Comparison of Clinically Adjudicated Versus Flow-Based Adjudication of Revascularization Events in Randomized Controlled Trials

Abstract

Background: In clinical trials, the optimal method of adjudicating revascularization events as clinically or nonclinically indicated (CI) is to use an independent Clinical Events Committee (CEC). However, the Academic Research Consortium-2 currently recommends using physiological assessment. The level of agreement between these methods of adjudication remains unknown. Methods: Data for all CEC adjudicated revascularization events among the 3457 patients followed-up for 2-years in the TALENT trial, and 3-years in the DESSOLVE III, PIONEER, and SYNTAX II trial were collected and readjudicated according to a quantitative flow ratio (QFR) analysis of the revascularized vessels, by an independent core lab blinded to the results of the conventional CEC adjudication. The κ statistic was used to assess the level of agreement between the 2 methods. Results: In total, 351 CEC-adjudicated repeat revascularization events occurred, with retrospective QFR analysis successfully performed in 212 (60.4%). According to QFR analysis, 104 events (QFR ≤0.80) were adjudicated as CI revascularizations and 108 (QFR >0.80) were not. The agreement between CEC and QFR based adjudication was just fair (κ=0.335). Between the 2 methods of adjudication, there was a disagreement of 26.4% and 7.1% in CI and non-CI revascularization, respectively. Overall, the concordance and discordance rates were 66.5% and 33.5%, respectively. Conclusions: In this event-level analysis, QFR based adjudication had a relatively low agreement with CEC adjudication with respect to whether revascularization events were CI or not. CEC adjudication appears to overestimate CI revascularization as compared with QFR adjudication. Direct comparison between these 2 strategies in terms of revascularization adjudication is warranted in future trials. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: TALENT trial: NCT02870140, DESSOLVE III trial: NCT02385279, SYNTAX II: NCT02015832, and PIONEER trial: NCT02236975.