The Cost of Innovation and Evidence in Cardiac Surgery.

This review summarizes a systematic analysis of 216 randomized trials of cardiovascular interventions performed during 2008-2019, according to the source of trial funding. The systematic analysis showed that on average the results of each trial would change significance if only 5 patients experienced different outcomes. Industry-sponsored trials were more likely to use composite endpoints, noninferiority designs, and twice as likely as nonindustry trials to report results favoring the device arm. Over 80% of industry trials used reporting strategies or "spin" suggesting the device arm was advantageous versus fewer than half of non-industry trials. The review discusses the implications of these findings.