Evaluating the Medicare Coordinated Care Demonstration
The Medicare Coordinated Care Demonstration tested whether care coordination and disease management programs lowered costs for and improved the health of beneficiaries with chronic illnesses covered by fee-for-service Medicare. These improvements were to result from programs teaching patients how to better adhere to self-care and medication regimens and from improving communication among physicians and between patients and physicians. In January 2002, the Centers for Medicare & Medicaid Services (CMS) selected and funded 15 demonstration programs to participate in the evaluation. Each program began enrolling beneficiaries between April and September 2002 and was authorized to operate for four years. Mathematica's study, the largest random assignment study to date of care coordination programs, examined whether the programs met their goals of reducing costs, improving quality of care, and improving patient satisfaction with care. It also included a qualitative assessment of the strengths and weaknesses of each program.
The third report to Congress synthesized findings from more than four years of the demonstration programs’ operations and provided the most comprehensive and rigorous estimates ever presented on the effectiveness of care coordination interventions in a Medicare fee-for-service setting. Only 2 of the 15 programs resulted in reduced hospital admissions, and none of the programs generated net savings. There were few effects on beneficiaries’ adherence to prescribed medication, diet, or exercise regimens. The report described ways in which the more successful programs differed from the unsuccessful ones and concluded that programs with the differentiating features could generate savings for Medicare if implemented for a high-risk target population.
Data Collection Highlights
The analysis relied on Medicare claims data covering the four demonstration years. The study also included a patient survey conducted 6 to 12 months after enrollment and two rounds of physician surveys. For the patient survey, samples were selected from Medicare beneficiaries who enrolled in the care coordination programs. The survey took place between May 2003 and June 2004, with patients interviewed 7 to 12 months after enrollment. At each program, the target for the survey was completion of 600 interviews, with 300 treatment group members and 300 control group members. Altogether, 7,357 interviews were completed, with a completion rate of 92 percent. The response rate, subtracting ineligible patients, was 95 percent. For the physician survey, primary care physicians were identified from patient consent forms of treatment group members. Physicians to be interviewed in each program were selected with probability proportional to the number of patients who had named them on their consent forms. Interviews were completed with 699 physicians, 65 percent of the eligible sample.
CMS extended the end dates for 11 of the 15 programs from 2006 to 2008, and 2 of the 11 were further extended through March 2010 as a result of having demonstrated potential cost neutrality. The extended evaluation has three components:
(1) An implementation analysis using information from discussions with program staff, case presentations, and direct observation to document the operations of the two programs extended to 2010.
(2) An impact analysis based on Medicare claims data to estimate program effects on expenditures, service use, and quality-of-care indicators through 2010.
(3) A synthesis combining implementation findings and impact estimates to draw inferences about program features that contributed to any improved health or cost outcomes.
Deborah Peikes, senior researcher; Arnold Chen, senior clinician researcher; Jennifer Schore; senior researcher; and Randall Brown, vice president and director of Health Research won the 2010 AcademyHealth article-of-the-year award as well as the National Institute for Health Care Management Foundation's 2010 research award for excellence in original and creative health care research for a paper from the Medicare Coordinated Care Demonstration. The winning article, "Effects of Care Coordination on Hospitalization, Quality of Care, and Health Care Expenditures Among Medicare Beneficiaries: 15 Randomized Trials," appeared in the Journal of the American Medical Association in February 2009.
View all publications from the original study.