Investigator-reported vs adjudicated cause of death in the TIOSPIR™ trial

Background: Accurate determination of cause of death (CoD) is challenging in some chronic obstructive pulmonary disease (COPD) patients. In the TIOtropium Safety and Performance In Respimat ® (TIOSPIR™) trial (n=17,135), comparing safety and efficacy of tiotropium Respimat 5/2.5µg with HandiHaler ® 18µg in COPD patients, all-cause mortality was a primary endpoint. A mortality adjudication committee (MAC) assessed all investigator (I)-reported deaths. Aim: To investigate discordance in MAC- and I-reported CoD. Methods: The MAC provided continuous, systematic, independent, blinded, expert assessment of I-reported deaths (n=1302); it reviewed medical documentation, case-report forms and witness statements and assigned an underlying CoD. Results: Concordance between I- and MAC-reported CoD was highest for cancer deaths (Table). Differences were observed, in particular cardiac deaths (16% I vs 5% MAC) and deaths assigned to general disorders including cardiac-related deaths i.e. of unknown cause and sudden, (17% I vs 25% MAC). Reasons for shifts included a switch to the underlying CoD and insufficient information available. Conclusion: In TIOSPIR™ there was discordance between MAC- and I-adjudicated CoD, especially higher I-reporting of cardiac deaths. The data confirm the importance of using composite endpoints including all-cause mortality and major cardiac adverse events in COPD trials given the difficulty in ascertaining an accurate CoD.