Diagnostic value of combined pelvic floor ultrasound parameters for pelvic floor dysfunction and risk factors.

OBJECTIVE: To investigate the diagnostic value of pelvic floor ultrasound parameters in combination for pelvic floor dysfunction (PFD), and to explore the risk factors. METHODS: Forty PFD patients treated from April2019to December 2020(case group) and another 40 healthy women (control group) were enrolled. Their clinical data were collected, and both groups received three-dimensional (3D) ultrasound of the pelvic floor. The diagnostic value of pelvic floor ultrasound parameters for PFD was assessed by receiver operating characteristic (ROC) curves. The risk factors of PFD were evaluated by multivariate logistic regression analysis. RESULTS: The area under the ROC curve (AUC), sensitivity, and specificity of the parameters in combination for predicting PFD were 0.851 [95% confidence interval (CI): 0.743-0.959], 0.901, and 0.812, respectively, indicating acceptable accuracy. Results of logistic regression analysis revealed that spontaneous delivery, lateral episiotomy/laceration, and large bladder neck rotation angle, posterior urethrovesical angle (PUA), bladder neck tilt angle, bladder neck distance (BND), levator hiatus area (LHA) (at anal contraction), R-LHA and V-LHA were risk factors for PFD (p < 0.05), while physical exercise was a protective factor (p < 0.05). ROC curve analysis revealed that the AUC, sensitivity, and specificity of the forest map model were 0.822 (95% CI: 0.759-0.885), 0.942, and 0.601, respectively, indicating acceptable accuracy of the model. Internal data validation of the model demonstrated consistence of the predicted occurrence of PFD with the actual one. CONCLUSIONS: Spontaneous delivery, lateral episiotomy/laceration, and large bladder neck rotation angle, PUA, bladder neck tilt angle, BND, LHA (at anal contraction), R-LHA and V-LHA were risk factors for PFD.