Aims: The clinical validity of uroflowmetry in women is attenuated by lack of absolute normal values. A peak flow <15 mL/sec and/or residual urine >50 mL with a minimum total bladder volume of 150 mL before voiding (volume voided+residual) (method A) and the 10th centile curve of the Liverpool Nomogram (method B) for the maximum urine flow rate have been identified as useful discriminants when diagnosing voiding difficulties in women. This study compares the two methods and analyses the validity of uroflowmetry in female voiding disturbances. Methods: A total of 348 women underwent a full urogynaecologic work-up. Evaluable results of uroflowmetry (229 with method A and 224 with method B) were analysed and compared in terms of the following clinical variables: age, parity, previous urogynaecologic surgery, prolapse grade, symptoms, postvoid residue, and incontinence. Uroflowmetry results were compared with pressure/flow study results as indicated by four different cut-offs. Results: The odds ratio that a subject with voiding difficulty has abnormal flow is 3.7 (95% CI, 1.9-7) in the patients analysed with method A and 2.8 (95% Cl, 1.6-5.2) with method B. A good accordance emerged between the two methods in flowmetry results. Uroflowmetry has a specificity of >70% and a sensitivity of 50 to 100% depending on the cut-offs. Uroflowmetry results in women can be analysed by using either of the methods. Conclusions: Uroflowmetry has a good specificity, a high negative predictive value, and a good diagnostic capacity such as to make it useful as the first diagnostic approach in urogynaecologic patients. (C) 2003 Wiley-Liss, Inc.
