Mortality is an important endpoint in chronic obstructive pulmonary disease (COPD) trials, although accurately determining cause of death is difficult. In the Understanding the Potential Long-term Impacts on Function with Tiotropium (UPLIFT®) trial, a mortality adjudication committee (MAC) provided systematic, independent and blinded assessment of cause-specific mortality of all 981 reported deaths. Here we describe this process of mortality adjudication and methodological revisions introduced to help standardise the adjudication of two areas recognised to pose particular difficulty; firstly, the classification of fatal COPD exacerbations that occur in the setting of pneumonia and secondly, the categorisation of sudden death. In addition MAC determined cause of death was compared with that reported by site investigators (SIs). MAC-assigned causes of death were: respiratory, 35%; cancer, 25%; cardiovascular, 11%; sudden cardiac death, 4.4%; sudden death, 3.4%; other, 8.8%; unknown, 12.4%. Cancer/cardiac deaths were more common in Global Initiative for Chronic Obstructive Lung Disease stage II, respiratory deaths in stages III and IV. Agreement between MAC and SI regarding cause of death was complete (50.2%), incomplete (18.5%) or none (31.3%). The SI classified deaths as cardiac three-fold more frequently than MAC (incidence rate [IR]/100 patient-years 0.797 vs. 0.257), although IR ratios for cardiac deaths for tiotropium vs. control were similar between SI and MAC. Discrepancies between MAC- and SI-adjudicated causes of death are common, especially increased reporting of cardiac deaths by the SI. Future multicentre COPD trials should plan appropriate infrastructure before study initiation to ensure collection and interpretation of fatal events data.
Copyright © 2011 Elsevier Ltd. All rights reserved.