Strengths and Limitations
This retrospective study used descriptive data rather than laparoscopic
images to simulate endometriosis staging, and therefore might not be as
representative as real time intraoperative staging. A prospective study
where scoring and staging is performed in the usual contemporaneous
fashion would overcome this. The three assessors in this study were
either MIGS or fellows in their final year of MIGS training, and data
were obtained from tertiary pelvic pain referral centres. Selection bias
has been identified as a challenge in endometriosis research (22),
whereby sample populations recruited to studies are not always
representative of the 11% (23) of the female population that suffer
from endometriosis. These findings therefore might not be generalizable
to general gynaecology clinics and generalist obstetricians and
gynaecologists who might use this tool, for example.
A strength of our study was the use of a consensus process to reduce
heterogeneity between assessors interpreting the AAGL system. The
rationale for this was that the authors identified several potential
areas of ambiguity within the tool. While this improved consistency
between the three assessors in this experiment, it does not necessarily
represent the real-world application of endometriosis staging tools. In
addition, the fact that three observers were used and the best of three
results was considered for discussion demonstrates that all due effort
was made assess the AAGL system fairly. Bias was mitigated by blinding
between observers, and also between the staging and skill level
allocations.