**Background:** Eosinophilic esophagitis (EoE) is a chronic immune-mediated esophageal condition diagnosed by symptoms and histology (≥15 eosinophils per high power field). The EoE Endoscopic Reference Score (EREFS) is a validated system (score 0–9) that grades five endoscopic features: exudates, rings, edema, furrows, and strictures. Despite its diagnostic and prognostic value, EREFS remains underused in clinical practice. This quality improvement (QI) initiative aimed to increase EREFS documentation during repeat endoscopies for known EoE patients to >90% and to assess correlation with histologic disease activity.
**Methods:** The QI project was conducted at Children’s Hospital Colorado Digestive Health Institute from October 2018 through February 2020. Inclusion criteria: any endoscopy in the Main Campus Procedure Center for patients with a prior EoE diagnosis (defined by problem list, procedure indication, or prior histology >15 eos/hpf plus clinical documentation). Baseline EREFS completion rate was determined via electronic medical record (EMR) review. Three PDSA interventions were implemented sequentially: (1) June 2019 – automatic drop-down EREFS menu added to the EGD documentation template; (2) August 2019 – in-person educational session for faculty and fellows on EREFS completion and clinical value; (3) November 2019 – addition of EREFS menus to EGD+dilation and EGD+colonoscopy templates, placement of laminated EREFS visual guides at workstations, and an educational session at the monthly faculty meeting. Data collected included demographics, procedure type, endoscopist (faculty ± fellow), EREFS total and subscores, biopsy locations, and peak esophageal eosinophil count. Statistical analyses included descriptive statistics, Fisher exact test, chi-square test, Spearman’s rank correlation, segmental linear regression, and nonparametric ANOVA (Kruskal-Wallis and Wilcoxon rank-sum). Significance level was 0.05; 90% confidence intervals were reported for one-tailed inference.
**Key Results:** Over 12 months, 542 endoscopies were performed on 410 patients (68% male, mean age 10.9 years [SD 5.7], mean EREFS score 2.14 [SD 1.88], mean peak eosinophil count 30.9 eos/hpf [SD 37.1]). Baseline EREFS completion rate was 72.7% (90% CI, 67.4-77.4) from 170 of 234 cases. After the first intervention, completion was 80.6% (90% CI, 70.9-88.1; 54/67); after the second, 78.9% (90% CI, 71.2-85.2; 82/104). After all three interventions, completion rate significantly improved to 94.9% (90% CI, 90.6-97.6; 130/137; P<0.001), exceeding the 90% target. By procedure type, pre-intervention rates were 66.0% for EGD, 66.7% for EGD+colonoscopy, and 66.7% for EGD+dilation; post-intervention rates rose to 94%, 100%, and 95.8%, respectively (P<0.0001 for EGD, P=0.09 for EGD+colonoscopy, P=0.05 for EGD+dilation). A significant positive correlation was found between EREFS total score and peak eosinophil count in the total cohort (Spearman ρ=0.61, P<0.001), and this correlation persisted pre- and post-interventions. For each EREFS component (edema, exudate, furrows, rings, stricture), a positive score was associated with significantly higher peak eosinophil counts compared to a score of 0.
**Clinical Implications:** This QI initiative demonstrates that simple, targeted interventions—particularly EMR template integration and provider education—can dramatically increase and sustain EREFS documentation rates in a pediatric tertiary care setting. Achieving >90% completion ensures that endoscopic findings are systematically captured, providing an additional objective measure alongside symptoms and histology for assessing treatment response in EoE. The maintained correlation between EREFS and eosinophil counts supports the validity of the scoring system even when used by a diverse group of endoscopists. These findings have direct implications for standardizing EoE care, improving clinical decision-making, and facilitating multi-center research. Limitations include the single-center design, potential selection bias from a specialized EoE team, and inclusion of trainees who may not represent all practice settings. Ongoing PDSA cycles are needed to assess long-term durability.