Development and internal validation of a screening tool for chronic prostatitis (S-CP)

Abstract

Abstract Purpose We developed a simple self-checkable screening tool for chronic prostatitis (S-CP) and internally validated it to encourage men (in the general population) with possible chronic prostatitis to consult urologists. Methods The expert panel proposed the S-CP, which comprises three domains: Area of pain or discomfort (6 components), accompanying Symptom (6 components), and Trigger for symptom flares (4 components). We employed logistic regression to predict chronic prostatitis prevalence with the S-CP. We evaluated the predictive performance using data from a representative national survey of Japanese men aged 20 to 84. We calculated the optimism-adjusted area under the curve using bootstrapping. We assessed sensitivity/specificity, likelihood ratio, and predictive value for each cutoff of the S-CP. Results Data were collected for 5,010 men—71 (1.4%) had a chronic prostatitis diagnosis. The apparent and adjusted area under the curve for the S-CP was 0.765 [95% confidence interval (CI) 0.702, 0.829] and 0.761 (0.696, 0.819), respectively. When the cutoff was two of the three domains being positive, sensitivity and specificity were 62.0% (95% CI 49.7, 73.2) and 85.4% (95% CI 84.4, 86.4), respectively. The positive/negative likelihood ratios were 4.2 (95% CI 3.5, 5.2) and 0.45 (95% CI 0.33, 0.60), respectively. The positive/negative predictive values were 5.7 (95% CI 4.2, 7.6) and 99.4 (95% CI 99.1, 99.6), respectively. Conclusion The reasonable predictive performance of the S-CP indicated that patients (in the general population) with chronic prostatitis were screened as a first step. Further research would develop another tool for diagnostic support in actual clinical settings.