🤖 AI Summary
Predicting patient enrollment in clinical trials is a critical challenge for trial planning. This paper proposes a deep probabilistic forecasting model that jointly leverages unstructured protocol text and structured site-level features. It employs a pretrained language model to encode trial protocols and constructs tabular representations from site characteristics; cross-modal attention aligns textual and tabular embeddings. A Gamma-distributed output layer explicitly models the Poisson–Gamma enrollment process, enabling principled uncertainty quantification and prediction interval estimation. Evaluated on real-world multicenter clinical trial data, the model significantly outperforms conventional statistical methods and end-to-end deep learning baselines—achieving superior accuracy and reliability in both enrollment count and timeline predictions. The approach delivers interpretable, robust decision support for trial resource allocation and operational planning.
📝 Abstract
Clinical trials are a systematic endeavor to assess the safety and efficacy of new drugs or treatments. Conducting such trials typically demands significant financial investment and meticulous planning, highlighting the need for accurate predictions of trial outcomes. Accurately predicting patient enrollment, a key factor in trial success, is one of the primary challenges during the planning phase. In this work, we propose a novel deep learning-based method to address this critical challenge. Our method, implemented as a neural network model, leverages pre-trained language models (PLMs) to capture the complexities and nuances of clinical documents, transforming them into expressive representations. These representations are then combined with encoded tabular features via an attention mechanism. To account for uncertainties in enrollment prediction, we enhance the model with a probabilistic layer based on the Gamma distribution, which enables range estimation. We apply the proposed model to predict clinical trial duration, assuming site-level enrollment follows a Poisson-Gamma process. We carry out extensive experiments on real-world clinical trial data, and show that the proposed method can effectively predict the number of patients enrolled at a number of sites for a given clinical trial, outperforming established baseline models.