HEALTH REFORM: Mixed Results from Medicare Experiments

In a promising sign for health care delivery system reform, a Pennsylvania nonprofit substantially reduced costs and improved outcomes for Medicare patients with chronic illness in a care management study. The success of this trial—and the limited accomplishments of others in this experiment—provides more evidence of the need to involve physicians in any disease management reform.

The February edition of JAMA describes 15 care coordination trials started in 2002 by CMS. The pilots tested the ability of care coordination programs to keep elderly, chronically ill patients out of the hospital, on diet and exercise regimens, and otherwise improve health and lower spending.

The results demonstrate three major points. First, the concept of a patient-centered medical home may not be easy to put into practice, but it can work. The Pennsylvania pilot gave high-utilization patients monthly visits and coordinated care from a nurse. The nurse encouraged lifestyle changes and developed a relationship with the patient. The pilot restructured care around the patient, and the nurse served as a sort of medical home for the patient.

Second, as Robert Berenson of the Urban Institute has posted on this blog before, successful disease management must involve the provider. Both of the successful pilots, in Pennsylvania and Iowa, included regular personal contact between a provider (in these cases, a nurse) and the patient. Since patients look to their providers, generally physicians, for responsibility for their health care, any disease management reform must include provider involvement. Past Medicare pilots that failed to save money because they attempted to involve outside entities in a patient's health rather than the provider.

Third, as the study author pointed out, it is challenging to get either providers or patients to change their behaviors. Most of the pilots struggled to get patients to exercise, improve their diet, stop smoking, and make other lifestyle changes necessary to improve health and reduce cost. Similarly, physicians who are used to operating as "lone rangers" without oversight of their work have a difficult time changing to a new model. The lesson here is that patients and providers need incentives to change their behavior. (See the USA Today story on financial incentives to quit smoking.) Any comprehensive reform has to provide appropriate incentives to improve health and cut costs.

So it is possible to save money and improve outcomes with a new care management model, but to work, it must be structured for success. (See our post on a previous Medicare pilot.) It needs to provide appropriate incentives, involve providers, and focus care on the patient. Hopefully the results of this experiment will inform future attempts at delivery system reform.