2.2 Eligibility for HFrEF trials 61 diuretics (99.5% vs 82%) as this was a requirement for participation in the BIOSTATCHF cohort. Between one and fourteen eligibility criteria were applied in the estimation of eligibility. Summarising across 163 HFrEF trials, about one-fifth of the combined target population were eligible (median eligibility score 0.19 (95% CI 0.14, 0.24)). Figure 2 shows that the distribution of eligibility scores across trials were broadly similar between Asian and European populations. Median eligibility was marginally higher in the Asian patients (0.20 vs 0.17) but was not statistically significant (p=0.3). Table 3 displays median eligibility scores by trial characteristics. Eligibility for trials declined with time by more than half from 0.40 to 0.14 between trials initiated in 1985-2005 and 2006 - 2015. Interestingly, trials from the recent seven years show a reversal, increasing to median eligibility of 0.19 (p-value=0.02). By intervention type, drug trials enrolled a more representative pool of participants (median score 0.24) compared to device and procedural or diagnostic trials (median were both 0.09, p<0.001). Further, trial eligibility differed by primary funding source; eligibility was highest among academic /healthcare institution-funded trials, followed by those funded by NIH and lastly, industry-sponsored trials. The anticipated size of trials, however, was not predictive of eligibility (p=0.4). Comparing impact of individual criterion by target population Trial eligibility can be limited when one or more exceptionally restrictive criterion is present. Of the criteria assessed, prior HF hospitalization, MRA background treatment and anemia were most restrictive with eligibility scores of 0.38, 0.56 and 0.61 respectively (Figure 3). Eligibility based on single criterion were comparable between Asian and European patient populations with a few exceptions. Prior HF hospitalization, history of MI, normal sinus rhythm and cardiac devices were more restrictive among European patients resulting in 26%, 20%, 20% and 13% lower eligibility compared to Asian patients. On the other hand, for trials which focus on devices or iron supplementation, QRS duration and anaemia status or serum haemoglobin were comparatively more restrictive in Asian patients with relative differences of 33% and 14% lower eligibility. Upon inverse-frequency weighting of each criterion, the most restrictive were prior MI, NYHA functional class, age and prior HF hospitalization. (Supplementary figure 1) LVEF was not compared because it is present in 100% of trials.
RkJQdWJsaXNoZXIy MjY0ODMw