Construction and evaluation of a model based on clinical factors and multimodal ultrasound parameters for diagnosing postpartum pelvic floor myofascial pain.

Background: Pelvic floor myofascial pain (PFMP) is an important yet often overlooked cause of chronic pelvic pain (CPP). The currently available diagnostic methods are limited due to issues such as insufficient efficacy and difficulties in practical implementation, while alternatives have yet to be established. The aims of this study were to evaluate the utility of multimodal ultrasound consisting of shear wave elastography (SWE) with two-dimensional and four-dimensional imaging in diagnosing postpartum PFMP and to develop a diagnostic model. Methods: A total of 346 postpartum women were prospectively enrolled and randomly allocated to training and validation sets at a 7:3 ratio, with 242 in the training set and 104 in the validation set. Nineteen clinical and ultrasound parameters were analyzed. Significant variables (P<0.05) on univariate analysis were entered into backward multivariate regression with multicollinearity assessment. Independent predictors were incorporated into a nomogram. The nomogram was validated via 1,000 bootstrap resamples. Performance was assessed by discrimination with the area under the receiver operating characteristic (ROC) curve (AUC), calibration with the Hosmer-Lemeshow (HL) test and calibration plots, and clinical utility with decision curve analysis (DCA). Results: Four diagnostic indicators were independently associated with postpartum PFMP and were retained in the final model: mean shear wave velocity (SWV), anorectal angle (ARA) excursion during maximum Valsalva maneuver (VM), percentage change in levator hiatal area, and weight gain during pregnancy. The nomogram demonstrated excellent discrimination after 1,000 bootstrap resamples, with AUC of 0.880 [95% confidence interval (CI): 0.836-0.924] and 0.861 (95% CI: 0.790-0.931) in the training and validation sets, respectively. The calibration plots confirmed high consistency between predicted probabilities and actual observations. DCA demonstrated net benefit across a wide threshold probability range in the training set (11-100%) and validation set (3-87% and 92-100%). Conclusions: The developed nomogram model demonstrated robust performance in diagnosing postpartum PFMP and thus may have value as an adjunct diagnostic tool in clinical practice. It may serve as a noninvasive approach to mitigating the diagnostic constraints in clinical practice.