Kappa statistics in the screening of malignancy of prostate

The optimal upper limit of the normal range for prostate specific antigen (PSA) is suggested, but review of literature reveals that, malignancy of prostate can occur below that range and some benign prostatic diseases are occasionally associated with higher levels. The aim of the study is for early detection of prostatic malignancy. We tried to evaluate digital rectal examination (DRE), estimation of PSA and transrectal ultrasound (TRUS) guided core needle biopsy and histopathological examination in the patients. Seventy-two consecutive patients with lower urinary tract symptoms were taken in this retrospective study from January 2005 to February 2006. PSA level was measured by automated chemilumininescence system. Prostatic biopsies were taken for histopathological examination and stained with haematoxylin and eosin. Gleason grading was applied in case of adenocarcinoma of prostate. For detection of malignancy, sensitivity, specificity, predictive value for positive test and of negative test, percentage of false negatives and false positives, p-values, confidence interval and kappa statistical calculations were done. It was found 19 cases with PSA level > 4 ng/ml had benign diseases of prostate and one person having PSA level < 4 ng/ml had adenocarcinoma of prostate. Seven DRE positive cases had benign disease of the prostate and 5 DRE negative patients had adenocarcinoma of prostate. When compared, serum total PSA value alone and combined PSA and DRE, the later combined approach was found to be more useful. We recommend the study of DRE, PSA and TRUS guided core needle biopsy for detection of prostatic cancer at localised and potentially curable stage.