RESUMO
OBJECTIVE: Our aim was to perform a systematic review and meta-analysis of the most commonly used examinations for rectosigmoid lesions of deeply infiltrating endometriosis, transvaginal sonography (TVS) and magnetic resonance imaging (MRI), to compare their diagnostic accuracy and enhanced or non-enhanced techniques. METHODS: A systematic search was performed until March 2018 without time or language restrictions. Eligibility criteria included studies that compared the accuracy of TVS and MRI for diagnosis of rectosigmoid endometriosis. The quality of the studies was assessed by means of Quality Assessment of Diagnostic Accuracy Studies-2 and Preferred Reporting Items for Systematic Reviews and Meta-Analyses recommendations. Bivariate and hierarchical analysis were performed. The difference in the accuracy of TVS and MRI was tested, and heterogeneity was addressed by means of meta-regression, sensitivity, or subgroup analysis. RESULTS: A total of 1754 studies were screened; 105 studies were eligible, and 11 studies were included in the meta-analysis. Overall pooled sensitivity, specificity, and area under the receiver operating characteristic curve were 0.80, 0.94, and 0.95, respectively. The measures for MRI were 0.82, 0.94, and 0.95, respectively. There was no statistical difference between the accuracy values of TVS and MRI (Pâ¯=â¯0.90). The use of bowel preparation and vaginal contrast could enhance the accuracy of MRI. Along with rectosigmoid prevalence, bowel and vaginal contrast explained a significant proportion of the statistical heterogeneity. CONCLUSIONS: Both TVS and MRI showed high diagnostic accuracy for rectosigmoid deeply infiltrating endometriosis lesions. There is no strong evidence suggesting that the two diagnostic methods might differ in specificity or sensitivity, but enhanced techniques may increase the accuracy measures.
