DISPARITIES IN OSTOMY CREATION FOR PERIANAL FISTULIZING COMPLICATIONS IN PEDIATRIC PATIENTS WITH CROHN'S DISEASE WITHIN A MULTI-INSTITUTIONAL COHORT
Samir Gadepalli*1, Brendan M. Boyle2, Mallory Chavannes3, Matthew D. Egberg4, Jason S. Frischer5, Christopher Gayer3, Prashanthi Kandavel3, Sandra C. Kim6, Lisa McMahon7, Phillip P. Minar5, Kevin Mollen6, Benedict Nwomeh2, Brad Pasternak7, Michael R. Phillips4, Jeremy Adler1
1Pediatric Surgery, University of Michigan, Ann Arbor, MI; 2Nationwide Children's Hospital, Columbus, OH; 3Children's Hospital Los Angeles, Los Angeles, CA; 4The University of North Carolina at Chapel Hill, Chapel Hill, NC; 5Cincinnati Children's Hospital Medical Center, Cincinnati, OH; 6UPMC, Pittsburgh, PA; 7Phoenix Children's Hospital, Phoenix, AZ
Background: When children with Crohn's disease (CD) have medically refractory disease, a diverting ostomy may be needed. As part of an antibiotic stewardship study, we sought to determine risk of ostomy in a large pediatric cohort.
Methods: We identified pediatric patients (≤21 years old) diagnosed with CD and enrolled in ImproveCareNow Network (ICN), a multicenter, multidisciplinary pediatric inflammatory bowel disease quality improvement collaborative. Patients within 90 days of CD diagnosis (2007-2018), and with ≥3 visits, were identified. The registry includes prospectively collected data from outpatient pediatric gastroenterology visits. We gathered data on demographics (age, sex, race, insurance), medications used, and presence of ostomy. Insurance was categorized as commercial insurance (with or without public insurance) or non-commercial (public or no insurance). Disease behavior uses Paris phenotype classification (B1-non-stricturing/non-penetrating, B2-stricturing, B3-penetrating, B2/B3-stricturing/penetrating, PFC-perianal fistulizing complications). Primary outcome was creation of ostomy from onset of CD diagnosis in those with and without PFC. Descriptive data and logistic regression were performed with p<0.05 considered significant.
Results: Of 5,720 patients from 80 ICN sites (median age 13.5yr, 40.5% female, 18.7% non-White, 78% commercial insurance), 1,023 (30.7%) had PFC without differences by sex (p=0.70), age at diagnosis (p=0.19), race (p=0.14), or insurance (p=0.27). Ostomies were not more common with PFC (2.9%vs.2.3%; p=0.21) and antibiotic treatment of PFC was not different among those with an ostomy (21.1%vs.26.9%; p=0.6) or by insurance (p=0.50).
Although no differences phenotype classification by race (p=0.77), B2/B3 phenotypes were more common in those without commercial insurance (18.6%vs.13.1%; p=0.001), older age at CD diagnosis (p<0.001) and in male patients (p=0.013). Ostomies were more common among those without commercial insurance (3.8%vs.2.3%; p=0.010), among Hispanic patients (5.3%vs.2.4%; p=0.008) and among Non-White patients (3.5%vs.2.3%; p=0.027). Ostomies were not associated with age at CD diagnosis (<10yr, 1.9%; 10-17yr, 2.6%; ≥17yr, 2.4%; p=0.44), or sex (male 2.2%, female 2.8%; p=0.17). Ostomies were more common among patients with aggressive phenotype: B1,1.8%; B2,5.3%; B3,6.2%; B2/B3,10.1% (p<0.001). After controlling for disease severity, age, sex, presence of PFC, and medications (Table), White patients with and without commercial insurance were least likely to need an ostomy (OR 0.42, p=0.009, OR 0.37, p=0.029).
Conclusion: Ostomies for CD are more commonly done in children without commercial insurance, in Non-White patients, and with an aggressive disease phenotype. These disparities are unexplained by antibiotic use or presence of PFC.
Back to 2023 Abstracts