The impact of an end-point committee in a large multicentre, randomized, placebo-controlled clinical trial - Results with and without the end-point committee's final decision on end-points

Background A multicentre study permits rapid recruitment of a large number of patients. However, there is a risk of heterogeneities in end-point evalu ations, as complex definitions of criteria are interpreted by several local investigators from different hospitals. Reports discussing endpoint evalua tion are sparse. The TRIM trial was a multicentre trial of a thrombin inhib itor in patients with unstable angina or non-Q myocardial infarction. In th is study, an independent end-point committee evaluated all the reported eve nts of death, acute myocardial infarction and refractory angina pectoris in order to obtain uniform judgements of major end-points. Study aims To describe the work of the end-point committee, to analyse its possible effect on the final study results and to discuss the impact on the design of future trials. Method The end-point committee consisted of four members, one from each par ticipating country. After the data were processed by the study monitors, co mpleted case record forms and patient files for patients with reported end- points were mailed to the national member of the end-point committee for ju dgement. The end-point committee met regularly and made final decisions abo ut the end-points. The work of the end-point committee was documented on a separate case record form. Results The end-point committee assessed 246 events of death, acute myocard ial infarction and refractory angina pectoris in 187 of the 1209 patients ( 15.5%) in the TRIM trial. Misinterpretation of the index event, an inclusio n myocardial infarction, as an early cardiac event was found in 12 patients . After end-point committee judgements, the number of patients with acute m yocardial infarction or refractory angina pectoris during 30 days of follow -up was reduced from 177 to 153 (13.6% reduction). The classification of ev ents was changed in 53 of 187 patients (28.3%) with death, acute myocardial infarction or refractory angina pectoris. The data assessed by the safety committee was significantly different from the final database after end-poi nt committee judgements. Conclusion The end-point committee corrected misinterpretations in such a h igh proportion of cases that the final results differed significantly from the preliminary results delivered to the safety committee. End-point judgem ents by an end-point committee should be performed in multicentre clinical trials to improve the quality and reliability of study results.