, Wd Z Ϯ͘Ϯ 54 ABSTRACT Background Decisions on eligibility criteria for randomized trials rely on clinical experience and lessons from prior trials. With growing computing capabilities and data access, possibilities have opened for data-guided criteria selection. This study aims to evaluate the magnitude and predictors of clinical trial and registry mismatch based on trial inclusion and exclusion criteria. Methods Interventional phase 3 trials registered for heart failure (HF) in ClinicalTrials.gov as of end 2021 were identified. Natural language processing was used to extract and structure the eligibility criteria for quantitative analysis. These criteria were ranked by frequency. The most common ones were applied to estimate eligibility, as a proportion of registry patients. Patient eligibility for HF with reduced ejection fraction (HFrEF) trials were compared between the ASIAN-HF and BIOSTAT-CHF registries. Results One hundred and sixty-three HFrEF trials were identified. LVEF aside, the most frequently used inclusion criteria were NYHA functional class (69%), worsening HF and natriuretic peptides whereas the most common comorbidity-based exclusion criteria were acute coronary syndrome (64%) and valvular heart disease (47%). On average, 20% of registry patients were eligible for enrolment. Eligibility distributions did not differ between Asian (N=4868) and European (N=2545) registry populations. With time, HFrEF trials became more restrictive, with a change in eligibility from 0.4 in 1985-2005 to 0.19 in 2017-2021. When frequency in trials is taken in consideration, prior MI, NYHA functional class, age and prior HF hospitalization had the highest impact on restrictiveness. Conclusion Based on data for 14 eligibility criteria, one-fifth of registry patients were eligible for phase 3 HF trials and eligibilities were comparable between Asian and European registry patients.
RkJQdWJsaXNoZXIy MjY0ODMw