Influence of Publication of US and European Prostate Cancer Screening Trials on PSA Testing Practices

Steven B. Zeliadt ) 0 1 Richard M. Hoffman 0 1 Ruth Etzioni 0 1 John L. Gore 0 1 Larry G. Kessler 0 1 Daniel W. Lin 0 1 0 Excellence, VA Puget Sound Health Care System , 1100 Olive Way, Ste 1400, Seattle, WA 98101 ( 1 Affiliations of authors: Health Services Research and Development Center of Excellence (SBZ) and Department of Urology (DWL), VA Puget Sound Health Care System, Seattle, WA; Department of Health Services (SBZ, RE, LGK) and Department of Urology (JLG, DWL), University of Washington, Seattle, WA; Medicine Service, Department of Veterans Affairs Medical Center, and Department of Medicine, University of New Mexico , Albuquerque , NM (RMH); Public Health Sciences, Fred Hutchinson Cancer Research Center , Seattle, WA (SBZ, RE, JLG, DWL) In 2009, results from the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial indicated no difference in mortality between the screening and the control groups (rate ratio = 1.13, 95% confidence interval = 0.75 to 1.70), whereas those from the European Randomized study of Screening for Prostate Cancer trial indicated a 20% reduction in mortality among the screening group (rate ratio = 0.80, 95% confidence interval = 0.65 to 0.98). In this study, we examined whether prostate-specific antigen (PSA) testing has changed following these publications. The primary outcome measure was the proportion of men seen at least once in a primary care or urology clinic between August 1, 2004, and March 31, 2010, who received a PSA test. Following the publications, PSA use declined slightly-by 3.0 percentage points and 2.7 percentage points among men aged 40-54 and 55-74 years, respectively. PSA testing among men older than 75 years initially declined slightly following the recommendations by the US Preventive Services Task Force in 2008 and continued to decline after the trial publications. - In the last two decades, limited evidence about the benefit of prostate-specific antigen (PSA) screening to prevent deaths from prostate cancer, relative to potential harm from screening, has led to mixed endorsements of PSA testing. Some clinical organizations including American Cancer Society (ACS) and American Urological Association (AUA) recommend screening, whereas others including the US Preventive Services Task Force (USPSTF) advise patients to discuss the potential benefits and risks of the PSA test with their physicians (13). Despite limited evidence and consensus regarding the benefits of PSA testing, it has become a widely adopted screening test in the United States with more than 75% of men older than 50 years having been tested at least once (4). Notably, PSA testing has been the highest among older men. Longitudinal patterns of testing in the 1990s suggest that 45% of white men aged 7079 years received annual PSA tests compared with 35% of men aged 5059 years (5). This may reflect overuse of PSA testing as most recommendations guard against screening men with less than a 10-year life expectancy because of the increased likelihood of experiencing harm from screening relative to any potential benefit (6). Two recent sets of publications received broad media coverage, potentially influencing the way PSA testing is viewed among clinicians and the general public. First, in August 2008, the USPSTF released updated recommendations suggesting an upper age limit of 75 years for PSA screening, citing accumulating evidence of high rates of screening among elderly men for whom a small benefit is likely. Second, in March 2009, the long-awaited results of the Prostate, Lung, Colorectal and Ovarian (PLCO) and the European Randomized study of Screening for Prostate Cancer (ERSPC) screening trials were simultaneously published, generating considerable attention because of their mixed findings (7,8). The PLCO trial reported similar death rates in both the screening and the control groups (rate ratio = 1.13, 95% confidence interval = 0.75 to 1.70), whereas the ERSPC reported a 20% reduction in prostate cancerspecific deaths with screening (rate ratio = 0.80, 95% confidence interval = 0.65 to 0.98). To explore whether these publications have begun to influence the patterns of PSA testing, we conducted a cross-sectional study from August 2004 to March 2010 to assess the frequency of PSA testing in the Veterans Health Administration (VHA) Pacific Northwest Network. To specifically examine changes in PSA testing patterns potentially associated with the USPSTF guidelines and the publication of the trial results (7,8), we focused on the 8-month period starting in August 2008, when the USPSTF guidelines were published, and ending in March 2009, when the trial results were published. We reviewed testing in four periods before the USPSTF recommendation update was released (period 1: August 2004March 2005; period 2: August 2005March 2006; period 3: August 2006 March 2007; period 4: August 2007March 2008). In period 5 (August 2008March 2009), we examined testing after the USPSTF update, and finally, in period 6 (August 2009 March 2010), we examined PSA testing patterns after the publication of the screening trials (7,8). During each of the six periods, we calculated the proportion of eligible men who had a PSA test (numerator) from among all eligible men with a primary care or urology clinic visit (denominator). We excluded men who were diagnosed with prostate cancer before the date of the first clinic visit in each period. The total number of eligible men ranged between 125 000 and 140 000 in each period (Table 1). We observed a slightly increased trend in PSA testing for men aged 4054 and 5574 years before the publication of the trials (7,8), with minimal changes in testing among men aged 75 years and older. The updated USPSTF guidelines in 2008 did not appear to correspond to a change in PSA testing for men younger than 75 years, as PSA testing rates continued to increase slightly for all men aged 4074 years. Patient age, y Period No. (calendar period) Period 6 vs periods 1-5 combined 25 495 (24.2) 23 953 (24.6) 24 713 (25.7) 25 869 (25.4) 25 514 (24.3) 26 384 (22.1) P < .001 Z statistic = 29.61 P = .842 P < .001 * This study period was after publication of updated US Preventive Services Task Force guidelines. This study period was after the publication of the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial and the European Randomized study of Screening for Prostate Cancer trial (7,8). Two-sample Z test of proportions (two-sided). However, we observed a decrease in testing among men aged 75 years and older from 25.4% in the period just before the USPSTF update to 24.3% in the period just after the USPSTF update (P = .036). We observed a decrease in PSA testing among all three age groups, 4054, 5574, and 75 years, by 3.0 percentage points, 2.7 percentage points, and 2.2 percentage points, respectively, following the publication of the PLCO and ERSPC trial results (7,8). This trend was statistically significant when comparing th (...truncated)