Performance of the 2021 AAGL endometriosis classification

The 2021 AAGL endometriosis classification shows variability in interpretations among observers

Key Points

Importance:

• Inter-observer agreement for the American Association of Gynecologic Laparoscopists (AAGL) 2021 Endometriosis Classification System has not yet been described.

Highlights:

• In this retrospective scoring study, the 2021 AAGL endometriosis classification appears to have some limitations in its effectiveness.
• Applying a similar methodology in a prospective study could address the challenges of exact staging and scoring with this classification system.

What's done here:

• A group of Australian and New Zealand researchers aimed to determine the performance of the AAGL 2021 Endometriosis Classification system and compare it to the rASRM staging system.
• The study included 379 women with suspected endometriosis who underwent laparoscopy performed by one of seven minimally invasive surgeons in Sydney, Australia.
• This study has a multi-center, retrospective inter-observer approach, and excluded women whose diagnostic laparoscopy results were negative for biopsy-proven endometriosis.
• Each of the three observers applied the AAGL system two times for each patient to their systematically collected surgical findings before and after a consensus meeting.

Key Results:

• Out of 376 patients, 62 were excluded because of negative laparoscopy. The surgical data of 317 patients was retrospectively analyzed after the exclusion of 62 patients because of negative laparoscopy.
• Before the consensus meeting, Observer 1 assigned higher ratings compared to Observer 2 and Observer 3; with Observer 2 rating moderately higher than Observer 3, reflecting a lack of strong inter-observer agreement.
• Following the consensus meeting, Observers 1 and 2 had odds ratios of 1.81 and 1.59, respectively, which were higher than that of Observer 3.
• These findings indicate good inter-observer agreement for the AAGL 2021 endometriosis classification."
• However,as the inter-observer agreement changed under the optimized conditions given during the meeting, this classification may not be superior to the rASRM classification of endometriosis.

Lay Summary

The most common scoring system is the revised American Society for Reproductive Medicine (rASRM) staging system for endometriosis. However, this classification system was widely criticized for failing to correlate surgical complexity and to show patients' pain scales and fertility indexes. It also has a poor inter-observer agreement. 

To replace rASRM with an improved system, the American Association of Gynecologic Laparoscopists (AAGL) announced a new endometriosis classification in 2021. The primary objective was to correlate surgical complexity and usability, and the second objective was related to baseline clinical factors.

A large prospective cohort study showed high concordance between the AAGL stage and surgical complexity level, but inter-observer agreement was not assessed.

Hence, Mak et al., from the Nepean Hospital, New South Wales, Australia, planned a prospective study to determine the inter-observer agreement of the AAGL endometriosis classification system. They utilized 31 cases from previously published endometriosis studies; after excluding 62 because of negative laparoscopy. Three minimally invasive surgery expert observers were asked to allocate the AAGL classification for each case using the point-scoring shema described in the reference paper. This scoring was performed on systematically collected surgical data, was not real-time, and was not aided by intraoperative images. 

The 3 expert observers initially scored the first case based on their interpretations of the AAGL system. Following this, they held a consensus meeting to address any uncertainties and reach a unified decision. They then reassessed and scored the AAGL classification for each case based on this joint discussion. The results showed poor inter-observer agreement for the first round and a better one for the second. The authors indicated that a prospective study in which scoring and staging are performed using the usual rational rules, like those used in older rASRM systems, would help to reach conclusions.

This study was recently published in the Australian and New Zealand Journal of Obstetrics and Gynaecology (ANZJOG).  

Research Source: https://pubmed.ncbi.nlm.nih.gov/38896105/