The center effect in heart transplantation.

Transplantation outcomes vary across centers, prompting interest in the notion of a "center effect." The components of this effect are not well understood, although experience is often regarded as the primary factor. Most studies, however, have failed to confirm an association between transplant program activity level and outcome. While there have been several published reports on the kidney transplant center effect, only 1 such (unpublished) report exists concerning heart transplantation. Data for this analysis were obtained from several sources that include: the Registry of the International Society for Heart and Lung Transplantation; the Medicare Provider Analysis and Review data maintained by the Health Care Financing Administration (HCFA), various hospital-specific data compiled by the American Hospital Association (AHA); and finally, 91% of all eligible heart programs participated in a special purpose survey intended to obtain critical data on transplant center characteristics. These 4 data sources were combined into a single data base representing 1,602 patients at 114 centers. The data were then analyzed using a discrete piecewise exponential hazards model. This is a nonparametric approach toward the modeling of risk, wherein no assumption is made about the shape of the survival curve. Risk was assessed up to 2 years posttransplant. Overall 1-year patient survival was 82.4%. In the multivariate models, neither recipient nor donor sociodemographic characteristics were associated with patient survival. Clinical characteristics were the most critical predictors of outcome, including use of an artificial device and retransplantation, both of which had a strong adverse effect. This was predictable based upon univariate analyses alone. The results of this analysis suggest that experience is positively related to heart transplant patient survival; however, due to sample size and other considerations, the experience effect did not achieve statistical significance. Nevertheless, while outcomes did vary by center, the extent of variation was not nearly as great as some commentators have argued. This suggests that heart transplantation is a technology that can be acceptably applied in diverse settings with excellent outcomes.