Endometriosis Index: A software-derived score to\npredict the presence and severity of the disease

Objective: To develop a clinical index that will positively predict the presence of endometriosis.\n\nDesign: Prospective single-center observational study. Setting: A hospital-based institute.\n\nPatients: 120 patients affected by chronic pelvic pain, infertility or with clinical suspicion of endometriosis.\n\nInterventions: Electronic processing of clinical data with software-assistance at the end of each\n\nconsultation and digital video recording of surgeries.\n\nMain Outcome Measures: Endometriosis Index (EI), the score calculated using clinical parameters\n\ncorrelated with macroscopic/microscopic presence or absence of endometriosis.\n\nResults: Endometriosis was staged and treated in 95 cases, the remaining 25 women presented benign\n\npathology with no endometriosis. Patients with positive operative findings of endometriosis had\n\na mean (± standard deviation) pre-operative EI score of 22 ± 12, while mean EI value of patients with\n\nno operative findings of endometriosis was 8 ± 6 (p<0.001). Cumulative distribution of EI shows increasing\n\nvalues from controls to peritoneal, ovarian and deeply infiltrating endometriosis (DIE) (p for trend\n\n 28 (75th percentile value) versus women with EI score ≤ 28; with EI score > 28 this noninvasive\n\ntest was predictive of DIE with a sensitivity of 72.4% and a specificity of 90.1%.\n\nConclusions: These data suggest that a dedicated Endometriosis Index is effective in identifying patients\n\nwho would benefit from early surgical management. We propose the use of this non-invasive\n\ntool to reduce the delay between the onset of symptoms and a surgical diagnosis of endometriosis.