Estimate criteria for diagnosis and severity in benign prostatic hyperplasia

Abstract

Background: Standard severity criteria for benign prostatic hyperplasia (BPH) are needed to facilitate clinical and epidemiologic investigations in the common terminology. Patients and Methods: A total of 692 BPH patients from 8 institutions were judged by specialized physicians for their overall disease severity and severity in 4 domains: symptom, function, anatomy, and quality of life (QOL). The items measuring severity in these 4 domains were the international Prostate Symptom Score (I-PSS) questionnaire, uroflowmetry, transrectal ultrasonography (TRUS), and questionnaires on bother and I-PSS QOL index, respectively. By Spearman's rank correlation and multiple regression analysis using physicians' judgement as the criterion variable (gold standard) the items and their cut-off levels were selected so that high predictivity, practicality and clinical significance were maintained, while the number of items in use was minimized. Results: The item(s) and their ranges of values for each severity grade were summarized as follows; the evaluation domain (the selected item(s): range of mild, moderate, severe): symptom (I-PSS: 0-7, 8-19, 20-35), function (maximum flow rate and residual urine volume: ≤ 15 mL/s and < 50 mL, ≤ 5 mL/s and < 100 mL, < 5 mL/s or ≤ 100 mL), anatomy (prostate volume on TRUS: < 20 mL, < 50 mL, ≤ 50 mL), QOL (QOL index: 0 and 1, 2 to 4, 5 and 6). The overall severity was efficiently predicted when defined according to the combinations of severity grades of the 4 domains. These criteria predicted physicians' judgement with accuracy of about 80% for overall severity and severity of individual domains. Conclusion: The proposed criteria are valid, practical, and would be useful for the common scale of the BPH severity in clinical and epidemiologic studies.