Accepted for/Published in: JMIR Medical Informatics
Date Submitted: Apr 12, 2024
Date Accepted: Oct 10, 2024
Medication Prescription Policy for US Veterans with Metastatic Castration-Resistant Prostate Cancer: Causal Machine Learning Approach
ABSTRACT
Background:
Metastatic castration-resistant prostate cancer (mCRPC) is an advanced stage of prostate cancer with a high risk of mortality. Though ARTAs are preferred therapies for their less toxic profile and improved survival, no study has used machine learning with patients’ demographics, test results, and comorbidities to identify heterogeneous treatment rules that might improve the survival of patients with mCRPC.
Objective:
To measure patient-level heterogeneity in the association of medication prescribed with overall survival and arrive at a set of medication prescription rules using patient demographics, test results, and comorbidities.
Methods:
The dataset included 3675 veterans treated for mCRPC who were prescribed either abiraterone or enzalutamide from 2014 to 2017, with follow-up to 2020. A causal survival forest was used to estimate the treatment effect of the medication prescribed conditional on patient demographics, test results, and comorbidities. The unit level of analysis was the patient, and the main outcome of this study was follow-up days indicating survival days while on the medication.
Results:
Enzalutamide is associated with an average of 40.15 (95% CI, 12.61 to 67.69) more days of survival than Abiraterone. The increase in overall survival for the two drugs varied across patient demographics, test results, and comorbidities. Race, Prostate-Specific Antigen (PSA), abnormal gait, vision, and age were the top five identified covariates contributing to the high variations. The analysis resulted in a set of rules based on a few covariates that permit patient-specific prescriptions for the two ARTAs.
Conclusions:
In this study of 3675 veterans, findings showed evidence for heterogeneity, i.e., based on the medication prescribed, survival days could be improved for certain mCRPC patients. Findings suggest that it is possible to arrive at a set of rules for the two ARTAs based on patient demographics, test results, and comorbidities to optimize treatment decisions to improve survival.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.