Procedural failures of physicians or teams in interventional healthcare may positively or negatively predict subsequent patient outcomes. We identify this “learning from failure”-effect by applying (non-)linear dynamic panel methods using data from the Belgian Transcatheter Aorta Valve Implantation (TAVI) registry containing information on the first 860 TAVI procedures in Belgium. Using bias-corrected fixed effects linear probability models and the split-panel jackknife estimator proposed by Dhaene and Jochmans (2015), we find that a previous death positively and significantly predicts subsequent survival of the succeeding patient. Moreover, our results also provide evidence for learning from failure for stroke. We find that these learning from failure effects are not long-living and that learning from failure is transmitted across adverse events, e.g., a stroke affects subsequent survival.

, ,
, , ,
HEDG working paper
Health Economics
Department of Applied Economics

van Gestel, R., Müller, T., & Bosmans, J. (2017). Learning From Failure in Interventional Care. Health Economics. Retrieved from