Data Challenges for Externally Controlled Trials: Viewpoint

Abstract

The preferred evidence of a large randomized controlled trial is difficult to adopt in scenarios, such as rare conditions or clinical subgroups with high unmet needs, and evidence from external sources, including real-world data, is being increasingly considered by decision makers. Real-world data originate from many sources, and identifying suitable real-world data that can be used to contextualize a single-arm trial, as an external control arm, has several challenges. In this viewpoint article, we provide an overview of the technical challenges raised by regulatory and health reimbursement agencies when evaluating comparative efficacy, such as identification, outcome, and time selection challenges. By breaking down these challenges, we provide practical solutions for researchers to consider through the approaches of detailed planning, collection, and record linkage to analyze external data for comparative efficacy.

Augmenting external control arms using Bayesian borrowing: a case study in first-line non-small cell lung cancer

“…The applications and challenges of working with RWD for ECA analyses are well understood and have been discussed extensively in the literature [ 9 , 49–51 ] – including many of the limitations encountered in this study. However, previous evaluations of the performance of ECA analyses at replicating RCTs in metastatic NSCLC have shown the method to be viable [ 14 , 49 , 51 , 52 ].…”

Examining external control arms in oncology: A scoping review of applications to date

“…These data types, particularly non‐clinical trial data, although they reflect real‐world clinical practice, they might not guarantee the controlled environment typical of RCTs, potentially introducing heterogeneity and bias. 4 , 24 , 62 With respect to matching control data to the intervention group, 26% (6/23) of the studies applied propensity score weighting for adjustments, indicating a focus on achieving balanced and comparable patient cohorts. However, nearly 48% (11/23) of studies did not provide clear information on covariate adjustments—this concern has been underscored in the existing literature as it relates to the failure in accounting for baselines imbalances.…”

“…Contrarily, 17% (4/23) of the studies relied on administrative health databases, while another 17% (4/23) utilized electronic medical records as their data sources. These data types, particularly non‐clinical trial data, although they reflect real‐world clinical practice, they might not guarantee the controlled environment typical of RCTs, potentially introducing heterogeneity and bias 4,24,62 . With respect to matching control data to the intervention group, 26% (6/23) of the studies applied propensity score weighting for adjustments, indicating a focus on achieving balanced and comparable patient cohorts.…”