Abstract 140: Majority of High-Risk Cardiovascular Device Studies Do Not Include Comparative Efficacy Data at the Time of FDA Approval

Abstract

Introduction Comparative effectiveness data is important in informing regulatory and treatment decisions with information on the clinical significance of the measured outcomes. There is increasing utilization and significance of cardiovascular devices in the United States, thus it is important to know the comparative efficacy of novel cardiovascular devices at the time they receive Food and Drug Administration (FDA) approval. Hypothesis We assessed the hypothesis that clinical trials results submitted to the FDA for approval of novel, high-risk cardiovascular devices often provide comparative efficacy data. Methods Comparative efficacy was defined as the presence of an active control. For all high-risk cardiovascular devices approved by the FDA from 2000 through 2011, we extracted information on comparative efficacy data from Summaries of Safety and Effectiveness Data (Summary) which include study data used to justify FDA approval. All identified studies were examined to determine if they relied upon active controls, historical controls, objective performance criteria, or if they had no control. The proportion of studies containing comparative efficacy data was calculated and cross-tabulated by approval year and device class. A multivariate logistic model was used to assess trends over time and the Kruskal-Wallis test was used to examine differences between device sub-types. Results We examined 114 Summaries which contained 340 cardiovascular device studies. Of these studies, 140 (41%) contained an active control. Historical controls 48 (14%) and objective performance 83 (24%) criteria were commonly used. Sixty-nine (20%) of the studies were single-arm studies without controls. In our Summary-level analysis, 74 (65%) devices had an active control in at least one supporting study. Approval year was not significantly associated with type of control arm after controlling for device type in our multivariate logistic analysis. Use of comparative efficacy data was significantly associated with device type. Use ranged from 15/19 (79%) of hemostasis devices to 0/3 (0%) of ventricular assist devices. Conclusions A minority of studies for the highest risk cardiovascular devices approved by the FDA since 2000 included comparative efficacy data through the use of an active comparator arm. Increasing use of active controls would help to better inform regulatory and treatment decisions at the point of FDA approval.