How reliable is a single measurement of urinary flow in the diagnosis of obstruction in benign prostatic hyperplasia?

Twenty-six elderly patients with obstructive symptomatology and an initial low peak urinary flow rate (< 15 ml/s) were observed for 6 to 12 months. Repeated assessments were made of flow rates and residual volumes. Within patients variation of the maximal urinary flow was significant; the standard deviation (SD) varied from 0.8 to 5.5 ml/s. The SD varied with the mean peak flow rates and this suggests that the precision of a single determination of a patient's peak flow is inversely related to the peak flow itself. In all, 333 determinations of peak flow were obtained. Only 80% were below 2 SD of the mean Siroky nomogram. Multiple determinations were used to estimate the sensitivity rate for each patient, i.e. the proportion of peak flows that were less than 2 SD below the mean of Siroky's nomogram. The average sensitivity for all of the flow values was 0.813 when applied to voided volumes and increased only slightly to 0.838 when applied to total bladder volume. Given a prevalence of 0.70 of obstruction ascribed to benign prostatic hyperplasia (BPH) among elderly men, the positive predictive value of an abnormally low peak flow was 0.97. A single low peak flow based on assessment of voided volume may serve as an almost sure indication of obstruction. If a patient has a single normal peak flow rate, he still has a 0.31 probability of having evidence of obstruction in his following uroflow determinations. Using total bladder volume does not make a difference in most patients and is not worth the effort and discomfort. The only patients who require accurate residual measurements are those with a normal uroflow but clear-cut obstructive symptomatology or with borderline peak flow. The results of this study are relevant to everyday clinical practice and to the evaluation of studies on alternatives to surgery in BPH.