Sensitivity and Specificity of Administrative Medical Coding for Pediatric Eosinophilic Esophagitis

Examining Disparities in Pediatric Eosinophilic Esophagitis

BACKGROUND

Little is known regarding the impact of race, ethnicity, and socioeconomic status on the health outcomes of children with eosinophilic esophagitis (EoE).

OBJECTIVE

To (1) identify demographic characteristics of children diagnosed with EoE in a large tertiary care center, and (2) determine associations between a patient's demographics and depth of evaluation or treatment choices.

METHODS

This retrospective cohort study included children 0 to 18 years old seen in Children's Hospital Colorado between January 1, 2009, and December 31, 2020. Demographics were extracted from the electronic medical record. Rural-Urban Commuting Area taxonomy codes were used to classify urbanization. Area Deprivation Index (ADI) scores were used to categorize neighborhood advantage/disadvantage. Data were analyzed using descriptive statistics and regression analysis.

RESULTS

The study included 2,117 children with EoE. Children with higher state ADI scores (greater neighborhood disadvantage) had less radiographic evaluation of their disease (odds ratio [95% CI] per unit increase in state ADI = 0.93 [0.89-0.97]; P = .0002) and had esophageal dilations at younger ages (r = -0.24; P = .007). Black children compared with White children were younger at diagnosis (8.3 y vs 10.0 y; P = .002). Children from rural areas were seen less by feeding therapy (3.9% vs 9.9%; P = .02), but were younger at their visits (2.3 y vs 4.3 y; P < .001).

CONCLUSIONS

In this study of children with EoE cared for in a large tertiary care center, we found differences in presentation and care depending on race, urbanization, and socioeconomic status.

Eosinophilic Gastritis and Enteritis Are Increased in Families With Eosinophilic Esophagitis

INTRODUCTION

There are limited data on the familial risk of distal eosinophilic gastrointestinal diseases (EGIDs) in patients with eosinophilic esophagitis (EoE). We analyzed the risk of eosinophilic gastritis/gastroenteritis (EG/EGE) and eosinophilic colitis (EC) as forms of distal EGIDs using International Disease Classification-9/10 codes in subjects with EoE and their relatives.

METHODS

The Utah Population Database is a resource that links genealogy information and medical records in Utah. We identified EGIDs in probands and their first-degree (FDRs), second-degree (SDRs), and third-degree (TDRs) relatives in the Utah Population Database. Relative risk and 95% confidence intervals were estimated. All individuals with inflammatory bowel disorder were eliminated to avoid misdiagnosis with EGIDs.

RESULTS

We included 8,455 subjects with EoE, 396 with EG/EGE, and 172 with EC. Probands with EoE were at increased risk of EG/EGE and EC. Risks of EG/EGE were increased among FDRs and SDRs of probands with EoE , even without concomitant EoE in the relatives. Increased risk of EG/EGE in FDRs and SDRs was also present for EoE probands without EG/EGE or EC. We observed no isolated familial aggregation of EG/EGE after excluding cases with comorbid EoE. EC probands without EoE were at increased risk of EG/EGE, but no evidence of familial risk of EC was observed.

DISCUSSION

The relative risk of EG/EGE is significant among relatives of patients with EoE, suggesting that shared genetic factors exist among these EGIDs. EG/EGE and EC showed limited familial clustering, although sample sizes were small.

Esophageal Eosinophilia Is Common Among Relatives of Eosinophilic Esophagitis Patients

BACKGROUND & AIMS

Familial clustering of eosinophilic esophagitis (EoE) has been described, and we report on the biopsy-assessed prevalence of esophageal eosinophilia (EE) in first-degree family members. The aim was to determine the prevalence of EE in first-degree adult relatives (FDRs) of EoE patients.

METHODS

Index EoE patients diagnosed by EE (>15 eosinophils per high-power field) and proton pump inhibitor nonresponsiveness were identified and family trees were constructed. Adult FDRs were invited to undergo upper endoscopy with esophageal biopsies and to complete reflux, dysphagia, and allergy/atopy questionnaires. Questionnaire information was gathered only for those who responded as per institutional review board purview. Records from other children and adult FDRs with prior EoE diagnoses also were obtained when permission was obtained. Simple and multivariable logistic regression models were used to evaluate the unadjusted and odds ratios of EoE for demographic and clinical variables.

RESULTS

A total of 239 FDRs from 37 index EoE patients were identified. Seventy-one of 239 adult (age, >18 y) FDRs completed endoscopy and questionnaires and 18 of 71 FDRs had EE. An additional 17 FDRs were confirmed to have EE after external medical record retrieval, resulting in a total of 35 of 239 (14.6%) FDRs with EE. Significantly more male FDRs had EE compared with female FDRs (P = .027). Proton pump inhibitors, dysphagia, gastroesophageal reflux disease, asthma, and reflux symptoms predicted EE in FDRs. FDRs who had EE reported hay fever, allergic eye symptoms, and food allergy more frequently than those without EE (P = .03, P = .001, and P = .02, respectively). Specifically, younger age, higher serum eosinophils, being male, and having food allergies all were associated with higher odds of EoE (P = .0211, P = .0031, P = .0362, and P = .0089, respectively).

CONCLUSIONS

The prevalence of esophageal eosinophilia is extremely high and male-predominant in first-degree relatives of EoE patients. Symptoms of hay fever, allergic eye symptoms, and food allergy were predictors of EE in FDRs. Dysphagia did not predict esophageal eosinophilia. Family members of EoE patients are at risk for EE, particularly those who have atopic symptoms.