Development of a clinical nomogram for predicting bladder outlet obstruction in male LUTS/BPH / 实用医学杂志

ABSTRACT

Objective To develop a clinical nomogram for predicting the probability of bladder outlet obstruction (BOO) in male LUTS/BPH using the most common and noninvasive parameters in clinical practice , with the hope of detecting BOO individually and precisely. Methods Retrospectively analyze the outpatients and inpatients of male LUTS/BPH from November 2003 to November 2015 in Guangzhou First People′s Hospital. Collect the Pressure-flow study parameters and other clinical parameters including Qmax , PV, TZV, TZI, PSA, and PVR. Find out the best independent predictors on the diagnosis of BOO and develop the nomogram for pre-dicting BOO. Results The data from 1 599 patients were analyzed. The areas under the ROC curve (AUCs)of PV, TZV, TZI, PSA, Qmax, and PVR were 0.803, 0.807, 0.698, 0.775, 0.742, and 0.641, respectively. Qmax, PV, and PSA were selected as the best clinical parameters to predict BOO. The Logistic regression equa-tion is Log(p)=0.332 4 - 0.201 8*Qmax + 0.026 6*PV + 1.135 1*PSA. Finally, a nomogram model was developed by R statistical software. This nomogram showed a concordance index of 0.854 according to the inter-nal validation of the model. Conclusions The clinical nomogram presented a high accuracy (85.4%) in de-tecting BOO, which would help predicting BOO in male LUTS/BPH noninvasively, individually, accurately, and providing valuable reference and guidance in clinical decision.