O2174: To Biopsy or Not? Development of a Risk Calculator to Stratify Children Presenting with Constipation for Hirschsprung Disease Assessment

Background: Hirschsprung Disease (HD) is in the differential diagnosis for the child presenting with constipation, however, this diagnosis is frequently delayed, especially in children with shorter segment disease. Diagnosis can only be ruled out by rectal biopsy or anorectal manometry, but undergoing these procedures in every child adds unnecessary risk and cost. Risk calculators have been used in order to risk-stratify patients who should proceed with further work-up for many conditions. The purpose of this study is to develop a risk calculator based on initial clinical and radiological findings to aid in determining the appropriateness of rectal biopsy in children with chronic constipation.

Methods: A multicenter study was performed in conjunction with the American Pediatric Surgical Association Hirschsprung Disease Interest Group and the Pediatric Colorectal and Pelvic Learning Consortium. All children aged 2-15 years presenting to a gastroenterologist or pediatric surgeon with chronic constipation who underwent an eventual confirmatory study were included. Multivariable analysis with backwards stepwise variable selection was performed to determine risk factors associated with the confirmation of HD. Each risk factor was given an associated score based on the effect size of its association. Sensitivity and specificity of different cut points for the total score was used to derive an optimal score to rule out the need for further HD work-up.

Results: A total of 984 children met criteria (median age: 6 years, 55.6% male), of which 67 (6.8%) were eventually diagnosed with HD. On univariable analysis, significant differences between HD and non-HD groups included male sex (76.1% vs 54.1%, p=0.001), family history of HD (9.0% vs 1.4%, p=0.001), abdominal distension on presentation (40.3% vs 22.6%, p=0.002), length of constipation (new onset: 14.9% vs 36.6%, p< 0.001), and contrast enema results (positive or equivocal for transition zone or reversed rectosigmoid ratio: 44.8% vs 6.9%, p< 0.001). After backwards stepwise variable selection, age < 6 years (adjusted odds ratio [AOR]: 2.11 (95% confidence interval [CI]: 1.20-3.69), male sex (AOR: 2.46, 95% CI: 1.33-4.54), family history of HD (AOR: 7.60, 95% CI: 2.46-23.4), length of constipation (not new onset, AOR: 3.52, 95% CI: 1.70-7.29), and findings on contrast enema (positive or equivocal results for HD, AOR: 11.3 (95% CI: 6.28-20.3) remained in the model and were used to design the risk calculator (Table). A cut off score of ≥3 demonstrated a sensitivity of 91.0% and specificity of 47.1% for the prediction of HD (c-statistic: 0.814, Figure).

Conclusion: While HD is in the differential diagnosis for children presenting with constipation at an older age, invasive work-up among all presenting children leads to an unnecessary risk of complications and increased cost. Using data from a multicenter study, we have developed a novel risk calculator that can accurately determine which children should proceed with further work-up.