Objective: To assess the criterion validity of the American Urological Association (AUA) symptom index for benign prostatic hyperplasia (BPH) in urological practice.
Patients and methods: The study comprised 1414 consecutive men > or = 50 years of age, newly referred to a urologist in a stratified sample of 12 hospitals throughout the Netherlands, who completed the AUA symptom questionnaire at their first visit. The 39 urologists completed a questionnaire on the diagnosis. The discriminative power of the index for BPH versus other urological diagnoses was calculated by receiver operating characteristic (ROC) analysis. Subsequently, criterion sensibility was studied by assessing the effect of particular variables on the probability of the diagnosis of BPH.
Results: ROC areas were low for BPH versus prostate cancer (0.57, SE 0.04), chronic prostatitis/prostatodynia (0.65, SE 0.03), and other diseases of the lower urinary tract (0.57, SE 0.04). Satisfactory to good values were found for BPH versus diseases of the upper urinary tract (0.79, SE 0.03), impotence/infertility (0.79, SE 0.04), penile-scrotal diseases (0.85, SE 0.02), and no abnormalities (0.84, SE 0.03). The diagnosis of BPH, as made by urologists in daily practice, appeared to be a weak criterion. Irrespective of other characteristics, the chance of being diagnosed with BPH was affected by differences in definition and the type of hospital.
Conclusion: Although the AUA index discriminated fairly well between BPH and diseases of the upper urinary tract, as well as genital diseases, it has no potential for distinguishing BPH from other diseases of the lower urinary tract. In addition, the lack of consensus in defining clinical BPH underscores the weakness of the index as a tool in the diagnostic process of this disease.