Frequency of Medical Reversal Among Published Randomized Controlled Trials Assessing Cardiopulmonary Resuscitation (CPR)

Objective: To characterize what proportion of all randomized controlled trials (RCTs) among patients experiencing cardiac arrest find that an established practice is ineffective or harmful, that is, a medical reversal. Methods: We reviewed a database of all published RCTs of cardiac arrest patient populations between 1995 and 2014. Articles were classified on the basis of whether they tested a new or existing therapy and whether results were positive or negative. A reversal was defined as a negative RCT of an established practice. Further review and categorization were performed to confirm that reversals were supported by subsequent systematic review, as well as to identify the type of medical practice studied in each reversal. This study was conducted from October 2017 to June 17, 2019. Results: We reviewed 92 original articles, 76 of which could be conclusively categorized. Of these, 18 (24%) articles examined a new medical practice, whereas 58 (76%) tested an established practice. A total of 18 (24%) studies had positive findings, whereas 58 (76%) reached a negative conclusion. Of the 58 articles testing existing standard of care, 44 (76%) reversed that practice, whereas 14 (24%) reaffirmed it. Conclusion: Reversal of cardiopulmonary resuscitation practices is widespread. This investigation sheds new light on low-value practices and patterns of medical research and suggests that novel resuscitation practices have low pretest probability and should be empirically tested with rigorous trials before implementation.