Summary 211 towards the primary goal of enhancing diversity and representation of people with heart failure in clinical trials. Chapter 1 introduces the topic of clinical trial generalizability and an outline of the projects within this thesis. Chapter 2 focuses on assessing the eligibility of heart failure registry patients for trials as a measure of trial generalizability and the extent to which the two populations differ. In chapter 2.1, we compared the characteristics between five heart failure trials and two patient registries to evaluate the gap between them. We confirmed prior evidence that trial patients were younger by about 10 years, a-third less females, of lower disease severity and had less coexisting medical illness. Although the probability of death appeared lower among trial patients compared to registry patients, this difference was fully explained by disparities in individual characteristics and background disease between the groups. However, when we looked specifically at heart disease-related deaths, they were not the same between the trial and registry groups such that trial patients were 30% more likely to die of heart-related conditions. The fact that known background risk could not fully explain this difference suggests the presence of remaining unmeasured differences between the two patient populations. These differences may arise from other undocumented medical illness or factors relating to whether a person is given the option or chooses to participate in trials. In chapter 2.2, we extended the scope from five trials to all HFrEF trials registered on a clinical trial repository, ClinicalTrials.gov and summarized the main entry criteria to these trials. Rather than taking the earlier approach of side-by-side population comparison, we assessed hypothetical scenarios by applying specific entry criteria of each trial on individuals from an Asian and a European heart failure registry to calculate the percentage which would have entered the trial. We did this for 163 trials and found that only a-fifth of registry patients were eligible to participate in the trials. Overall, both Asian and European registry patients were equally eligible to be enrolled into the trials. Several entry criteria carried more weight in excluding patients from trials. These were previous hospitalization for heart failure, previous heart attack, New York Heart Association class (a measure of heart failure severity) and age. From this study, we demonstrated that it is possible to project the impact of specific trial entry criteria on trial generalizability during design phase of a new study.
RkJQdWJsaXNoZXIy MjY0ODMw