Barrettʼs Surveillance Interval Repeat Ratio: A Novel Utilization Metric

Abstract

Purpose: Surveillance endoscopy (EGD) programs in patients with Barrett's Esophagus (BE) are recommended to detect development of dysplasia or early esophageal adenocarcinoma (EAC). Current guidelines recommend repeat EGD within one year for patients without dysplasia and interval EGD every 3 years subsequently; and for patients with low grade dysplasia repeat EGD within 6 months and annually subsequently. The number of patients diagnosed with BE has increased in the recent decades and this increase has significant implications for health care resource utilization and costs. Currently there are no objective criteria which can be used as a utilization measure. We proposed to devise a novel metric of BE Interval Repeat Ratio (BEIRR) to optimize utilization of surveillance EGD in BE. Methods: Retrospective review of medical records for a cohort of patients (≥ 18 years) who had EGD which detected mucosal changes suspicious for BE between January 2004 to December 2006 done by 6 well experienced gastroenterologists at the University setting. These patients were followed till May 2010. BEIRR was calculated as Total number of repeat EGD's + Total number of Initial EGD's/Total number of Initial EGD's for 6 providers separately and statistical analysis was performed. Patients diagnosed to have EAC and patients whose pathology reports could not be obtained were excluded from the final analysis. Results: 151 patients were identified with mucosal changes suspicious for BE (111 Male; Mean Age 61.1 years) and 13 were excluded. 82 patients had BE without dysplasia, 4 patients had BE with low grade dysplasia and 52 patients did not have microscopic changes of BE. 136 EGD's were repeated over 6 years for patients with BE without dysplasia with a cumulative BEIRR of 2.8. Individual BEIRR for 6 providers were 3, 2.26, 3.5, 2.1, 2.1 and 2.3, respectively, with a mean of 2.54, SD 0.57; 95% CI 1.4-3.68. For 4 patients with low grade dysplasia 24 EGD's were repeated with cumulative BEIRR of 7. For patients without BE on microscopy 54 EGD's were repeated of which 33 (61%) were repeated for surveillance and none of the subsequent EGD's showed any evidence of metaplasia, dysplasia or cancer. One patient with LGD progressed to have high grade dysplasia and one patient progressed to have adenocarcinoma. Conclusion: BEIRR's can be used as a benchmark metric to optimize utilization of surveillance EGD's for BE. We found a very low incidence of high grade dysplasia and adenocarcinoma in our cohort of patients consistent with prior reported data. Surveillance EGD for patients with no microscopic evidence of BE did not detect even a single case of metaplasia, dysplasia or cancer.