How Robust Are Studies in the American Board of Emergency Medicine Maintenance of Certification Lifelong Learning and Self-assessment? An Examination of Fragility and Bias of Included Randomized Controlled Trials.

Objective: Critics have raised concerns regarding the validity of maintenance of certification (MOC) programs. We sought to examine the quality of the randomized controlled trials (RCTs) selected for the lifelong learning and self-assessment (LLS) component of the American Board of Emergency Medicine (ABEM) MOC program.

Methods: We systematically reviewed the ABEM LLS reading lists from 2004 to 2017 to identify RCTs with dichotomous outcomes and superiority designs. A fragility index (FI) was calculated using Fisher's exact test for all statistically significant dichotomous outcomes. Bivariate correlation was performed to examine associations between the FI and RCT study characteristics. Each included study was evaluated with the Cochrane Collaboration risk-of-bias (ROB) tool.

Results: Thirteen superiority RCTs with dichotomous outcomes were included in the 2004-2017 LLS reading lists. Ten had a statistically significant outcome, and the majority were robust and at low ROB. The median trial size was 511 patients (interquartile range [IQR] = 251-1,517), and the median FI was 10 (IQR = 7-18); i.e., if 10 patients in the treatment arm had not had events, the results would not have been statistically significant.

Conclusions: The majority of RCTs included in the LLS are robust and at low ROB.