Freely Available Prostate Specific Antigen Testing in a Population: Testing Patterns and Outcomes on Prostate Cancer THE SASKATCHEWAN EXPERIENCE
Methods: Individual records were retrieved from the two labs in Saskatchewan capable of analyzing PSA serum samples. The PSA data represented almost all PSA tests in the population for the five-year period 1997-2001. The PSA data included date of the PSA test, a unique identifier of the men tested, test results including total PSA and the free-PSA amount (for November 1999 to December 2001), and an ID of the physician who ordered the test. This data was linked to the population-based Saskatchewan Cancer Registry to determine who had a previous or subsequent prostate cancer diagnosis and to secure tumour characteristics and clinical management data. This combined data was then linked to Saskatchewan Health data files to obtain information about biopsy procedures and to determine the geographic residence of men at the time of their PSA tests. De-identified data was returned for descriptive analysis.
Results: Over 60% of men aged 50 and over had at least one PSA test during 1997 to 2001. Even among men 40-49, 27% had at least one PSA test and there were over 5,300 tests done in men under 40 years of age. Sixteen percent of men 40-49 who had PSA tests had more than one and this percentage increased with age to 59.4% for men in their 70s. Over 80% of all PSA tests were ordered by general practitioners and there were significant geographic variations in testing patterns. Knowledge of the free-PSA-ratio, which began in 1999, reduced biopsy rates 4.7% and increased cancer detection 8.7% for men with total PSA test results in the 4.0-10.0ng/ml range, however these rates were also very age specific. The age-adjusted incidence rate of organ confined disease increased from 38.5 per 100,000 to 108.8 per 100,000 from 1985 to 2001. Almost 80% of prostate cancers were detected by needle biopsy in 2001 compared to only 34% in 1985, while 20% of cases in 2001 were treated with radical surgery compared to only 2% for 1985. Mortality rates have remained stable up to 2001.
Conclusion: PSA testing is very common in Saskatchewan consistent with extensive screening activity. Conflicting guidelines and the universal availability of the test has resulted in significant inappropriate testing and considerable variation of use. Most prostate cancers are now found by needle biopsy and are organ confined at the time of diagnosis. No benefit in prostate cancer mortality has yet been realized in Saskatchewan from extensive PSA testing.
Advisor: Muhajarine,Nazeem; Reeder, Bruce; Skarsgard, David; Leis, Anne; Hakulinen, Timo; Chibbar, Rajni; McGregor, Liz
School:University of Saskatchewan
School Location:Canada - Saskatchewan
Source Type:Master's Thesis
Keywords:psa testing prostate cancer epidemiology
Date of Publication:04/01/2009