All 15 experimental coordinated care programs launched by Medicare in 2002 failed to generate net savings, and only two of them reduced hospital admission rates among patients with chronic diseases, including diabetes.
The results of the experiment, which ran from April 2002 through June 2005, were published in the February 11, 2009, issue of JAMA (Journal of the American Medical Association).
The experiment was launched because people with chronic disease place the biggest strain on the Medicare system, the result, officials believed, of inadequate counseling on diet, medication, and self-care; lack of access to medical help other than emergency rooms; and poor communication betweens patients and doctors. They wanted to test whether care coordinators could decrease costs by improving counseling, access, and communication.
In 2002, the Centers for Medicare and Medicaid Services (CMS) awarded contracts to 15 demonstration programs operated by various healthcare entities. Program participants included Medicare patients with such chronic diseases as diabetes, congestive heart failure, and coronary artery disease.
The study tracked the rate of hospitalizations and Medicare costs for 18,309 patients in the 15 programs, whose participants ranged in number from 178 to 2,657. The care coordinators turned out to be mostly nurses who monitored and educated patients primarily by phone and acted as go-betweens with doctors.
By the study’s end, 13 of the 15 programs had not significantly affected the number of hospitalizations. At the statistical extremes, Mercy Medical Center in Iowa reduced hospitalizations by 17 percent, while Maryland’s Charlestown retirement community saw a 19 percent increase in hospitalizations.
In an editorial that accompanied the JAMA article, John Z. Ayanian, MD, MPP, at Brigham and Women’s Hospital and Harvard Medical School in Boston, wrote that the study offered two important insights:
- 1. Care coordinators must interact with patients in person and not by telephone. He said that only four of the 15 programs emphasized in-person contact between coordinators and participants.
- 2. Care coordinators should be based in doctors’ offices or attend medical appointments with their patients “to have a reasonable prospect of influencing care.”
Dr. Ayanian did not explain how Medicare would fund such high levels of interaction between care coordinators and their patients. For example, a care coordinator can reach far more people per day by telephone than by personal contact. To give Medicare patients the benefit of face-to-face contact would require hiring more nurses-highly unlikely in today’s recessionary economy.