[Relationship between screening by stratifying cases into groups on prostate specific antigen level and the positive...

To evaluate the detection of prostate cancer in different prostate specific antigen (PSA) level and the predict value of PSA, digital rectal examination (DRE), transrectal ultrasound scan (TRUS) and PSA density (PSAD).

The clinical data of 634 cases who had underwent transrectal ultrasound guided systematic sextant prostate biopsies between April 1996 to December 2002 due to being suspicious of prostate cancer were retrospectively analyzed. The detection of prostate cancer in different PSA groups, namely PSA < or = 4.0, 4.1-, 10.1-, > 20.0 microg/L, and the predict values of PSA, DRE, TRUS and PSAD were statistically analyzed using t test, chi2 test and logistic regression analysis.

The rates of prostate cancer detection in different PSA groups were 11.6%, 26.8%, 39.8% and 68.6%, respectively. The higher the PSA, the higher the rate of prostate cancer detection, the same was the positive predictive value of DRE and TRUS. The sensitivity and specificity of PSA > 4.0 microg/L were 93.0% and 33.0%, and the efficiency of DRE and TRUS were very low. Logistic regression analysis indicated that PSAD was the most risk factor of prostate cancer in the group of PSA 4.1-20.0 microg/L (OR = 687.09 +/- 646.96, P = 0.000).

The rates of prostate cancer detection in different PSA groups are different compared with other countries. The screening roles of DRE and TRUS are dependent on PSA level. Utilization of the screening protocol which to stratify cases into three PSA groups, namely PSA < or = 4.0, 4.1 - 20.0, > 20.0 microg/L, can elevate the positive rate of prostate biopsies without sacrificing cancers detected.