M. Carter1, J. Afowork1, J. B. Pitt1, S. A. Ayala1, S. D. Goldstein1 1Ann and Robert H. Lurie Children’s Hospital of Chicago, Division Of Pediatric Surgery, Chicago, IL, USA
Introduction: Children who present with intussusception without evidence of bowel compromise or pathologic lead point typically undergo non-operative management by enema reduction. However, failure of such necessitates operation. Predictors of failed non-operative management are not routinely considered in these patients. The purpose of this study is to create a scoring system that predicts failure of non-operative management and need for surgical intervention.
Methods: After institutional approval (IRB #2023-6231), children diagnosed with intussusception upon presentation to the emergency department of a high-volume, tertiary children’s hospital between January 2019 and December 2022 were retrospectively identified. Patients were stratified into two cohorts based on success or failure of non-operative management that presentation. Continuous variables were split into groups with optimal cut-point identified by Youden’s method and demographics and clinicopathologic variables were compared by Chi-squared. Univariable logistic regression was performed to identify predictors of non-operative failure (p<0.05) and used as covariates for multivariable logistic regression with final model determined by backward elimination method. Odds ratios (OR) for final predictors independently associated with non-operative failure were rounded to nearest whole number to create the scoring system. Optimal cut-points were identified indicative of low, moderate and high risk of failure.
Results: We identified 121 patients accounting for 143 presentations. The majority were their 1st presentation for intussusception (84.6%), and 115 (80.4%) were managed non-operatively while 28 (19.6%) required operative intervention. Most were male (64.3%), Hispanic (43.4%) and publicly insured (54.5%) with no difference between cohorts. Independent predictors of failed non-operative management identified on univariable and multivariable analysis include age ≥4 years (OR 32.7, 95% CI 5.9-224.5), ≥2 reduction attempts (OR 39.3, 95% CI 11.0-182.8) presenting heart rate ≥128 (OR 4.7, 95% CI 1.4-19.0) and presenting systolic blood pressure ≥ 115 mmHg (OR 4.8, 95% CI 1.2-20.3) and were utilized as the risk scoring system (maximum score 82, AUC 0.89 95% CI 0.81-0.96, Figure). Scores <33 were determined to have low failure risk (4.2%), 33-44 moderate failure risk (63.3%), and >44 high failure risk (100%).
Conclusion: Using basic clinical parameters, we produced a risk scoring system with strong predictive ability for failure of non-operative management in children with intussusception. Additional studies are warranted to evaluate this risk calculator’s performance when applied prospectively to a diverse multi-institutional test cohort.
