Predictive Model for the Non-Invasive Diagnosis of Endometriosis Based on Clinical Parameters

Abstract

Objectives: Are other pain symptoms in addition to dysmenorrhea, dyspareunia, dyschezia, dysuria, and chronic pelvic pain correlated to endometriosis and suitable for a clinical prediction model? Methods: We conducted a prospective study from 2016 to 2022, including a total of 269 women with numerous pain symptoms and other parameters. All women filled out two questionnaires and were examined by palpation and transvaginal ultrasound (TVUS). In cases of suspected deep endometriosis, magnetic resonance imaging (MRI) was performed. After the operation, endometriosis was diagnosed by histological examination. Results: All in all, 30 significant parameters and 6 significant numeric rating scale (NRS) scores associated with endometriosis could be identified: 7 pain adjectives, 8 endometriosis-associated pain symptoms, 5 pain localizations, 6 parameters from the PainDETECT, consumption of analgesics, and allergies. Furthermore, longer pain duration (before, during, and after menstruation) was observed in women with endometriosis compared to women without endometriosis (34.0% vs. 12.3%, respectively). Although no specific pain for endometriosis could be identified for all women, a subgroup with endometriosis reported radiating pain to the thighs/legs in contrast to a lower number of women without endometriosis (33.9% vs. 15.2%, respectively). Furthermore, a subgroup of women with endometriosis suffered from dysuria compared to patients without endometriosis (32.2% vs. 4.3%, respectively). Remarkably, the numbers of significant parameters were significantly higher in women with endometriosis compared to women without endometriosis (14.10 ± 4.2 vs. 7.75 ± 5.8, respectively). A decision tree was developed, resulting in 0.904 sensitivity, 0.750 specificity, 0.874 positive predictive values (PPV), 0.802 negative predictive values (NPV), 28.235 odds ratio (OR), and 4.423 relative risks (RR). The PPV of 0.874 is comparable to the positive prediction of endometriosis by the clinicians of 0.86 (177/205). Conclusions: The presented predictive model will enable a non-invasive diagnosis of endometriosis and can also be used by both patients and clinicians for surveillance of the disease before and after surgery. In cases of positivety, as evaluated by the questionnaire, patients can then seek advice again. Similarly, patients without an operation but with medical therapy can be monitored with the questionnaire.

Surgical Treatment of Patients with Endometriosis in the Certified Endometriosis Centers of the DACH Region – A Subanalysis of the Quality Assurance Study QS ENDO pilot

“…Our group recently showed that targeted taking of patientsʼ history using a patient questionnaire has a high sensitivity (90.4 %) for predicting the presence of endometriosis [25]. The pain intensity can therefore function as a discriminating parame-ter, although no conclusions can be drawn about the extent of disease.…”

“…Unsere Arbeitsgruppe konnte kürzlich zeigen, dass durch eine gezielte Anamnese anhand eines Patientinnenfragebogens das Vorliegen einer Endometriose mit hoher Sensitivität (90,4 %) vorausgesagt werden kann [25]. Insofern kann die Schmerzintensität durchaus als diskriminierender Parameter dienen, allerdings ohne Rückschlüsse auf die Ausdehnung zuzulassen.…”