Identification of patients with metastatic castration-sensitive or metastatic castration-resistant prostate cancer using administrative health claims and laboratory data
Current Medical Research and Opinion, 2021
Objective
To develop algorithms to identify metastatic castration-sensitive prostate cancer (mCSPC) patients and castration-resistant prostate cancer (mCRPC) patients, using health claims data and laboratory test results.
Methods
A targeted literature review summarized mCSPC and mCRPC patient selection criteria previously used in real-world retrospective studies. Novel algorithms to identify mCSPC and mCRPC were developed based on diagnosis codes indicating hormone sensitivity/resistance, prostate-specific antigen (PSA) test results, and claims for castration and mCRPC-specific treatments. These algorithms were applied to claims data from Optum Clinformatics Extended DataMart (Date of Death) Databases (commercial insurance/Medicare Advantage [COM/MA]; 01 January 2014-31 July 2019) and Medicare Fee-for-Service (Medicare-FFS; 01 January 2014-31 December 2017).
Results
Previous real-world studies identified mCSPC primarily based on metastasis diagnosis codes, and mCRPC based on mCRPC-specific drugs. Using the current study's algorithms, 7034 COM/MA and 19,981 Medicare-FFS patients were identified as having mCSPC, and 2578 COM/MA and 11,554 Medicare-FFS as having mCRPC. Most mCSPC patients were identified based on evidence of being hormone/castration-naive. Patients were identified as having mCRPC most commonly based on rising PSA (COM/MA), or at the metastasis diagnosis date if it occurred after castration (Medicare-FFS). Among patients with mCSPC, 14-17% had evidence of progression to castration resistance during a median 1-year follow-up period, mostly based on use of mCRPC-specific drugs.
Conclusions
Comprehensive algorithms based on claims and laboratory data were developed to identify and distinguish patients with mCSPC and mCRPC. This will facilitate appropriate identification of mCSPC and mCRPC patients based on health claims data and better understanding of patient unmet needs in real-world settings.
Authors
Freedland SJ, Ke X, Lafeuille MH, Romdhani H, Kinkead F, Lefebvre P, Petrilla A, Pulungan Z, Kim S, D'Andrea DM, Francis P, Ryan CJ