Comparison of Investigator-Reported and Clinical Event Committee–Adjudicated Outcome Events in GLASSY

Abstract

Background: Event adjudication by a clinical event committee (CEC) provides a standardized, independent outcome assessment. However, the added value of CEC to investigators reporting remains debated. GLASSY (GLOBAL LEADERS Adjudication Sub-Study) implemented, in a subset of the open-label, investigator-reported (IR) GLOBAL LEADERS trial, an independent adjudication process of reported and unreported potential outcome events (triggers). We describe metrics of GLASSY feasibility and efficiency, diagnostic accuracy of IR events, and their concordance with corresponding CEC-adjudicated events. Methods: We report the proportion of myocardial infarction, bleeding, stroke, and stent thrombosis triggers with sufficient evidence for assessment (feasibility) that were adjudicated as outcome events (efficiency), stratified by source (IR or non-IR). Using CEC-adjudicated events as criterion standard, we describe sensitivity, specificity, positive and negative predictive value, and global diagnostic accuracy of IR events. Using Gwet AC coefficient, we examine the concordance between IR- and corresponding CEC-adjudicated triggers. There was sufficient evidence for assessment for 2592 (98.3%) of 2636 triggers. Results: Overall, the adjudicated end point-to-trigger ratio was high and similar between IR- (88%) and non-IR–reported (87%) triggers. The global diagnostic accuracy and concordance between IR-reported and CEC-adjudicated outcome events was 0.70 (95% CI, 0.65–0.74) and 0.54 (95% CI, 0.45–0.62), respectively, for myocardial infarction; 0.77 (95% CI, 0.75–0.79) and 0.71 (95% CI, 0.68–0.74) for bleeding; 0.70 (95% CI, 0.62–0.79) and 0.59 (95% CI, 0.43–0.74) for stroke; 0.59 (95% CI, 0.52–0.66) and 0.39 (95% CI, 0.25–0.53) for stent thrombosis. For IR bleedings, the concordance with the CEC on type of events was generally weak. Conclusions: Implementing CEC adjudication in a pragmatic open-label trial with IR events is feasible and efficient. Our findings of modest global diagnostic accuracy for IR events and generally weak concordance between investigators and CEC support the role for CEC adjudication in such settings. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT03231059.