Timo Soeterik

47 Active Surveillance: Follow-up Statistical analysis Possible significant differences in mean values between hospitals were assessed using using one-way analysis of variance. We evaluated differences in the proportions of patients using the Fisher’s Exact Test. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using univariable cox-regression. Analysis was performed using SPSS for Windows v.24.0 (IBM Corp., Armonk, NY, USA). A p -value of <0.05 was considered significant. RESULTS Study population A total number of 1181 patients were diagnosed with PCa and included for AS between January 1, 2008 and December 31, 2014 at six Santeon hospitals. This included an initial 181/1181 (15%) of patients with incidental tumours (cT1a/b), who were later excluded. Baseline characteristics for the 1000 patients with screen-detected PC on AS are presented in Table 1. The table shows the differences between hospitals with regard to baseline serum PSA, PSAD, number of positive biopsy cores and proportions of PRIAS-eligible and PRIAS-ineligible patients. PSA follow-up testing A total of 958 patients had treatment-free follow-up of more than 6 mo. The percentages of patients receiving PRIAS-concordant and -discordant PSA testing are presented in Figure 1. The variation between hospitals was considerable. Hospital 1 had the highest compliance, as 83% of patients had PRIAS-concordant PSA monitoring. The least strict monitoring occurred in hospital 4, where only 55% of the patients had PRIAS-concordant PSA monitoring. Overall, 706/958 patients (74%) had PRIAS-concordant PSA monitoring. The proportion of patients with a PSA doubling time (PSADT) of <3 yr did not differ significantly between the groups with discordant and concordant PSA follow-up (42% vs 48%; p = 0.156). The group of patients with discordant PSA testing was slightly younger (mean age 66.8 vs 68.7 yr; p < 0.001). 3