MP32-18 A COMPARISON BETWEEN UROLOGY AND OTHER SPECIALITIES IN THE UTILIZATION OF TWITTER AT INTERNATIONAL CONFERENCES

Abstract

diagnostic prostate-biopsy, as well as treatment for early-stage prostate-cancer, in Israel. METHODS: This is a population-based cohort study, using Clalit electronic-databases. Clalit is the largest HMO in Israel and insures over 4,000,000 subjects. Our cohort included all 220,861 Clalit male patients who were 75 or older in January 1st 2009. We recorded all PSA tests, diagnostic biopsies, and treatments for early-stage prostate-cancer, between the years 2009-2013. We performed a uni and multivariable analyses to determine patient and doctor related variables that were associated with PSA testing among elderly patients. RESULTS: Excluding patients who had a prior prostate-cancer diagnosis our final cohort included 213,814 subjects. During the study period, 187,228 PSA tests were performed on 55,886 subjects (26.1%). The median PSA was 2.15 ng/ml (IQR 0.92, 4.7). PSA tests were common even among subjects older than 85, with 10,594 of 79,731 subjects tested (13.3%). Prostate biopsies were performed on 1913 subjects. Of those, 818 were then diagnosed with prostate-cancer. One month following prostate biopsies we found 103 associated emergency department visits and even 4 deaths. On Multivariable analysis age, marital status and type of health insurance were independent predictors of PSA testing. Surprisingly, Charleston comorbidity was also an independent predictor (a higher Charleston score was associated with an increased risk of PSA testing). The vast majority of PSA tests were performed by family physicians (76%). The number of years of practice was the only Doctor associated predictors of PSA testing. Performing other screening tests was not associated with PSA testing. CONCLUSIONS: We have demonstrate a large overuse of PSA testing among elderly in Israel, even among those older than 85. As appose to PSA testing, which may be considered harmless, biopsies are defiantly not. Our study demonstrates that biopsies are also commonly performed and rarely may cause severe consequences. We are amongst the first studies to test doctor associated predictors. Future interventional studies are required to try and mitigate this phenomenon.