| Abstract: | Background: In Asian populations, prostate cancer mortality (PCM) has been increasing with the increasing use of PSA testing. Objectives: 1) To establish prevalence of elevated PSA at different cutpoints in Taiwan and to assess the predictive relationship of PSA with PCM. 2) To compare PCM between those with PSA tested and not tested from a cohort study. Methods: Among the cohort, 231,937, who received standard medical screening examination between 1994 and 2006 at a private screening clinic, 46,951 subjects, aged at least 50 years old, had PSA test results. Age and education were adjusted to those of Taiwan to arrive at national PSA prevalence. PCM was compared between the PSA untested (38,596) and PSA tested group (24,176), with the latter including those with a self-reported history of prostate cancer (1,675). Results: National prevalence of PSA was 8.8% (threshold at 4) and 1.9% (threshold at 10). With national mortality rate at 30.1/100,000, the chance of dying in ten years was 3.4%, 15.8% or 27.9% among those above threshold at 4, 10, or 15, respectively, yielding false positives rates, with mortality as the outcome, 96.4%, 84.2% or 72.1%, and false negative rates, 12%, 33% or 50%, respectively. Having PSA tested, when compared with not tested, was not associated with decreased mortality from prostate cancer, even though PSA tested group, a higher educational class, had significantly decreased mortality from all cause. Increasing PSA levels were associated with increasing HRs. Conclusion: PSA value above standard threshold of 4 is common in this Asian population (one out of 11 for age 50 and older), but only one in 29 tested positive was expected to die from prostate cancer within the next 10 years. This 10-year probability increased when threshold was 10 (1/6) but was still low. The high false positive rate of PSA (97%) and the harm associated with intervention among Asians should be communicated to the elderly patients, in view of the particularly low mortality rate of this cancer. |
