Time to Diagnosis of Eosinophilic Esophagitis Remains Prolonged Despite Increasing Knowledge of the Disease

Background: Eosinophilic esophagitis (EoE) represents a rapidly increasing cause of upper gastrointestinal morbidity, and patients often have symptoms for many years prior to diagnosis.While the number of studies describing this condition has grown exponentially since the 1990s and several iterations of guidelines have been published, it is unknown whether increased knowledge of this disease has translated into a more expedient diagnosis of symptomatic patients.Aim: To ascertain whether the symptomatic period before diagnosis decreased over time in a cohort of patients with EoE.Methods: We conducted a retrospective cohort study utilizing the University of North Carolina EoE Clinicopathologic Database between 2000 and 2016.This database contains data extracted from electronic medical records including: patient demographics, symptoms, treatment, endoscopic findings, and outcomes.We focused on the date of diagnosis and the length of symptoms reported prior to diagnosis.In addition to looking at symptom length per year, we divided dates into three time frames: prior to 2007 (before first guidelines), 2007-2011 (after first guidelines, and after 2011 (after second guidelines).Descriptive statistics were utilized to describe the cohort.Simple linear regression was used in the bivariate analysis.ANCOVA models assessed the relationship between symptom length before diagnosis and year of diagnosis adjusted for covariates.We also stratified by child (<18 yrs) or adult (≥18 yrs) status.Non-parametric testing was performed on variables non-normally distributed but did not change the results.Results: A total of 677 EoE cases were included in the study (mean age 26.1; 69% male; 81% white; baseline peak eosinophil count 65.4.eos/hpf; Table 1).The mean length of symptoms for all patients before diagnosis was 7.3 ± 8.6 years.Length of symptoms before diagnosis was 5.0 years from 2000-2006, 6.8 years from 2007 -2010, and 8.1 years from 2011 -2015 (p=0.07).In the ANCOVA model for all patients, the relationship was not significant (P = 0.68).When stratified by adults versus children, ANCOVA models also did not find a statistically significant correlation between length of symptoms and year of diagnosis (Figure 1), though there was an overall trend toward increased length of symptoms before diagnosis.Conclusions: Length of symptoms before diagnosis did not statistically correlate with year of diagnosis and has not decreased over time.Despite a rapidly increasing medical literature on EoE, there has paradoxically been a trend towards increased length of symptoms before diagnosis.Increased efforts should be made to raise awareness of this disease amongst the public and medical providers.