Overtreatment of prostate cancer in Florida.

Abstract

140 Background: Although most evidence suggests that older men with localized prostate cancer do not benefit from treatment, many men with relatively limited (<10 years) life expectancy continue to seek and receive treatment. Given mounting concerns regarding overtreatment, we sought to better understand geographic differences in the potential overtreatment of prostate cancer. Methods: We used the Florida Cancer Data System (FCDS) to identify treated and untreated cases of prostate cancer between 2004-2007. The FCDS is a large statewide cancer registry that collects clinical, demographic and treatment information for incident cancers. We defined potential overtreatment according to risk of disease (Gleason <=7, stage T2 and less) and age ( >75 years). We used Geographic Information Systems (GIS) to map differences in overtreatment across Health Service Areas (HSAs) in Florida. We then examined the relationship between the number of urologists and radiation oncologists reported in the Area Resource File and potential overtreatment using linear regression models. All analyses were performed using standard statistical software (SAS and ArcGIS). Results: Among men older than 75 years (n=10,411), 9.2% received surgery, 38.7% received external radiation therapy, 15% received brachytherapy and 10.5% received primary androgen deprivation therapy; only 26.6% of patients were managed expectantly. Overtreatment varied significantly across Florida. Among some HSA clusters in Central and Panhandle Florida, overtreatment was estimated in less than 21.6 % of cases. Much higher rates of overtreatment (36-52% of cases) was observed among other HSAs, particularly in South Florida. Although the number of urologists (per 100,000 male population) was not associated with an increased propensity for overtreatment in regression models, the concentration of radiation oncologists was associated with an increase of 0.64 cases of overtreatment (per 100,000 men) for each additional radiation oncologist (p<0.0001). Conclusions: Treatment among older men with low- and intermediate-risk localized prostate cancer is more common than expectant management. The propensity for overtreatment varies considerably according to geographic and healthcare market factors.