Overcoming the discrepancies between RCTs and real-world data by accounting forSelection criteria, Operations, andMeasurements of Outcome (SOMO)

Abstract

AbstractThe potential of real-world data to inform clinical trial design and supplement control arms has gained much interest in recent years. The most common approach relies on matching real-world patient cohorts to clinical trial baseline covariates using propensity score techniques. However, recent studies pointed out that there is a lack of replicability, generalisability, and consensus. Further, few studies consider differences in operational processes. Systematically accounting for confounders, including hidden effects related to the clinical treatment process and clinical trial study protocol, would potentially allow for improved translation between clinical trial and real-world data and enable learning across translational activities.In this paper, we propose an approach that aims to explore and examine these confounders by investigating the impact of selection criteria and operations on the measurements of outcome. We tested the approach on small cell lung cancer patients receiving platinum-based chemotherapy regimens (n=1,224).The results showed that the discrepancy between real-world and clinical trial data potentially depends on differences in covariate characteristics and operations (e.g., censoring mechanism, the process of pre-trial patient selection related to ECOG-performance status 2 patients).This work builds on current approaches and suggests areas of improvement for systematically accounting for differences in outcomes between study cohorts. Continued development of the method presented here could pave the way for transferring learning across studies and developing mutual translation between the real-world and clinical trials to inform future studies design.