借由回溯性研究，本文共包括121位前列腺癌患者和127位良性前列腺肥大症病人為對照組。在我們的前例腺癌病人中，幾乎大部份(46.3%)為癌症第IV期，局部性腫瘤(第I，II期)則只有46個案例(佔38%)。為了提高精確度(81.9%)以及較高的精確率(77.4%)，所以我們將前列腺特異抗原的正常值訂在10 ng/ml。肛門指診的敏感度、精確度及精確率分別為66.1%，92.9%以及79.8%。而經直腸超音波檢查則分別為59.9%，89.8%以及75%。雖然應用此三種檢查在偵測較晚期前列腺癌比較早期者有顯著的統計學上差異(肛門指診：p<0.001；經直腸超音波：p<0.001；前列腺特異抗原：p<0.05)，但對於所有前列腺癌或早期前列腺癌患者，此三種檢查在偵測能力上並無統計學上明顯的差異。同時在肛門指診，前列腺特異抗原及經直腸超音波都有不正常的檢查結果時，則最能精確地預測癌症的存在(勝算比=124.3)。所以在有更新的診斷方式取代之前應用此三種方法來診斷前列腺癌仍是無法避免的。

Retrospectively records of 121 patients of prostate cancer and 127 patients of benign prostate hyperplasia, as a control group, were reviewed. Almost half (46.3%) of the prostate cancer were Stage D, and 46 cases (38%) were clinically localized cancer (Stages A, B) in this study. The cutoff point of prostate specific antigen (PSA) was set at 10 ng/ml to have higher specificity (81.9%) and accuracy rate (80.7%). The sensitivity, specificity and accuracy rate of digital rectal examinations (DRE) were 66.1%, 92.9% and 79.8%; the statistical results of transrectal sonography (TRUS) were 59.5%, 89.8% and 75%, respectively. Although there were significant differences in applying each of these three methods in detecting prostate cancer in its advanced stages compared with localized stages (DRE:P<0.001; TRS:P<0.001; PSA:P<0.05), there was no significant difference among these three methods in detecting prostate cancer in either the whole cancer group (X2(2)=3.2391, P>0.1) or the localized cancer subgroup (X2(2)=4.9785, P>0.05). Combining findings from DRE, PSA and TRUS, it is possible to predict cancer with the highest degree of accuracy (Odds ratio=124.3). These results have allowed development of clinical guidelines for accurate diagnosis of prostate cancer.