RESUMO
BACKGROUND: Using a single cutpoint for the free-to-total (F/T) prostate specific antigen (PSA) ratio loses important diagnostic information. We evaluated the performance of multiple F/T PSA cutpoints in detecting prostate cancer in men with nonspecific PSA values. METHODS: We extracted sensitivity and specificity data from 12 studies reporting on >or=30 cancer patients with PSA values between 2.0 and 10.0 ng/mL. We calculated stratum-specific likelihood ratios (LR) and areas under the receiver operating characteristic (ROC) curves. RESULTS: Multiple cutpoints for the F/T PSA ratio significantly increased the area under the ROC (0.70) compared with the single investigator-selected cutpoint (0.62), P < 0.004. The LR for the most positive cutpoint stratum (2.62) was significantly higher than the LR for a positive test from the single cutpoint (1.36), P < 0.004. CONCLUSIONS: Using multiple cutpoints increased the discriminating power of the F/T PSA ratio and led to greater probability revisions in the most positive test-result strata.
