Systematic review reveals key clinical, laboratory, and ultrasound predictors of ovarian torsion in emergency patients, indicating composite multivariable models optimize diagnostic accuracy.
Background: Ovarian torsion is a gynecologic emergency in which nonspecific symptoms and imperfect imaging complicate early recognition in emergency care. In this study we aimed to analyze original articles on clinical, laboratory, and ultrasound predictors associated with surgically confirmed ovarian or adnexal torsion in females presenting with acute symptoms. Methods: A PRISMA-aligned systematic review framework was constructed for PubMed, Scopus, Web of Science Core Collection, and the Cochrane Library from inception through August 2026. Eligible studies evaluated preoperative predictors of ovarian or adnexal torsion in females assessed in emergency settings and used operative findings as the reference standard. Prediction-model and diagnostic-factor studies were analyzed qualitatively because predictor definitions, thresholds, age groups, and modeling strategies were heterogeneous. Results: Eighteen original studies were included. Recurrent predictors include unilateral or localized pain, nausea or vomiting, shorter symptom duration, leukocytosis or elevated neutrophil-to-lymphocyte ratio, ovarian enlargement or increased ovarian volume ratio, ovarian stromal edema, free pelvic fluid, abnormal ovarian position, and the whirlpool sign. Multivariable and composite models performed better than isolated symptoms, laboratory values, or Doppler findings. Pediatric models integrating age, inflammatory indices, and sonographic morphology reported discrimination in internal validation. Conclusions: Ovarian torsion prediction is strongest when history, examination, inflammatory markers, and ultrasound morphology are integrated.
No takes yet. Share an insight, caveat, or question.
Alanazi et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: