Ultrasound techniques in the diagnosis of deep pelvic endometriosis: algorithm based on a systematic review and meta-analysis

Objective: To collate all available evidence with respect to ultrasound techniques in the management of deep pelvic endometriosis (DPE) and compare the sensitivity and specificity of each to determine the most suitable site-specific method. We aim to provide clinicians with information to improve the diagnosis and management of patients with DPE.

Design: Systematic review of the literature and meta-analysis.

Setting: Not applicable.

Patient(s): None.

Interventions(s): Review of MEDLINE, EMBASE, ScienceDirect, Cochrane Library.

Main outcome measure(s): For each study we calculated the sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and positive/negative likelihood ratio regarding DPE sites. We then compared the specificity and sensitivity of each technique. Forest plots with the corresponding 95% confidence interval using fixed/random effects for each approach (both separately and summarized according to the weight of any single study) were used.

Result(s): A key word search strategy identified 441 manuscripts, 35 of which were eligible for the review (32 for meta-analysis). Standard transvaginal sonography (TVS) showed specificity greater than 85% for all DPE sites, despite sensitivity ranging between 50% (bladder, vaginal wall, and rectovaginal septum) and 84% (rectosigmoid). Modified techniques such as bladder site tenderness-guided TVS showed a value of 97.4% for both sensitivity and specificity. Rectal endoscopy-sonography and rectal water contrast TVS were both superior to TVS in detecting rectosigmoid endometriosis with sensitivities and specificities over 92%. Promising data were reported by using rectal water contrast TVS for rectovaginal septum disease (sensitivity, 97.1%; specificity, 99.3%).

Conclusion(s): The summary of data regarding diagnostic specificity and sensitivity of TVS in women undergoing surgery for deep endometriosis may allow us to conclude that TVS should remain the first-line method in the evaluation of patients with suspicion of DPE. When TVS is insufficient, second-line "modified-techniques" should be considered. Choosing the most effective technique is a challenge and should be based on patient history and clinical signs/symptoms.