Introduction: Active surveillance (AS) is recognized as a safe management option for low-risk prostate cancer, however, its utilization varies. We aim to characterize variability in AS utilization in the US and evaluate patient- and county- characteristics associated with the use of AS in men with low-risk prostate cancer.
Methods: We identified men diagnosed with localized prostate cancer in the Surveillance, Epidemiology, and End Results Prostate with Watchful Waiting (WW) database linked with the County Area Health Resource File for detailed county-level demographics and provider distribution data. Clinical and demographic variables were included in 3-level hierarchical mixed-effect logistic regression models to evaluate clustered random regional variation on utilization of AS/WW.
Results: The study cohort was comprised of 81,071 men with low risk disease by D’Amico criteria. Mean age at diagnosis was 63.1 years (SD 7.9). 14.1% of the cohort was Black, 9.4% Hispanic, and 4% Asian/Pacific Islander. Mean PSA at diagnosis was 5.5 ng/ml (SD 1.9). 93.8% were insured. Most (83.4%) had cT1c, mean percent cores positive of 27.6% (SD 20.1). Mean follow-up was 38.6 months (SD 20.4). The proportion of men in the AS/WW group decreased from 21.4% in 2010 to to 12.3% in 2015.
Registry region accounted for 38.4% of the variation in AS/WW utilization in the population. The odds of AS/WW increased with age (>80 years, OR 2.43, 95% CI 2.08-2.84 vs age<50 years) and decreased with greater PPC (>33% and <50%, OR 0.46, 95% CI 0.43-0.48; >50%, OR 0.30, 95% CI 0.28-0.32). Hispanic men were less likely to receive AS/WW compared to white men (OR 0.80, 95% CI 0.74-0.85) while the odds for black men did not differ significantly from white men (p=0.92). Compared to heavily populated metropolitan areas (>1 million people), men residing in all other city types were less likely to be managed with AS/WW. County-level socioeconomic characteristics such as household income and education, and measures of medical resources such as number of urologists, radiation oncologists, PCPs, and hospital beds were not associated with greater odds of AS/WW over other management strategies.
Conclusions: Overall, regional characteristics accounted for one-third of the variation in AS/WW utilization. Greater county-level densities of medical resources were not associated with increased odds of AS/WW over active treatment. Â Greater concordance with AS guidelines is needed. Source of
