Vicki Nisenblat, Lucy Prentice, Patrick MM Bossuyt, Cindy Farquhar, M Louise Hull, Neil Johnson

About 10% of women of reproductive age suffer from endometriosis, a costly chronic disease causing pelvic pain and subfertility. Laparoscopy is the gold standard diagnostic test for endometriosis, but is expensive and carries surgical risks. Currently, there are no non-invasive tests available in clinical practice to accurately diagnose endometriosis. This review assessed the diagnostic accuracy of combinations of different non-invasive testing modalities for endometriosis and provided a summary of all the reviews in the non-invasive tests for endometriosis series. To estimate the diagnostic accuracy of any combination of non-invasive tests for the diagnosis of pelvic endometriosis (peritoneal and/or ovarian or deep infiltrating) compared to surgical diagnosis as a reference standard. The combined tests were evaluated as replacement tests for diagnostic surgery and triage tests to assist decision-making to undertake diagnostic surgery for endometriosis. We did not restrict the searches to particular study designs, language or publication dates. We searched CENTRAL to July 2015, MEDLINE and EMBASE to May 2015, as well as the following databases to April 2015: CINAHL, PsycINFO, Web of Science, LILACS, OAIster, TRIP, ClinicalTrials.gov, DARE and PubMed. We considered published, peer-reviewed, randomised controlled or cross-sectional studies of any size, including prospectively collected samples from any population of women of reproductive age suspected of having one or more of the following target conditions: ovarian, peritoneal or deep infiltrating endometriosis (DIE). We included studies comparing the diagnostic test accuracy of a combination of several testing modalities with the findings of surgical visualisation of endometriotic lesions. Three review authors independently collected and performed a quality assessment of the data from each study by using the QUADAS-2 tool. For each test, the data were classified as positive or negative for the surgical detection of endometriosis and sensitivity and specificity estimates were calculated. The bivariate model was planned to obtain pooled estimates of sensitivity and specificity whenever sufficient data were available. The predetermined criteria for a clinically useful test to replace diagnostic surgery were a sensitivity of 0.94 and a specificity of 0.79 to detect endometriosis. We set the criteria for triage tests at a sensitivity of 0.95 and above and a specificity of 0.50 and above, which 'rules out' the diagnosis with high accuracy if there is a negative test result (SnOUT test), or a sensitivity of 0.50 and above and a specificity of 0.95 and above, which 'rules in' the diagnosis with high accuracy if there is a positive result (SpIN test). Eleven eligible studies included 1339 participants. All the studies were of poor methodological quality. Seven studies evaluated pelvic endometriosis, one study considered DIE and/or ovarian endometrioma, two studies differentiated endometrioma from other ovarian cysts and one study addressed mapping DIE at specific anatomical sites. Fifteen different diagnostic combinations were assessed, including blood, urinary or endometrial biomarkers, transvaginal ultrasound (TVUS) and clinical history or examination. We did not pool estimates of sensitivity and specificity, as each study analysed independent combinations of the non-invasive tests.Tests that met the criteria for a replacement test were: a combination of serum IL-6 (cut-off >15.4 pg/ml) and endometrial PGP 9.5 for pelvic endometriosis (sensitivity 1.00 (95% confidence interval (CI) 0.91 to 1.00), specificity 0.93 (95% CI, 0.80, 0.98) and the combination of vaginal examination and transvaginal ultrasound (TVUS) for rectal endometriosis (sensitivity 0.96 (95% CI 0.86 to 0.99), specificity 0.98 (95% CI 0.94 to 1.00)). Tests that met the criteria for SpIN triage tests for pelvic endometriosis were: 1. a multiplication of urine vitamin-D-binding protein (VDBP) and serum CA-125 (cut-off >2755) (sensitivity 0.74 (95% CI 0.60 to 0.84), specificity 0.97 (95% CI 0.86 to 1.00)) and 2. a combination of history (length of menses), serum CA-125 (cut-off >35 U/ml) and endometrial leukocytes (sensitivity 0.61 (95% CI 0.54 to 0.69), specificity 0.95 (95% CI 0.91 to 0.98)). For endometrioma, the following combinations qualified as SpIN test: 1. TVUS and either serum CA-125 (cut-off ≥25 U/ml) or CA 19.9 (cut-off ≥12 U/ml) (sensitivity 0.79 (95% CI 0.64 to 0.91), specificity 0.97 (95% CI 0.91 to 1.00)); 2. TVUS and serum CA 19.9 (cut-off ≥12 U/ml) (sensitivity 0.54 (95% CI 0.37 to 0.70), specificity 0.97 (95% CI 0.91 to 1.0)); 3-4. TVUS and serum CA-125 (cut-off ≥20 U/ml or cut-off ≥25 U/ml) (sensitivity 0.69 (95% CI 0.49 to 0.85), specificity 0.96 (95% CI 0.88 to 0.99)); 5. TVUS and serum CA-125 (cut-off ≥35 U/ml) (sensitivity 0.52 (95% CI 0.33 to 0.71), specificity 0.97 (95% CI 0.90 to 1.00)). A combination of vaginal examination and TVUS reached the threshold for a SpIN test for obliterated pouch of Douglas (sensitivity 0.87 (95% CI 0.69 to 0.96), specificity 0.98 (95% CI 0.95 to 1.00)), vaginal wall endometriosis (sensitivity 0.82 (95% CI 0.60 to 0.95), specificity 0.99 (95% CI 0.97 to 1.0)) and rectovaginal septum endometriosis (sensitivity 0.88 (95% CI 0.47 to 1.00), specificity 0.99 (95% CI 0.96 to 1.00)).All the tests were evaluated in individual studies and displayed wide CIs. Due to the heterogeneity and high risk of bias of the included studies, the clinical utility of the studied combination diagnostic tests for endometriosis remains unclear. None of the biomarkers evaluated in this review could be evaluated in a meaningful way and there was insufficient or poor-quality evidence. Laparoscopy remains the gold standard for the diagnosis of endometriosis and using any non-invasive tests should only be undertaken in a research setting.