Dynamic Anal Endosonography and MRI Defecography in Diagnosis of Pelvic Floor Disorders: Comparison With Conventional Defecography

BACKGROUND:Pelvic floor disorders are frequent, especially in women. Surgeons need more information on the accuracy of available diagnostic techniques to make therapeutic decisions. OBJECTIVE:This study aimed to compare the accuracy of dynamic anorectal endosonography and dynamic MRI defecography with conventional defecography as the criterion standard in the diagnosis of pelvic floor disorders. DESIGN:We used a prospective crossover design in which patients underwent each procedure in random order within the same month. SETTING:Investigations were conducted at a regional referral center in Marseille, France. PATIENTS:Women with dyschezia who were undergoing diagnostic evaluation were eligible. INTERVENTION:Dynamic anorectal endosonography, dynamic MRI, and conventional defecography were performed in all patients by 3 blinded operators. MAIN OUTCOME MEASURE:The accuracy of dynamic anorectal endosonography and dynamic MRI in the diagnosis of pelvic floor disorders was assessed by calculating sensitivity, specificity, positive and negative predictive values, correlation coefficients, concordance rates, and the Cohen κ statistic, with conventional defecography used as the criterion standard. RESULTS:The study comprised 56 women with a mean age of 50.7 (SD, 12.5) years. No significant differences were observed between dynamic anorectal endosonography and dynamic MRI in the number of patients with rectocele (P = .49), perineal descent (P = .11 when dynamic anorectal endosonography measured descent of the puborectalis muscle; P = .27 for bladder descent), or enterocele (P = .78); no differences were found between these techniques in sensitivity, specificity, or positive and negative predictive values. Diagnostic concordance with conventional defecography as the standard did not differ significantly between dynamic MRI and dynamic anorectal endosonographyConcordance rates for dynamic anorectal endosonography were 75% for rectocele, 64% for perineal descent, and 91% for enterocele (no rectal intussusception was found with dynamic anorectal endosonography); concordance rates for dynamic MRI were 82% for rectocele, 57% for perineal descent, 93% for enterocele, and 55% for rectal intussusception. Significantly more internal anal sphincter defects were found with dynamic anorectal endosonography than with dynamic MRI defecography21 patients (37.5%) vs 12 patients (21.4%); P = .02. Patient tolerance was significantly better for dynamic anorectal endosonography than for