.
Conclusions: The majority of the high-cost UCP users received ongoing, regular medical attention for chronic, progressive illnesses during the year. These users were not
‘catastrophic’ cases of individuals who required immediate but isolated medical care. However, most high-cost users were not high-cost users for more than one year, potentially indicating deferred care prior to a significant diagnosis.
Implications for Policy, Delivery, or Practice: High-cost users within a safety net program may be significant costdrivers within the system. While many of these high-cost users may use the system as insurance, the lack of reimbursement for some services (specifically rehabilitation and sub-acute care), and the lack of disease management programs may inhibit cost savings and lead to increased utilization. The
DHCFP has initiated a pilot program to explore whether some high-cost UCP users may be eligible for Medicaid or private insurance, and facilitate their enrollment into insurance.
Primary Funding Source: DHCFP
● Migrant and Seasonal Farmworkers: Health Insurance
Coverage and Access to Care
Peter Shin, Ph.D., M.P.H., Sara Rosenbaum, J.D.
Presented By: Peter Shin, Ph.D., M.P.H., Assistant Research
Professor, Department of Health Policy, George Washington
University, 2021 K St., NW Suite 800, Washington, DC 20006;
Tel: 202-530-2313; Fax: 202-296-0025; Email: pshin@gwu.edu
Research Objective: Migrant and seasonal farmworkers
(MSFW) and their families confront health challenges stemming from the nature of their work, their extreme poverty and mobility, and living and working arrangements that impede access to health coverage and care. This analysis provides an overview of MSFW and the health challenges they face and considers options for improving the health coverage and access to care for a highly mobile working population.
Study Design: The information used to develop this analysis comes from a review of the literature and state programs as well as two data sources: the 2000 National Agricultural
Worker Survey (NAWS), a national survey of farmworkers conducted by the United States Department of Labor; and the
Uniform Data System (UDS) maintained on federally funded health centers by the United States Department of Health and
Human Services, Health Resources and Services
Administration. The NAWS analysis is based on a sub-sample of approximately 1400 workers who identified themselves as migrant farmworkers. Experts in migrant health care view
NAWS as providing an accurate portrait of MSFW for purposes of designing health care services. The UDS identifies MSFWs, including their dependents. In 2002, 372 out of 843 federally funded health centers (44%) served
708,611 persons identified as migrant and season workers and family members. Among this broader group of grantees receiving both general and migrant health center funding, 15 health centers received MSFW grants exclusively and served a total of 58,350 patients. Although other data sources exist for describing the migrant labor force, only these two data sets provide information on migrant health status and health care access.
Population Studied: Approximately three to five million workers earn their living through migrant or seasonal farm labor. The NAWS data exclude secretaries and mechanics, as well as workers who are non-immigrants working in the U.S. under a temporary visa issued pursuant to the Immigration and Nationality Act.
Principal Findings: • Almost all migrant and seasonal farmworkers are foreign-born with only 6% reporting being born in the United States. The majority (70%) permanently reside in the United States. Although concentrated in certain areas of the country, MSFW reside in all states. They travel frequently between states for their employment. • MSFW face significant language barriers—about 9 in 10 say they read and speak little or no English. • Even though MSFW report working five to six days a week, they are extremely poor. In 2000, the median income for MSFW was $6,250, compared to $42,000 for U.S. workers overall. • MSFW and their families are overwhelmingly uninsured. In 2000, 85% of MSFW were uninsured, compared to 37% of low-income adults nationally.
• Nine in ten children in MSFW families were uninsured compared to less than a quarter (22%) of low-income children nationally. * Reflecting their low levels of coverage, MSFW and their families use very little health care compared to other lowincome people. In 2000, only 20% of migrant and seasonal farmworkers reported using any health care services in the preceding two years.* Fifteen health centers were funded exclusively with migrant grant funding and tended to be far smaller than other health centers, and, because the overwhelming majority of their patients are uninsured (92%), their revenues primarily come through their federal grants
(80%). Centers that receive migrant funding do differ from other centers in that they often offer services tailored to
MSFW families’ needs, such as outstationed services.
