Abstract
The extent of effective prostate‐specific antigen (PSA) contamination in the Rotterdam section of the ongoing European Randomized Study of Screening for Prostate Cancer (ERSPC) trial was evaluated and defined as when opportunistic PSA testing of ≥ 3.0 ng/ml was followed by biopsy, similar to the regular procedure within the trial. Records of participants aged 55–74 years at entry were linked to the regional database of the general practitioner (GP) laboratory to obtain PSA tests requested by GPs in the period 1 July 1997 to 31 May 2000 (2.9 years), and to the national pathology database to quantify the number of biopsies. All men randomized were included, only those with prostate cancer screen‐detected or clinically diagnosed before July 1997 were omitted from the analyses. 2,895 out of the 14,349 men (20.2%) in the control arm and 1,981 out of the 14,052 men (14.1%) in the screening arm were PSA‐tested, at an average annual rate of 73 and 52 per 1,000 person‐years, respectively. These rates were higher than those recorded at the national and regional levels, 33 and 38 per 1,000 person‐years, respectively. Opportunistic PSA testing in the control arm reached a peak within the first months of randomization, after which it decreased to around 70 per 1,000 person‐years. An opposite pattern was observed in the screening arm, where participants already had received the scheduled screening within the trial. The proportion of men in the control arm with PSA ≥ 3.0 ng/ml followed by biopsy and prostate cancer was 7–8% and 3%, respectively (3% and 0.4–0.6% in the screening arm), over the whole study period. Over a 4‐year rescreening interval, the average PSA and effective contamination amount were approximately 28% and 10%, respectively. PSA testing in the control arm in the Rotterdam ERSPC section is high, but was not followed by a substantial increase in prostate biopsies. Although the reasons for ordering PSA test or indicating biopsy are unknown, effective PSA contamination in the Rotterdam ERSPC section is low and not likely to jeopardize the power of the trial. © 2003 Wiley‐Liss, Inc.