Reproducibility of the interpretation of pelvic x-ray 3 months after hysteroscopic sterilization with Essure

Objective To estimate the diagnostic accuracy and the interobserver reproducibility of pelvic x-rays in the diagnosis of successful bilateral sterilization with Essure after a 3-month follow-up period. Design Interobserver study. Setting Outpatient department of obstetrics and gynecology in a Dutch teaching hospital. Patient(s) Patients with successful bilateral Essure placement. Intervention(s) Hysteroscopic sterilization with Essure and pelvic x-ray and hysterosalpingography after a 3-month follow-up period. Main Outcome Measure(s) Six observers evaluations of 47 pelvic x-rays from 47 patients 3 months after a technical successful bilateral placement of microinserts to estimate the reliability of the sterilization. Diagnostic accuracy of pelvic x-ray per observer in detecting incorrectly positioned microinserts was expressed in terms of sensitivity and specificity, with hysterosalpingography as the reference strategy. Reproducibility of the interpretation of the pelvic x-ray was expressed as κ-values. Result(s) The sensitivity and specificity for x-rays read by gynecologists was 0.67 (95% confidence interval [CI], 0.29–0.96) and 0.79 (95% CI, 0.58–1.00) and for radiologists 1.0 and 0.5 (95% CI, 0.36–0.64). The interobserver agreement in reliability of pelvic x-ray of hysteroscopic sterilization assessment with Essure ranged from slight (κ-value = 0.09) for gynecologists to moderate (κ-value = 0.52) for radiologists. Conclusion(s) Test characteristics of pelvic x-ray as the imaging technique to assess the position of the Essure microinserts and tubal patency were poor, as was the reproducibility, particularly if gynecologists performed the evaluation. We do not recommend the use of pelvic x-ray for the assessment of the positioning of microinserts after hysteroscopic sterilization.