Background Results of transcatheter implantation of the Rashkind double umbrella occluder for treatment of patent ductus arteriosus (PDA) have improved over time.We evaluated factors associated with changes in outcomes seen in the early clinical experience with this device.Methods and Results We reviewed records of otherwise healthy pediatric patients undergoing occluder implantation for isolated PDA between 1982 and 1987 at six major US and Canadian referral centers.Logistic regression was used to assess the impact of ductal size, patient age and weight, calendar time, prior intrainstitutional occluder experience, and residual center effects on closure rates. There were no study deaths. Of 180 occluder placement attempts, 76.7% (95% confidence limits=70.5%, 83.0%) produced PDA closure by auscultation at 14-month cardiology follow-up or its equivalent. Success probabilities increased in a linear fashion as ductus size decreased, ranging from 41.7% for large (4.1 to 9.0-mm) ducts to 89.4% for small (1.5 to 2.5-mm) lesions. PDA occluder closure rates also improved over time, increasing from 47.4% in 1982-1983 to 84.6% in 1986 and 83.1% in 1987. Multivariate logistic regression indicated that improved outcomes were mainly attributable to decrease in the size of treated ducts over time. Independent of this patient selection, improvements in outcome were less strongly associated with cumulative intracenter experience, calendar time, and age.Conclusions Increasing closure rates seen over time primarily reflected more judicious selection of patients as experience accumulated.However, other factors also contributed to the observed trends. An analysis of current treatment results would be of considerable interest. (Circulation. 1994;90(part 2): II-36-II-42.)