Baijiang Yuan, Muammar Kabir, Jiang Chen He, Yuchen Li, Benjamin Grant, Sharon Narine, Mattea Welch, Sho Podolsky, Ning Liu, Rami Ajaj, Luna Jia Zhan, Aly Fawzy, Janine Xu, Yuhua Zhang, Vivien Yu, Wei Xu, Rahul G Krishnan, Steven Gallinger, Kelvin K W Chan, Monika K Krzyzanowska, Tran Truong, Geoffrey Liu, Robert C Grant
{"title":"Development of Machine Learning Systems to Predict Cancer-Related Symptoms With Validation Across a Health Care System.","authors":"Baijiang Yuan, Muammar Kabir, Jiang Chen He, Yuchen Li, Benjamin Grant, Sharon Narine, Mattea Welch, Sho Podolsky, Ning Liu, Rami Ajaj, Luna Jia Zhan, Aly Fawzy, Janine Xu, Yuhua Zhang, Vivien Yu, Wei Xu, Rahul G Krishnan, Steven Gallinger, Kelvin K W Chan, Monika K Krzyzanowska, Tran Truong, Geoffrey Liu, Robert C Grant","doi":"10.1200/CCI-25-00073","DOIUrl":null,"url":null,"abstract":"<p><strong>Purpose: </strong>Cancer and its treatment cause symptoms. In this study, we aimed to develop machine learning (ML) systems that predict future symptom deterioration among people receiving treatment for cancer and then validate the systems in a simulated deployment across an entire health care system.</p><p><strong>Methods: </strong>We trained and tested ML systems that predict a deterioration in nine patient-reported symptoms within 30 days after treatments for aerodigestive cancers, using internal electronic health record (EHR) data at Princess Margaret Cancer Centre (3,229 patients; 20,267 treatments). The primary performance metric was the area under the receiver operating characteristic curve (AUROC). The best-performing systems in the held-out internal test set were then externally validated across 82 cancer centers in Ontario (12,079 patients; 77,003 treatments) by adapting techniques from meta-analysis.</p><p><strong>Results: </strong>The best ML systems predicted symptom deterioration with AUROCs ranging from 0.66 (95% CI, 0.63 to 0.69) for dyspnea to 0.73 (95% CI, 0.71 to 0.75) for drowsiness in the internal test cohort. Treatments flagged as high-risk were significantly associated with future symptom deterioration (odds ratios [ORs], 2.53-6.56; all <i>P</i> < .001) and emergency department visits for dyspnea (OR, 1.85; <i>P</i> = .008), depression (OR, 1.84; <i>P</i> = .04), and anxiety (OR, 2.66; <i>P</i> < .001). In the external validation cohort, the AUROCs for different symptoms meta-analyzed across centers ranged from 0.67 (95% CI, 0.66 to 0.68) to 0.73 (95% CI, 0.72 to 0.74). Performance across centers displayed significant heterogeneity for six of nine symptoms (I<sup>2</sup>, 46.4%-66.9%; <i>P</i> = .004 for dyspnea, <i>P</i> < .001 for the rest).</p><p><strong>Conclusion: </strong>ML can predict future symptoms among people with cancer from routine EHR data, which could guide personalized interventions. Heterogeneous performance across centers must be considered when systems are deployed across a health care system.</p>","PeriodicalId":51626,"journal":{"name":"JCO Clinical Cancer Informatics","volume":"9 ","pages":"e2500073"},"PeriodicalIF":2.8000,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12487662/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"JCO Clinical Cancer Informatics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1200/CCI-25-00073","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/9/25 0:00:00","PubModel":"Epub","JCR":"Q2","JCRName":"ONCOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Purpose: Cancer and its treatment cause symptoms. In this study, we aimed to develop machine learning (ML) systems that predict future symptom deterioration among people receiving treatment for cancer and then validate the systems in a simulated deployment across an entire health care system.
Methods: We trained and tested ML systems that predict a deterioration in nine patient-reported symptoms within 30 days after treatments for aerodigestive cancers, using internal electronic health record (EHR) data at Princess Margaret Cancer Centre (3,229 patients; 20,267 treatments). The primary performance metric was the area under the receiver operating characteristic curve (AUROC). The best-performing systems in the held-out internal test set were then externally validated across 82 cancer centers in Ontario (12,079 patients; 77,003 treatments) by adapting techniques from meta-analysis.
Results: The best ML systems predicted symptom deterioration with AUROCs ranging from 0.66 (95% CI, 0.63 to 0.69) for dyspnea to 0.73 (95% CI, 0.71 to 0.75) for drowsiness in the internal test cohort. Treatments flagged as high-risk were significantly associated with future symptom deterioration (odds ratios [ORs], 2.53-6.56; all P < .001) and emergency department visits for dyspnea (OR, 1.85; P = .008), depression (OR, 1.84; P = .04), and anxiety (OR, 2.66; P < .001). In the external validation cohort, the AUROCs for different symptoms meta-analyzed across centers ranged from 0.67 (95% CI, 0.66 to 0.68) to 0.73 (95% CI, 0.72 to 0.74). Performance across centers displayed significant heterogeneity for six of nine symptoms (I2, 46.4%-66.9%; P = .004 for dyspnea, P < .001 for the rest).
Conclusion: ML can predict future symptoms among people with cancer from routine EHR data, which could guide personalized interventions. Heterogeneous performance across centers must be considered when systems are deployed across a health care system.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
