In March 2022 Australian Medicare Benefits Schedule updated gynaecological item numbers relating to endometriosis surgery. Specific item numbers include 35637, 35631 & 35632 (a division of 35638) and 35641. Each item number correlates to staging as per rAFS(rASRM) stage 1-4, respectively. We aimed to predict to predict four MBS item numbers using ultrasound.

Design:

Retrospective observational study

Setting:
Conducted across five countries and six centres. August 2018-November 2019

Patients or Participants:

Patients included when attending gynaecology outpatient ultrasound clinic with symptoms of endometriosis and progressed to laparoscopic surgery for management.

Interventions:

Endometriosis ultrasound was performed as part of their assessment and recorded in line with the IDEA consensus statement. This data has previously been published in evaluating the IDEA consensus endometriosis prediction.

Here, data is used to assess an ultrasound based rASRM endometriosis staging and by-proxy the billing codes for endometriosis surgery. Subgroup analysis was performed by dichotomizing ultrasound and surgical stage/MBS item into a low and high-stage system.

Measurements and Main Results:

273 patients included, 54 excluded due to incomplete data. Prediction of MBS item by rASRM-U stage showed weak agreement, weighted Kappa value 0.31 (0.24 to 0.38, 95% CI). rASRM-U predicted a 0.68 higher surgical-stage, alongside large variability with limits of agreement +/- 1.62 supporting 'over-staging'. Although poor for low stage item numbers, sensitivity of 0.96 and specificity 0.44 related to item 35641. Dichotomised staging shows sensitivity in rASRM-U predicting high-stage of 0.99, specificity 0.34, NPV 0.95. Prediction of low-stage: sensitivity 0.34, specificity 0.99 and NPV 0.79.

Conclusion:

The IDEA based rASRM-U has relatively poor agreement with corresponding MBS item numbers, tending to predict a higher-stage item. This model is unlikely to predict low stages incorrectly. Although unable to accurately predict individual item numbers, this system predicts high vs low surgical complexity when ultrasound and surgical groups are dichotomized, thus could allow for improved surgical planning and financial consent.