We summarise a recent article, published in Journal of the American Medical Informatics Association, that evaluated the impact of eligibility criteria on recruitment and clinical outcomes of COVID-19 clinical trials using electronic health records (EHRs).
It has been nearly a year since SARS-CoV-2 first appeared and propagated rapidly across the world. Despite the rising number of confirmed cases and deaths, there is no consensus on a definitive therapy for COVID-19. Many clinical trials have been launched to evaluate the efficacy and safety of experimental agents. Nonetheless, there are concerns regarding the robustness of trial designs. This is particularly related to insufficient sample sizes, inadequate statistical adjustments and inconsistent definitions for endpoints. To date, no studies have evaluated the influence of the eligibility criteria of the COVID-19 studies on these issues.
Trials with restrictive eligibility criteria do not provide evidence regarding the risk and benefits of tested drugs for excluded patients. This limits their generalisability to real-world patients with the target condition. On the other hand, applying overly permissive eligibility criteria has its on challenges. For example, it may result in a heterogenous cohort and reduce the probability of detecting a drug’s true effect. As a result, it is important to balance the trade-off between internal validity and external validity during eligibility criteria definitions.
Data-driven eligibility criteria
In this study, researchers aimed to measure the influence of eligibility criteria on patient recruitment and outcomes in a retrospective cohort. They first identified the frequently used eligibility criteria in COVID-19 trials. Then, using EHR data, they assessed the influence of individual eligibility criteria based on the number of patients that could be included or excluded by each criterion. The outcome events were compared between the groups to estimate how eligibility criteria could be modified to optimise the balance between internal and external validity.
There were 3,251 patients diagnosed with COVID-19 from the CUIMC EHR included in the analysis. During follow-up, the composite events occurred in 18.1% of the COVID-19 cohort. In a hypothetical trial with common eligibility criteria, 33.6% (690/2051) were eligible among patients with evaluable data. Out of these, 22.2% (153/690) had the composite event. By adjusting the thresholds of common eligibility criteria based on COVID-19 patient characteristics, the team observed more composite events from fewer patients.
This research demonstrates the potential use of EHR data of COVID-19 patients to help inform the selection of eligibility criteria and their thresholds. The team believe this method promises to improve the feasibility and efficiency of COVID-19 clinical trial recruitment.
Image credit: By kjpargeter – www.freepik.com