Machine learning for identifying Randomized Controlled Trials: An evaluation and practitioner's guide

TLDR

Machine learning algorithms are highly accurate for identifying randomized controlled trials, yet their practical adoption remains limited because optimal workflow integration is unclear. The study evaluates ML models for RCT classification and provides practical guidance on their use in systematic and rapid reviews, including recommended probability cutoffs and open‑source software. The authors trained and optimized support vector machine and convolutional neural network models on the Cochrane Crowd RCT set, then evaluated them on the external Clinical Hedges dataset, comparing performance to traditional database search filters and reporting AUROC metrics. ML approaches outperformed traditional database search filters at all sensitivity levels, with the best model achieving an AUROC of 0.987 (95% CI 0.984–0.989), the highest reported for this task.

Abstract

Machine learning (ML) algorithms have proven highly accurate for identifying Randomized Controlled Trials (RCTs) but are not used much in practice, in part because the best way to make use of the technology in a typical workflow is unclear. In this work, we evaluate ML models for RCT classification (support vector machines, convolutional neural networks, and ensemble approaches). We trained and optimized support vector machine and convolutional neural network models on the titles and abstracts of the Cochrane Crowd RCT set. We evaluated the models on an external dataset (Clinical Hedges), allowing direct comparison with traditional database search filters. We estimated area under receiver operating characteristics (AUROC) using the Clinical Hedges dataset. We demonstrate that ML approaches better discriminate between RCTs and non‐RCTs than widely used traditional database search filters at all sensitivity levels; our best‐performing model also achieved the best results to date for ML in this task (AUROC 0.987, 95% CI, 0.984‐0.989). We provide practical guidance on the role of ML in (1) systematic reviews (high‐sensitivity strategies) and (2) rapid reviews and clinical question answering (high‐precision strategies) together with recommended probability cutoffs for each use case. Finally, we provide open‐source software to enable these approaches to be used in practice.