Jerry Cromwell, part-time faculty member in the College of Nursing and Health Sciences, coauthored an article in the new issue of the New England Journal of Medicine. A summary from that article is below.
A complete version of the article, "Results of the Medicare Health Support Disease-Management Pilot Program," is available on the New England Journal of Medicine website.
In the current healthcare system, Medicare patients must navigate a system structured and financed to manage their acute, rather than longer-term chronic healthcare needs. They are often treated in costly, uncoordinated settings rather than in a holistic fashion. This article evaluated a radically new approach by Medicare to address the fragmented care faced by seniors with heart failure and serious diabetes. Eight commercial disease management companies participated in the Medicare Health Support Pilot Program in different regions of the country. Each was randomly assigned an intervention population of 20,000 eligible beneficiaries. Nurses began enrolling willing patients and providing mostly remote telephone support designed to improve self-management of their conditions, medications, lifestyles, and interactions with healthcare providers. Another 10,000 equally sick beneficiaries were set aside as a control group. Firms were paid substantial monthly fees for managing patients (usually more than $100 per month per patient), but were at risk of paying back the portion of their fees not covered by Medicare savings on medical services.
A team of researchers at Research Triangle Institute, led by Nancy McCall and Jerry Cromwell, was funded by Medicare to determine whether the eight interventions saved the program money over three years while maintaining or improving quality. Medicare claims were used to evaluate many care processes (e.g., cholesterol screening), utilization (e.g., hospitalizations and emergency room visits), and Medicare payments for all covered services. Firms provided additional information on nurse contact rates with patients. Multivariate statistical methods were used to adjust for any severity differences between study and control patients.
Quality improvements were uneven. Only 14 significant improvements were found in 40 process-of-care measures. Firms had little success in slowing the rate of growth in hospital admissions or emergency room visits. Only three out of the eight firms had slower rates of increase in Medicare costs than their control groups and none of the three was statistically significant. Several unsuccessful firms dropped out of the pilot program before the end of the 36-month period.
It was not possible to disentangle the several reasons why this approach to disease management failed to achieve its goals. Lack of personal contact with patients may be one reason. Also, managing the chronic elderly population is far more challenging than a younger population with just one health problem. For example, four out of five hospital admissions for the intervention group were for reasons other than heart failure or diabetes. This made it very difficult to prevent many of the costs that patients incurred. Moreover, each firm prioritized which patients to follow more closely based usually on their baseline health status and costs. Yet, because use of the healthcare system goes in cycles, even for most of the chronically ill, the intervention likely failed to identify initially low-cost patients as future high-cost users. The fact that patients were contacted only every two to three months on average, and usually after a hospitalization, resulted in a limited intervention.
More about Cromwell
Cromwell has a PhD in Economics from Harvard University. His areas of expertise include healthcare financing and reimbursement, healthcare efficiency and equity, federal-state Medicaid cost sharing, and Medicare demonstration design and evaluation. He teaches Health Economics, Healthcare Finance, Cost Effectiveness, Secondary Data Analysis and Quantitative Research Colloquium courses in the PhD Program in Nursing, which focuses on health policy and population health. Cromwell is also a member of the PhD Committee and is working with doctoral students on their dissertations. View Cromwell's full biography.