Deep learning-based risk prediction for interventional clinical trials based on protocol design: A retrospective study

Sohrab Ferdowsi; Julien Knafou; Nikolay Borissov; David Vicente Alvarez; Rahul Mishra; Poorya Amini; Douglas Teodoro

doi:10.1016/j.patter.2023.100689

Deep learning-based risk prediction for interventional clinical trials based on protocol design: A retrospective study

Patterns (N Y). 2023 Feb 10;4(3):100689. doi: 10.1016/j.patter.2023.100689. eCollection 2023 Mar 10.

Authors

Sohrab Ferdowsi^{1

2}, Julien Knafou², Nikolay Borissov^{3

4}, David Vicente Alvarez^{1

2}, Rahul Mishra¹, Poorya Amini^{3

4}, Douglas Teodoro^{1

2

5}

Affiliations

¹ Department of Radiology and Medical Informatics, University of Geneva, Geneva, Switzerland.
² Geneva School of Business Administration, HES-SO University of Applied Sciences and Arts of Western Switzerland, Geneva, Switzerland.
³ Clinical Trials Unit, University of Bern, Bern, Switzerland.
⁴ Risklick AG, Bern, Switzerland.
⁵ Swiss Institute of Bioinformatics, Lausanne, Switzerland.

Abstract

Success rate of clinical trials (CTs) is low, with the protocol design itself being considered a major risk factor. We aimed to investigate the use of deep learning methods to predict the risk of CTs based on their protocols. Considering protocol changes and their final status, a retrospective risk assignment method was proposed to label CTs according to low, medium, and high risk levels. Then, transformer and graph neural networks were designed and combined in an ensemble model to learn to infer the ternary risk categories. The ensemble model achieved robust performance (area under the receiving operator characteristic curve [AUROC] of 0.8453 [95% confidence interval: 0.8409-0.8495]), similar to the individual architectures but significantly outperforming a baseline based on bag-of-words features (0.7548 [0.7493-0.7603] AUROC). We demonstrate the potential of deep learning in predicting the risk of CTs from their protocols, paving the way for customized risk mitigation strategies during protocol design.

Keywords: clinical trials; deep learning; graph neural networks; neural language models; risk prediction; text classification; text mining; transformer-based language models.