# Non-Invasive Diagnosis of Endometriosis by Questionnaires in Patients Using Contraception

## TL;DR

This study shows that questionnaires can reliably predict endometriosis in patients using contraception, potentially reducing diagnostic delays.

## Contribution

The study introduces a non-invasive questionnaire-based method for diagnosing endometriosis with high accuracy.

## Key findings

- EMS-positive patients had ~4-fold higher pain scores compared to EMS-negative patients.
- A decision tree achieved high sensitivity and specificity for predicting endometriosis.
- Combining dysuria and lightning-like pain improved diagnostic accuracy significantly.

## Abstract

Background/Objectives: The assessment of endometriosis (EMS)-associated pain is important, but only very few studies address the potential use of questionnaires for non-invasive prediction of the disease. Methods: In a prospective observational study from 2016 to 2024 with patients (n = 228) using hormonal contraception, all women with suspected EMS answered two questionnaires and were examined physically and with transvaginal ultrasound (TVUS). If deep infiltrating EMS (DIE) was suspected, magnetic resonance imaging (MRI) was performed. EMS diagnosis was confirmed by histological examination. Statistical analysis was mainly performed using 2 × 2 contingency tables. The decision tree was created manually. Results: The mean numerical rating scales (NRSs) of EMS-positive compared to EMS-negative patients were ~4-fold higher (4.45 and 1.15, respectively). Patients with EMS have, significantly, ~3 times more significant parameters compared to patients without EMS (18.5 and 5.9, respectively). In combination with dysuria and lightning-like pain, this resulted in very good prediction. A decision tree yielded a sensitivity of 0.924, a specificity of 0.917, a positive predictive value (PPV) of 0.924, a negative predictive value (NPV) of 0.917, and a positive likelihood ratio of 11.2, indicating a very good diagnostic test. There is no typical endometriosis pain, but various pain patterns are predictive of EMS. Conclusions: Thus, a reliable non-invasive EMS diagnosis by questionnaires is possible and could reduce the delay in the detection of EMS.