Generalizability of Randomized Controlled Trials in Heart Failure

Abstract

In the first part of this thesis, we explored the differences between trial participants and registry cohorts to explain the mismatch between heart failure trial participants and patient registry populations to identify actionable opportunities toward achieving generalizable trials in the future. From the findings, we confirm that underrepresentation of females, older people and people with multiple background conditions remain in recent trials. On disease outcomes, heart failure trial populations had higher risk of heart-related deaths compared to the routine care population and this is likely because trials seek patient types which gives the highest likelihood of detecting benefit from heart treatment. Interestingly, we found differences between males and females in trials, such that there is lower-than-expected heart-related deaths in trial females compared to registry females, suggesting that they may be less representative of the registry patients compared to males. This signals a need to advocate for future trials to be sufficiently balanced in numbers by both sexes, while overcoming possible psychosocial and logistical barriers to trial participation, particularly among females. Next, we investigate the incidence of heart failure hospitalization and its 10- year trends by age, sex and ethnicity in a multiracial population in Malaysia and found variations in heart failure outcomes between ethnicities. In relation to trial enrolment, insufficient information capture on race/ethnic and geographical subgroups in existing trials prevent assessment of trial generalizability in this respect, particularly for global trials. Lastly, we compared eligibility for trials between Asian and European heart failure patients and found that eligibility based on trial entry criteria did not differ between the two distinct populations.