OP31.10: Performance of the transvaginal sonographic ‘sliding sign’ in prediction of pouch of Douglas obliteration: does this improve over time?

Pouch of Douglas (POD) obliteration can be predicted with a high degree of certainty in women with symptoms suggestive of underlying endometriosis using transvaginal ultrasound (TVS). To evaluate preoperative real-time dynamic TVS in the prediction of pouch of Douglas (POD) obliteration in women undergoing laparoscopy for suspected endometriosis over an eight-year period. Prospective study between March 2009 and March 2017. Women who attended the Endogynecology clinic at Nepean Hospital underwent detailed ‘deep endometriosis scan’ and follow up laparoscopy. Each woman was assessed for the presence of ‘sliding sign’ on TVS at the both the retro-cervix(RC) and posterior uterine fundus(PUF). As long as the ‘sliding sign’ is considered positive in both of these anatomical regions the POD is recorded as ‘not obliterated’. If either of these anatomical regions demonstrates that the anterior rectal wall or rectosigmoid does not glide smoothly over the RC or PUF, respectively, then the ‘sliding sign’ is considered negative, and the POD is recorded as ‘obliterated’. We evaluated the performance of the ‘sliding sign’ to predict POD obliteration in terms of sensitivity, specificity, PPV & NPV for first 200 cases (March 2009 to January 2013) the second 200 cases (February 2013 to March 2017) as well as all 400 cases. 400 consecutive women were included in the study. Sensitivity, specificity, PPV & NPV for first 200 were 97.32%, 88.46 %, 96.03% & 92.00 % respectively, for second 200 were 97.35%, 66.67%, 90.18% & 88.89 % respectively and all the cases were 97.33%, 78.00 %, 92.99% & 90.70 % respectively. Although the detection rate for POD obliteration was stable between the two cohorts, the false positive rate increased over time. Eliciting the ‘sliding sign’ is an accurate way to predict POD obliteration in women with suspected endometriosis who are planning laparoscopy.