Search

The international pediatric oncology community has a long history of research collaboration. In the United States, the 2019 launch of the Children's Cancer Data Initiative puts the focus on developing a rich and robust data ecosystem for pediatric oncology. In this spirit, we present here our experience in constructing the Pediatric Cancer Data Commons (PCDC) to highlight the significance of this effort in fighting pediatric cancer and improving outcomes and to provide essential information to those creating resources in other disease areas. The University of Chicago's PCDC team has worked with the international research community since 2015 to build data commons for children's cancers. We identified six critical features of successful data commons design and implementation: (1) establish the need for a data commons, (2) develop and deploy the technical infrastructure, (3) establish and implement governance, (4) make the data commons platform easy and intuitive for researchers, (5) socialize the data commons and create working knowledge and expertise in the research community, and (6) plan for longevity and sustainability. Data commons are critical to conducting research on large patient cohorts that will ultimately lead to improved outcomes for children with cancer. There is value in connecting high-quality clinical and phenotype data to external sources of data such as genomic, proteomics, and imaging data. Next steps for the PCDC include creating an informed and invested data-sharing culture, developing sustainable methods of data collection and sharing, standardizing genetic biomarker reporting, incorporating radiologic and molecular analysis data, and building models for electronic patient consent. The methods and processes described here can be extended to any clinical area and provide a blueprint for others wishing to develop similar resources.

PURPOSE Severe and febrile neutropenia present serious hazards to patients with cancer undergoing chemotherapy. We seek to develop a machine learning–based neutropenia prediction model that can be used to assess risk at the initiation of a chemotherapy cycle. MATERIALS AND METHODS We leverage rich electronic medical records (EMRs) data from a large health care system and apply machine learning methods to predict severe and febrile neutropenic events. We outline the data curation process and challenges posed by EMRs data. We explore a range of algorithms with an emphasis on model interpretability and ease of use in a clinical setting. RESULTS Our final proposed model demonstrates an out-of-sample area under the receiver operating characteristic curve of 0.865 (95% CI, 0.830 to 0.891) in the prediction of neutropenic events on the basis of only 20 clinical features. The model validates known risk factors and offers insight into potential novel clinical indicators and treatment characteristics that elevate risk. It relies on factors that are directly extractable from EMRs, provided a tool can be easily integrated into existing workflows. A cost-based analysis provides insight into optimal risk thresholds and offers a framework for tailoring algorithms to individual hospital needs. CONCLUSION A better understanding of neutropenic risk on an individual level enables a more informed approach to patient monitoring and treatment decisions.