Barrett's esophagus: incidence and prevalence estimates in a rural Mid-Western population.

ABSTRACT

BACKGROUND AND AIMS: Barrett's esophagus (BE) predisposes to adenocarcinoma of the esophagus and survival in esophageal adenocarcinoma is low. We studied patients diagnosed with BE in the Marshfield Epidemiologic Study Area (MESA). Our objectives were to estimate the prevalence of diagnosed BE, estimate the annual incidence of initial diagnosis of BE, and characterize the demographics of patients diagnosed with BE. METHODS: We retrospectively reviewed medical records of patients diagnosed with BE until December 31, 2002. The esophagogastroduodenoscopy (EGD) reports were reviewed to establish the presence of columnar epithelium. All slides were retrieved and reviewed by a gastrointestinal pathologist to establish the presence of intestinal metaplasia and dysplasia. Chart abstraction was conducted using a standardized form. RESULTS: BE was histologically confirmed in 216 patients. All were white, 165 (76%) were male, and 81% had a hiatal hernia. Median age at diagnosis was 65.5 yr (range 17-94). Long-segment BE (LSBE) was present in 112 (51.9%) patients. The prevalence of histologically confirmed BE in MESA was 261.8 (95% CI 222.5-301.1) per 100,000 people. The incidence of an initial diagnosis of BE between 1996 and 2002 was 32.7 per 100,000 person-years (95% CI 27.1-38.2) and did not change significantly over the study period despite an increase in EGD rates. At the initial diagnosis, 41.7% of the patients were on proton pump inhibitors. Dysplasia was present in 24.5% of patients. CONCLUSION: The incidence of initial diagnosis of BE in a stable white population did not change significantly over a 7-yr period, despite an increase in EGD rates.