Developing and validating a new nomogram for diagnosing bladder outlet obstruction in women.

To develop and validate a nomogram for assessing bladder outlet obstruction (BOO) in women derived from concurrent Pdet.Qmax and Qmax based on radiographic evidence of increased urethral resistance. Retrospective analysis of prospectively acquired video-urodynamics and clinical data of 185 women (development cohort) was performed. The Pdet.Qmax were plotted against Qmax and cluster analysis was performed to determine an axis that best divided the definitively obstructed and unobstructed. Using data from a further 350 women (validation cohort), the sensitivity and specificity of the derived criterion was calculated. Finally, the data from both groups was pooled together and using binary logistic regression analysis, a nomogram was produced. Of the 535 patients in the two cohorts, (122 [22.8%]) demonstrated radiographic evidence of BOO. Cluster analysis identified the axis that best separates the radiographically obstructed and unobstructed as Pdet.Qmax  = 2*Qmax . Using the data from the validation cohort, the sensitivity and specificity for this was calculated as 0.94 and 0.93, respectively. A nomogram, representing the probability of BOO for concurrent Pdet.Qmax and Qmax measurements was derived by pooling data from both cohorts. Alternatively, a female BOO index (BOOIf) may be calculated mathematically using the formula BOOIf = Pdet.Qmax  - 2.2*Qmax, that is, BOOIf < 0, <10% probability of obstruction, BOOIf > 5 likely obstructed (50%) and If BOOIf > 18, obstruction almost certain (>90%). A female BOO nomogram (the SG nomogram) with high sensitivity and specificity is proposed. The nomogram can be used to stratify the degree of BOO or assess response to treatment. © 2017 Wiley Periodicals, Inc.

Citation

Eskinder Solomon, Habiba Yasmin, Megan Duffy, Tanwir Rashid, Emmanuel Akinluyi, Tamsin J Greenwell. Developing and validating a new nomogram for diagnosing bladder outlet obstruction in women. Neurourology and urodynamics. 2017 Jun 30

PMID: 28666055

search → result