The optimal cutoff value for levator‐urethra gap measurements using tomographic ultrasound imaging in avulsion diagnosis is population specific

• Published in: Neurourology and urodynamics 2020-06, Vol.39 (5), p.1401-1409
• Main Authors: Greenbaum, Hila, Klein, Linor, Alcalay, Meny, Kassif, Eran, Weisz, Boaz, Eisenberg, Vered H.
• Format: Article
• Language: English
• Subjects: Adult; Anal Canal - diagnostic imaging; avulsion defect; Body mass index; Contraction; Delivery, Obstetric - adverse effects; Female; Humans; levator‐urethra gap; Middle Aged; Pelvic Floor - diagnostic imaging; Pelvic Floor - injuries; pelvic organ prolapse; Population studies; Pregnancy; Retrospective Studies; Risk Factors; Sphincter; tomographic ultrasound imaging; transperineal ultrasound; Ultrasonic imaging; Ultrasonography - methods; Ultrasound; Urethra; Urethra - diagnostic imaging; Youden's index analysis
• ISSN: 0733-2467; 1520-6777
• DOI: 10.1002/nau.24353

Aims The levator‐urethra gap (LUG), the distance between the urethral lumen center and levator insertion on the inferior pubic rami, can be used for diagnosing levator avulsion, with a previously suggested cutoff of LUG ≥2.5 cm. The aim of this study is to determine an optimal cutoff value for LUG measurements in a high‐risk patient population. Methods Women followed prospectively after sustaining obstetric anal sphincter injury underwent an interview pelvic examination questionnaires and four‐dimensional‐transperineal ultrasound examination. Levator avulsion was diagnosed on contraction using tomographic ultrasound imaging. Ultrasound datasets were analyzed offline at a later time blinded to previous data. LUG was measured on each side of the three central slices, yielding six measurements and the highest available value was obtained on each side. Different cutoffs were evaluated using receiver‐operating characteristics (ROC) curve analysis and Youden's test. The cutoff was validated against symptoms and signs, and sonographic findings using logistic regression analysis. Results A total of 618 complete datasets were available for analysis, median age 29 years, median body mass index of 23.4 kg/m2, parity 1, and 26.4% instrumental deliveries. Youden's test and ROC curve analysis gave the best area under the curve of 0.869 for a cutoff of 2.305 (95% confidence interval, 0.839‐0.9). Women diagnosed with avulsion based on this cutoff were more symptomatic, whereas using larger cutoffs missed more avulsion defects. Conclusion LUG measurement is useful but should be individualized to the population studied, in our case, in a high‐risk population, 2.305 cm was the optimal cutoff. Using larger cutoffs may be more specific but is likely to miss more cases.