Geographic variation in patient surveillance after radical prostatectomy

Background: Prostate cancer is often diagnosed early enough in its clinical course to permit radical prostatectomy to be done with curative intent, ye t many patients experience tumor recurrence. Most patients receive postoper ative surveillance, but the intensity of testing varies appreciably. We sou ght to evaluate the influence of geographic location on the variability of surveillance intensity. Methods: Questionnaires pertaining to postoperative surveillance were maile d to 4467 members of the American Urological Association (AUA). Practice pa ttern variation was assessed among 24 large metropolitan statistical areas, among nine United States census regions, and by health maintenance organiz ation penetration rate. Results: Of 4467 urologists surveyed, 1416 (32%) responded and 1050 (24%) r esponses were evaluable. Correlation analysis showed that mean follow-up in tensity across modalities surveyed was highly correlated across tumor, node , metastasis (TNM) stages and years postsurgery. We found no significant ma in effects attributable to metropolitan statistical area, United Stares (US ) census region, or health maintenance organization (HMO) penetration rate for commonly used surveillance modalities: serum prostate-specific antigen (PSA), office visit, and urinalysis. For infrequently used modalities, ther e were minimal effects on testing intensity of US census region, metropolit an statistical area, and HMO penetration rate. Few two-way and three-way in teractions were significant. Conclusions: The utilization of commonly used surveillance modalities by ur ologists caring for patients after radical prostatectomy is not affected by metropolitan statistical area, US census region, or HMO penetration rate.