Objective: Significant interinstitutional variation in mortality after congenital heart surgery has been demonstrated. Noting an association between reduced mortality and higher volume, a center with a small annual case volume began in August 1998 to selectively refer to high-volume surgical centers based on published or "apparent" low mortality rates for specific cardiac lesions. This study was undertaken to evaluate the effect of evidence-based referral in this practice.
Design, setting, and participants: A retrospective cohort comparison over a 10-year period for a small Midwestern pediatric cardiology practice. The institutional database was retrospectively reviewed for children (<18 years) undergoing surgery from August 1992 to July 2002. Data were divided into 3 time periods (August 1992 to July 1995, period 1; August 1995 to July 1998, period 2; and August 1998 to July 2002, period 3). Hospital discharge abstract data from 5 states (California, Illinois, Massachusetts, Pennsylvania, and Washington) in 1992, 1996, and 1998 provided contemporaneous benchmarks. Risk adjustment was performed using the Risk Adjustment in Congenital Heart Surgery-1 method. Risk category, age at surgery, prematurity, and major noncardiac structural anomaly were entered into a multivariate logistic regression model to compare in-hospital mortality adjusting for case-mix differences.
Results: A total of 514 congenital heart surgical cases were identified from August 1992 to July 2002; 507 cases (98.6%) were assigned to a risk category and analyzed further. Unadjusted in-hospital mortality rates were 9.3% in period 1, 5.9% in period 2, and 1.3% in period 3. Unadjusted mortality rates for cases from benchmark data were 6.4% in 1992, 4.8% in 1996, and 3.7% in 1998. Risk adjusted mortality was comparable to the benchmark data in periods 1 and 2, but superior outcomes (odds ratio = 0.24) were demonstrated in period 3.
Conclusions: Evidence-based referrals from a small-volume pediatric cardiac center to large-volume institutions resulted in a reduction in mortality after congenital heart surgery.