2.2 Eligibility for HFrEF trials 73 rigorous data quality controls, completeness and patients that span multiple countries. Lastly, temporal characteristics for event or procedure-based criteria such as time from revascularization could not be determined in the patient data and as they were commonly a basis for exclusion, could result in underestimation of eligibility by these features. In the present study, we have shown value in characterizing eligibility in two distinct target populations. For instance, investigators for device trials who intend to select patients by history of implanted devices or QRS complex duration may need to be aware of lower eligible numbers in Asian sites whereas exclusion of patients with AF or history of MI will lead to comparatively slower accrual in European sites. Enrichment with criteria such as prior HF hospitalization potentially leads to higher enrolment rates in the Asian population, given that the proportion with prior hospitalization is already a-third higher in its a shorter observation period of 6 months. Understandably, these estimates can be sensitive to time and study site but here we demonstrate the feasibility of data-driven decisions at the design stage, which can potentially improve cost-efficiency of future trials. The use of observational patient data in guiding trial eligibility criteria is a fairly new concept. Several case studies have demonstrated promising opportunities in testing assumptions, simplify enrolment and expanding clinical trial access.10 Further studies on impact of trial criteria decisions on cardiovascular event rates or hazard can be useful especially at pre-trial design stages. Existing obstacles related to unstructured data formats, data sharing policies and data quality can be overcome to enable real-time understanding of varying eligibility criteria decisions on patient eligibility and outcome event rates for trials. Based on an analysis of 163 trials over 37 years, we show that one-fifth of registry patients were, on average, eligible for enrolment in phase III HFrEF trials, with comparable eligibilities between Asian and European populations. By individual criterion, previous HF hospitalization, requirement of MRA therapy, and anaemia were most restrictive and could adversely impact accrual and generalizability of individual trials. On a broad perspective for HFrEF therapeutics, criteria that had most impact by both patient selectivity and frequency in trials were prior MI, NYHA class, age and previous HF hospitalization.
RkJQdWJsaXNoZXIy MjY0ODMw