University of North Carolina School of Medicine Chapel Hill, North Carolina
Introduction: Diagnosing non-esophageal eosinophilic gastrointestinal diseases (EGIDs) remains challenging despite increasing awareness, and misdiagnosis is possible. We aimed to determine whether the length of time preceding EGID diagnosis has decreased over time, and to assess the presence of prior non-EGID diagnoses.
Methods: We conducted a retrospective cohort study utilizing the UNC EGID Clinicopathologic Database over a two decade time span. This database contains demographics, clinical characteristics, and procedural data extracted from electronic medical records. EGID subjects were diagnosed with eosinophilic gastritis (EoG) and/or enteritis (EoN) based on clinical presentation and biopsy results. We calculated the length of symptoms prior to diagnosis. ANCOVA models assessed the relationship between symptom length before diagnosis and year of diagnosis adjusted for covariates.
Results: A total of 67 patients were included, with 55% male and 72% white (Table). An atopic condition was diagnosed in 48% of adults and 65% of children. The length of symptoms prior to diagnosis for all patients was 37.1 ± 57.1 months, which was longer in adult patients compared to children (60.2±82.4 vs 22.9±25.6; p=0.01). Prior alternative diagnoses for EGID symptoms were common, with 55% carrying a different pre-EGID diagnosis. The most frequent included non-specific inflammation (10%), GERD (6%), other functional disorder (6%), IBS (4%), dysmotility (4%), infection (4%), and other (13%). Symptom length preceding diagnosis did not significantly associate with year of diagnosis on bivariate (p = 0.58) or multivariate analysis after adjusting for covariates (p = 0.88) (Figure). No association was found for symptomatic period and year of diagnosis when stratified by children (p = 0.26) or adults (p = 0.40). Additionally, multivariate linear regression between length of symptoms and year of diagnosis found no association and regression with year as a continuous variable was also not significant.
Discussion: Despite increased knowledge regarding EGIDs, a trend towards decreased symptom length prior to diagnosis was not found. Numerically, the time of diagnosis decreased somewhat after the year 2006, though this did not reach statistical significance. Prior to their EGID diagnosis, approximately 1 in 2 patients received an incorrect diagnosis. These data suggest that more efforts should be made to raise awareness of these diseases among medical providers.