• Over the years, a few states, including Wisconsin and Texas, have attempted to improve Medicaid’s ability to serve farmworkers. From these state experiences, we have learned that Medicaid can be made more accessible through rapid enrollment, accessible enrollment locations, acceptance of out-of-state enrollment cards, and payment for a broad range of services provided out-of-state.
Conclusions: Migrant and seasonal farmworkers are an integral support to the nation’s agricultural industry.
However, the health needs of farmworkers are considerable.
The MSFW face a number of barriers to obtaining Medicaid

coverage. Some of these problems affect the low-income population generally, but many of them are exacerbated by the characteristics of MSFW, such as their immigrant status, their fluctuating incomes, and their migratory patterns.
Implications for Policy, Delivery, or Practice: The evidence reviewed in the analysis presented here suggests the importance of an intervention and a range of viable options, including more energetic Medicaid interventions and other programs to complement Medicaid’s reach. More active attention to overcoming Medicaid barriers, coupled with a national program that uses a nationwide intermediary to bring benefits to families, would help surmount the inherent difficulties for this population created by state borders and state-based healthcare programs.
Primary Funding Source: Kaiser Family Foundation
● The Impact of Legal Regulations on State-Level
Distributions of Opioids
Jonathan Treadwell, Ph.D., Vivian Coates, M.B.A., Karen
Schoelles, M.D.
Presented By: Jonathan Treadwell, Ph.D., Senior Research
Analyst, Health Technology Assessment, ECRI, 5200 Butler
Pike, Plymouth Meeting, PA 19462; Tel: 610-825-6000; Fax:
610-834-1275; Email: jtreadwell@ecri.org
Research Objective: Some states have statutes and regulations that set legal restrictions on the prescribing of opioids and other controlled substances. For example, seven states set a maximum length of time for the validity of prescriptions for Schedule II substances (e.g., 14 days in
California). Such restrictions may discourage opioid use in the legitimate treatment of pain. We sought to measure the association between the existence of such laws and the amount of opioids that were distributed to different states in
2003. This work was conducted in the context of a systematic review on the use of opioids for the treatment of chronic, noncancer pain.
Study Design: Two kinds of legal restrictions were considered: a maximum length of prescription validity, and any additional restrictions on prescriptions. To measure opioid distributions, we used data from the Drug Enforcement
Administration's Automation of Reports and Consolidated
Ordering System (ARCOS). This system counts drug transactions for controlled substances at the manufacturing and wholesale distribution level. We examined data on the number of distributed grams of eight different opioids that are commonly prescribed for the treatment of pain (oxycodone, morphine, hydromorphone, fentanyl, meperidine, methadone, hydrocodone, and codeine). To control for confounding, we investigated 16 additional variables (ten medical and six demographic) that may also account for state-level differences in opioid distributions. Multiple regression was used to measure the association between legal restrictions and opioid distributions.
Population Studied: All 50 states and the District of
Columbia.
Principal Findings: States that set a maximum length of prescription validity had significantly lower meperidine distributions than other states, but for the other seven opioids there was no significant effect of this legal restriction. States with other additional restrictions on prescription had significantly lower distributions of oxycodone, methadone, and meperidine than other states. Such restrictions were also associated with greater distributions of codeine, suggesting a drug substitution effect (codeine can be the active ingredient in less-regulated Schedule III medications). However, we observed statistically non-significant effects for fentanyl, morphine, and hydromorphone. When we combined the different kinds of prescription restrictions (to flag any restriction on prescription), states with restrictions had significantly lower distributions of oxycodone, fentanyl, and meperidine.
Conclusions: Some legal restrictions can be shown to be associated with lower distributions of some opioids (e.g., triplicate prescription, special prescription forms). The associations, however, are not apparent for all opioids. We observed the strongest associations for oxycodone, fentanyl, and meperidine. As measured by ARCOS data, setting a maximum length of prescription validity appears to be unrelated to opioid distributions. These data, however, are assessed on the state level rather than on the patient level, and patient-level data may be more sensitive to the impact of laws pertaining to controlled substances.
Implications for Policy, Delivery, or Practice: Legal restrictions on controlled substances may potentially impede patients' access to pain medications.
Primary Funding Source: Mayday Fund and the Milbank
Memorial Fund