Risk of Barrett's Esophagus After a Negative Index Endoscopy: An Analysis Using the GIQUIC National Benchmarking Registry

Introduction: Current guidelines recommend single screening endoscopy in patients with multiple risk factors for Barrett’s esophagus (BE). Data suggesting a low risk of BE after a negative esophagogastroduodenoscopy (EGD) are limited by small sample size and short follow-up after initial EGD. There remains a possibility of missed or incident BE after a negative index EGD. With the advent of cost-effective, non-endoscopic BE screening tools, repeat screening may be a consideration in high-risk patients. We aimed to determine the prevalence and predictors of BE after a negative index evaluation, on repeat EGD in a large national endoscopic database. Methods: We analyzed data from the GI Quality Improvement Consortium Registry (GIQuIC), a large nationwide quality benchmarking clinical registry, from 2013-2020. We included patients who underwent at least 2 EGDs. Patients diagnosed with or with a history of BE or esophageal adenocarcinoma (EAC) at index EGD were excluded. Data on prevalence of BE/EAC on second/subsequent EGDs, and known risk factors for BE were collected. Univariate and multivariate logistic regression analyses were performed to assess association between predictors and outcome of BE/EAC on repeat EGD. Results: The prevalence of BE at index endoscopy in the GIQuIC database is 4.2%. A total of 124,223 patients underwent at least 1 EGD (mean number of repeat EGDs 1.39, range 1-63). Of these, 2,272 (1.83%, 95% CI 1.75, 1.90%) were found to have BE/EAC. Table shows the prevalence of BE/EAC stratified by risk factors. Risk factors associated with BE/EAC on repeat endoscopy included GERD (OR: 3.36, p < 0.01), male sex (OR: 1.85, p< 0.01), White race (OR: 1.76, p< 0.01), age 50-80 years (OR: 1.53, p< 0.01), and obesity (OR: 1.20, p=0.04). In patients with GERD and an additional risk factor, the prevalence of BE/EAC was higher at 3.4% at a mean (SD) time interval of 13.1 (14.1) months after a negative index EGD. The prevalence of BE/EAC increased with increasing number of risk factors (Figure). Conclusion: The prevalence of BE/EAC after an initial negative index EGD was approximately 2% (44% of baseline BE prevalence in GIQuIC), with most cases diagnosed within 5 years. In patients with two or more risk factors, the prevalence was two fold higher. Repeat evaluation for BE, particularly with non-endoscopic techniques, may be considered in patients with multiple risk factors, 1-5 years after initial negative evaluation.Figure 1.: The distribution of BE/EAC diagnosed after initial negative endoscopy stratified by the number of risk factors present per patient and GERD. Table 1. - Baseline characteristics of patients with and without BE/EAC on follow-up EGD after negative index EGD Characteristic No BE/EAC (N = 121,951, 98.17%) BE/EAC (N = 2,272, 1.83%) p-value Age (years) < 0.0001 < 50 27,466 (22.8%) 411 (18.1%) 50-80 85,262 (69.9%) 1,762 (77.6%) >80 8,890 (7.3%) 99 (4.4%) Male sex 48,224 (39.5%) 1,247 (54.9%) < 0.0001 White race 82,8384 (82.8%) 1,663 (90.0%) < 0.0001 GERD* 39,035 (32.0%) 1,346 (59.2%) < 0.0001 Obesity (BMI >30) 6,600 (35.5%) 217 (40.3%) 0.0204 Time interval between EGDs < 0.0001 < 1year 69,485 (57.0%) 1,475 (64.9%) 1 to < 3 years 41,828 (34.3%) 617 (27.2%) v3 to < 5 years 10,134 (8.3%) 174 (7.7%) >=5 years 504 (0.4%) 6 (0.3%) Indication for second EGD Screening for BE 1,218 (1.0%) 130 (5.7%) < 0.0001 Reflux 39,035 (32.0%) 1,346 (59.2%) < 0.0001 Ulcer 15,071 (12.4%0 264 (11.6%) 0.2890 Weight loss 2,493 (2.0%0 19 (0.8%) < 0.0001 Dysphagia 41,391 (33.9%) 547 (24.1%) < 0.0001 Esophagitis 11,607 (9.5%) 570 (25.1%) < 0.0001 GI bleeding/anemia 25,296 (20.7%) 227 (10.0%) < 0.0001 Vomiting 6,657 (5.5%) 84 (3.7%) 0.0002 Other 12,346 (10.1%) 185 (8.1%) 0.0019 *Based on indication of 2nd EGD