What is the accuracy of transvaginal ultrasound for endometriosis mapping prior to surgery when performed by a sonographer within an outpatient women's imaging centre?

INTRODUCTION

This study aimed to assess the accuracy of transvaginal ultrasound (TVUS) for the mapping of endometriosis before surgery when performed by sonographers in an outpatient women's imaging centre.

METHODS

A prospective longitudinal cohort study was performed. The study group comprised of 201 women who underwent a comprehensive TVUS assessment, performed by a sonographer. Laparoscopy was performed as the reference standard. Complete TVUS and surgical data were available for 53 women who were included in the final analysis.

RESULTS

Endometriosis was confirmed at a surgery in 50/53 (94.3%) participants, with 25/53 (47.2%) having deep endometriosis (DE) nodules and/or endometriomas present. TVUS for mapping of DE had an overall sensitivity of 84.0%, specificity of 89.3%, PPV of 87.5%, NPV of 86.2%, LR+ of 7.85, LR- of 0.18, and accuracy of 86.8% (P < 0.001). Ovarian immobility had poor sensitivity for detecting localised superficial endometriosis, DE, adhesions, and/or endometriomas (Left = 61.9% and right = 13.3%) but high specificities (left = 87.5% and right = 94.7%). Site-specific tenderness had low sensitivities and moderate specificities for the same. All soft markers of endometriosis failed to reach statistical significance except for left ovarian immobility (P = <0.001).

CONCLUSION

Sonographers well experienced in obstetric and gynaecological imaging, working in an outpatient women's imaging setting can accurately map DE; however, the performance of soft markers for detection of SE was poor.