LAUSR

Background Increasing evidence indicates an association between statins and reduced prostate cancer-specific mortality (PCSM). However, significant bias may exist in these studies. One particularly challenging bias to assess is the healthy user effect, which may be quantified by screening patterns. We aimed to evaluate the association between statin use, screening, and PCSM in a dataset with detailed longitudinal information. Methods We used the Veterans Affairs Informatics and Computing Infrastructure to assemble a cohort of patients diagnosed with prostate cancer (PC) between 2000 and 2015. We collected patient-level demographic, comorbidity, and tumor data. We also assessed markers of preventive care utilization including cholesterol and prostate specific antigen (PSA) screening rates. Patients were considered prediagnosis statin users if they had at least one prescription one or more years prior to PC diagnosis. We evaluated PCSM using hierarchical Fine-Gray regression models and all-cause mortality (ACM) using a cox regression model. Results The final cohort contained 68,432 men including 40,772 (59.6%) prediagnosis statin users and 27,660 (40.4%) nonusers. Prediagnosis statin users had higher screening rates than nonusers for cholesterol (90 vs. 69%, p  < 0.001) and PSA (76 vs. 67%, p  < 0.001). In the model which excluded screening, prediagnosis statin users had improved PCSM (SHR 0.90, 95% CI 0.84–0.97; p  = 0.004) and ACM (HR 0.96, 95% CI 0.93–0.99; p  = 0.02). However, after including cholesterol and PSA screening rates, prediagnosis statin users and nonusers showed no differences in PCSM (SHR 0.98, 95% CI 0.91–1.06; p  = 0.59) or ACM (HR 1.02, 95% CI 0.98–1.05; p  = 0.25). Conclusion We found that statin users tend to have more screening than nonusers. When we considered screening utilization, we observed no relationship between statin use before a prostate cancer diagnosis and prostate cancer mortality.