79. Predicting Adnexal Torsion in Pediatric and Adolescent Patients

Abstract

<h3>Background</h3> Diagnosis of adnexal torsion is complicated by symptom overlap with other diagnoses. Failing to accurately diagnose adnexal torsion can lead to devastating reproductive consequences. Previous small research studies on pediatric and adolescent gynecology patients have developed diagnostic algorithms and composite scores using clinical and radiologic findings to help predict adnexal torsion and streamline which patients require surgical intervention. In our study, we evaluate the validity of a previously developed algorithm in a retrospective cohort of girls undergoing diagnostic laparoscopy for concern for adnexal torsion. <h3>Methods</h3> After IRB approval, a retrospective review of patients < 21 years of age who underwent diagnostic laparoscopy for possible torsion between January 2016 and April 2020 was performed. Patients were included if they had a CPT code diagnosis of torsion or operative notes suggesting torsion or concern for torsion, and if they had a pelvic ultrasound done prior to surgery. Each case was assigned a score based on an algorithm including presence of vomiting, ovarian volume, adnexal ratio, and menarchal status (Table 1). A composite score of 0-6 was calculated for each patient and compared to surgical findings. Based on a prior study, a score of 0-1 conferred a 3% risk of torsion, a score of 2-3 had a 3-10% risk of torsion, and a score greater than 4 had a > 25% risk of torsion. <h3>Results</h3> Among the charts reviewed, 103 met inclusion criteria. 92.2% had adnexal torsion, while 7.8% had no torsion. Of those with a composite score of ≥4, 96.7% were found to have torsion, compared to 92.9% with scores of 2-3 and 73.3% with scores of 0-1. A higher composite score was significantly associated with surgically-identified adnexal torsion by chi-squared analysis, X2 (2, N = 103) = 9.14, p = 0.010. Further analysis of each component score showed a significant relationship between vomiting (p=0.0015) and adnexal torsion. Adnexal volume and ratio scores were not independent predictors of torsion (p=0.121 and 0.726, respectively). <h3>Conclusions</h3> This data suggests higher composite scores and associated vomiting indicate a greater likelihood of torsion. However, clinical significance of these results is limited by the fact that low scores are not sufficient to exclude torsion. Further prospective research evaluating the validity of this algorithm is needed, and currently underway, in order to better elucidate the composite score's utility in clinical practice.