Hypothesis / aims of study
Micturition physiology differs in men and women. However, the results in urodynamic studies (US) in women with lower urinary tract symptoms (LUTS) were extrapolated from studies in men. Nowadays, the only validated nomogram for females is Solomon-Greenwell’s (1). However, it only evaluated bladder outlet obstruction (BOO) without considering underactive bladder (UAB). This study aims to create a nomogram that includes an evaluation of UAB and BOO in non-neurogenic women and validate it against videourodynamic studies (VUS) along with other nomograms.
Study design, materials and methods
For the creation cohort, a total of 183 female patients over 18 years old who underwent VUS between 2021 and 2022 were included under the tenets of the Declaration of Helsinki after the approval of the institutional ethics committee. The VUS was performed under the International Continence Society (ICS) standardization. Excluded patients were those with neurological pathology and renal transplantation, and trouble performing the flow-pressure (QP) phase.
The baseline characteristics of the patients and the VUS values and diagnosis between two reviewers were evaluated. The QP Qmax and PdetQmax were significant predictors for BOO and UAB in a logistic regression. The results were plotted in a nomogram creating four groups: 1) No BOO – No UAB, 2) BOO, 3) UAB and 4) UAB-BOO.
To test the nomograms Blaivas-Groutz (BG), Solomon-Greenwell (SG), LinPURR and the our new one, and the BOO index (BOOI) and bladder contractility index (BCI), we included 142 patients from 2023 to 2024 under the same criteria as the creation cohort.
For validation with ROC curves, the qualifications in nomograms and indexes were recoded being non-BOO or non-UAB all with negative or equivocal values, BOO those with any kind of obstruction, and UAB the ones with hypocontractility in any degree. Diagnostic tests were performed for our new nomogram. All analyses were performed with SPSS 29 (IBM, Chicago, 2023) and RStudio (RStudio Team, Boston, 2020). A p<0.05 was considered significant.
Results
The median age of the creation cohort is 50 years old [IQR 39-63] and all patients were women with LUTS without a clear diagnosis in the first US. We found two predictors for BOO: QP Qmax (OR = 0.72, CI 95% [0.628-0.826], p<0.001) and QP PdetQmax (OR = 1.146, CI 95% [1.078-1.218], p<0.001). Also, two predictors for UAB: QP Qmax (OR = 0.846, CI 95% [0.748-0.958], p=0.008) and QP PdetQmax (OR = 0.82, CI 95% [0.736-0.914], p<0.001).
After the creation of the model, we clustered the predicted data in two groups (yes/no) for each model with a p<0.001 and graphed our new nomogram (figure 1).
The median age of the test cohort is 44 years old [IQR 33.75-59] being younger than the creation cohort (p=0.036). However, none other parameter of the VUS was different, so the age was not considered clinically significant. All data in table 1.
The ROC curve for BOO (figure 2.A) showed that the most accurate diagnostic nomogram or index was our new nomogram (85.4%, p=0.000), followed by Blaivas-Groutz (68.5%, p=0.000), Solomon-Greenwell (58.1%, p=0.089), BOOI (57.1%, p=0.135) and LinPURR (50%, p=1.000). For diagnostic tests, sensibility is 83.1%, specificity 87.7%, positive predictive value 88.8% and negative predictive value 96.6%.
The ROC curve for UAB (figure 2.B) showed that our new nomogram was also the most accurate diagnostic tool (80.2%, p=0.003), followed by BCI (76.6%, p=0.001) and LinPURR (70.1%, p=0.078). For diagnostic tests, sensibility is 71.4%, specificity 88.8%, positive predictive value 25% and negative predictive value 98.3%.
Interpretation of results
Free uroflowmetry and standard urodynamics are diagnostic tools that allow us to estimate the function and dysfunction of the LUT. However, the standardization of these paraclinical tests has been carried out on men despite the different micturition physiology in women.
To extrapolate the LinPURR nomogram (2) to the female population, we need to understand that women would need a lower Pdet to achieve a higher Qmax than a man, since relaxation of the pelvic floor would allow a lower opening pressure of the urethra. We found a diagnostic accuracy of 50% for BOO and 70.2% for UAB. This means we should not use it in women.
The BOOI, was also created for men. In our study, when compared to the VUS, it has an accuracy of 57.1% which means it can let us suspect BOO in women, but it cannot be used as a diagnostic tool. Solomon et al., recalculated the BOOI for the female population (BOOIf) creating the Solomon-Greenwell nomogram plotting the QP Qmax vs. PdetQmax, and establishing different probabilities for BOO (1). We validated it finding an accuracy of 58.1%. This means, that grouping BOOIf by probabilities still lead to suspicious but does not allow to make a definitive diagnosis for the patient.
One more nomogram created for the female population is the Blaivas-Groutz. It compares the relationship between the Qmax of the uroflowmetry and the maximum Pdet of the QP, considering the decrease in Qmax due to using a urethral catheter in the QF curve (3). The Blaivas-Groutz nomogram had not been validated until now, where we found an accuracy of 68.5%.
On the other hand, we only have one tool to predict UAB in women, which is de BCI, also created and used in men. We validated it finding an accuracy of 76.6%. It means we can use BCI to guide our suspicious but should not be used as definitive diagnosis.
Since we want to simplify the diagnosis of BOO and UAB in women and reduce the radiation and the time to diagnosis, we created this new nomogram named after our first author. We validated it against VUS, finding an accuracy of 80.2% for UAB and 85.4% for BOO.