From Data to Duty: An Informatics Blueprint for Pediatric Oncology Trials.
Researchers
Kirk D Wyatt, Andrew D Tran, Suzi Birz, Michael Watkins, Jeremy L Warner, Heather Wasserstrom, Grace E Rossi, Elyssa Rubin, Charles A Phillips, Angela J Waanders, David H Noyd, Nicole Wood, Reuven J Schore, Natasha Alexander, Brian Furner, Wayne H Liang, Jennifer Wilkes, Yang Yang Ding, Ajay Perumbeti, Tamara P Miller, Karen Lewing, Gerard Hills, Colin Moore, Ksenya Shliakhtsitsava, Tyler J Benning, Anna Keller, Emily M Webb, Hope Humann, Emily Larimer, Harry Hochheiser, Adam P Yan, Amir Mian, Amanda Haddock, Parker Moss, Molly S Talman, Lynda M Villagomez, Subhashini Jagu, Allen R Chen, Sara M Dryden, Richard Aplenc, Samuel L Volchenboum
Abstract
The data generated by cooperative-group clinical trials have transformed outcomes for children with cancer, yet the infrastructure underlying their collection and use has not kept pace with advances in informatics and data science. Data collection on pediatric oncology trials remains largely manual: Clinical research assistants abstract values from medical records and key them into electronic case report forms, a process that is costly, slow, inconsistent, and error-prone. Data are often siloed within trial networks for years after a study closes, limiting their value for secondary research, particularly for rare cancers. In this consensus blueprint, an interdisciplinary group of pediatric oncologists, clinical informaticians, data scientists, nurses, clinical research assistants, drug developers, and parent advocates outlines a pragmatic path forward across three components of the data lifecycle: collection, processing, and sharing. We argue that case report forms should be bound to standardized common data elements; that automated data abstraction from electronic health records, including via large language models with human-in-the-loop validation, should replace much of today's manual abstraction; that paper chemotherapy roadmaps should be reconceived as computable artifacts; and that clinical trials data should be deposited in publicly accessible repositories with appropriate governance as soon as practically feasible. We discuss facilitators and barriers within each component, including standards adoption, regulatory complexity, financial sustainability, and workforce capacity. We propose concrete actions for funders, clinical trials networks, professional societies, and industry. The technology to achieve these goals exists today, and widespread adoption requires coordinated leadership, sustainable funding models, and a cultural shift that recognizes data stewardship as an ethical imperative and a duty to the children and families we serve.Source: PubMed (PMID: 43062204)View Original on PubMed