When the Medicare Part D program began in 2006, it increased drug coverage for Medicare beneficiaries from 59 percent to 89 percent, says Christopher C. Afendulis, PhD, a lecturer in the department of health care policy at Harvard Medical School. Afendulis and colleagues sought to determine whether this change reduced hospitalization rates for conditions sensitive to drug adherence. Their findings, published in Health Services Research, indicate that for the conditions studied, Part D reduced the overall rate of hospitalization by 20.5 per 10,000 (4.1 percent) — about 42,000 admissions.
“There have been a few studies that have demonstrated that relatively small changes in copayment amounts can lead to large changes in terms of inpatient hospitalizations,” says Afendulis. “There may be a similar process at work when seniors gain drug coverage.”
He recommends that clinical executives “think creatively about how they might aggressively manage copayments for particular types of drugs or encounters with physicians.”
CHF, diabetes, asthma
Afendulis says, “We have confidence in the results for the summary measure (that is, hospitalization for any of the eight specific conditions), CHF, uncontrolled diabetes, and asthma.” The other conditions “fail to reach conventional levels of statistical significance.” That is, the change in hospitalization rates may simply be caused by chance.
“Impact” is the change in terms of the number of hospitalizations per 10,000. “Relative impact” is the change relative to the mean value of each hospitalization type.
For example, since hospitalizations for uncontrolled diabetes are relatively rare (3.2 per 10,000), the impact of the change from Part D is also small (-0.6 per 10,000). But as a percentage of the base hospitalization rate for this condition, the impact is large (15.5%).
Diabetes (short term)
Congestive heart failure
Acute myocardial infarction
Impact of coverage change
Relative impact of coverage change
Numbers in parentheses are standard errors. These are measures of the uncertainty associated with each estimate. The smaller this number is compared with the estimate, the more confident the researchers are that they observed a “real” effect as opposed to an effect that is caused by statistical chance.
Source: Afendulis CC, He Y, Zaslavsky AM, Chernew ME. The impact of Medicare Part D on hospitalization rates. Health Serv Res. 2011;46(4):1022–1038.