Interobserver reproducibility of two endometriosis scoring systems: a multicenter observational prospective study.

Abstract

Objective: To assess the interobserver reproducibility of the Revised American Society for Reproductive Medicine score (r-ASRM) and the Endometriosis Fertility Index (EFI) in women undergoing a conservative laparoscopy to treat endometriosis.

Methods: The r-ASRM stage and score and EFI were independently determined by two assessors participating in the surgery. Assessors were either a minimally invasive (MIS) specialist or fellow, a fertility specialist, or an obstetrics and gynecology resident. They both completed the score sheets separately, blindly to the other assessor. A consensus was then obtained, after discussion between the two assessors. Interobserver reproducibility was evaluated using Cohen's Kappa and intraclass correlation coefficient (ICC) for categories and continuous scores, respectively.

Results: In this multicenter cross-sectional cohort study, 100 women undergoing a laparoscopy for endometriosis were recruited between April 2020 and May 2023. Most participants had stage 3 (22%) or 4 (43%) endometriosis. The interobserver agreement was strong for the r-ASRM stage, and almost perfect for EFI categories (0-3, 4, 5, 6, 7-8, 9-10) and Least function score categories (1-3, 4-6, and 7-8). The agreement for the r-ASRM score and EFI score is excellent. Interobserver agreement remained strong regardless of the level of expertise, the use of preoperative suppression or history of a prior pelvic surgery.

Conclusion: r-ASRM and EFI classifications are highly reproductible between assessors, making them excellent tools for communication between health professionals. However both are limited by their capacity to explain variations in pain symptoms, which remains a challenge to be addressed in future work.