Objective: Cost-effective strategies for identifying patients with Barrett's esophagus who are most likely to develop cancer have not been developed. Surveillance endoscopy is currently used, and we hypothesized that more frequent surveillance intervals would identify patients with "transient positive" diagnoses of dysplasia--dysplasia found on one examination but not on subsequent ones. Our aim was to explore the potential economic impact of transient positive diagnoses of dysplasia on alternative surveillance strategies over a 10-yr period.
Methods: Data were derived from a 2-yr randomized, prospective study comparing omeprazole to ranitidine in 95 patients with Barrett's esophagus. A transient positive diagnosis of dysplasia was defined as a patient who was diagnosed with dysplasia during the study period but whose 24-month biopsies revealed no dysplasia. We calculated the number of transient positive diagnoses of dysplasia and modeled the potential economic impact of a diagnosis of dysplasia over a 10-yr period.
Results: Thirty patients (31%) had at least one reading of dysplasia during the study period. Nineteen patients (20%) had a transient positive diagnosis of dysplasia. During the study period, no cancers were found. A surveillance strategy of every other year and every 6 months for dysplasia would result in 1072 endoscopies over a 10-yr period at a discounted cost of $1,587,184. A total of 61% of endoscopies would be because of transient positive diagnoses of dysplasia. A strategy of yearly surveillance and every 6 months for dysplasia would result in 1404 endoscopies at a discounted cost of $2,096,733, of which 28% would result from transient positive diagnoses of dysplasia. The discounted incremental costs of more frequent surveillance in this cohort of patients over 10 yr is $509,549.
Conclusions: Based on current practice strategies, transient positive diagnoses of dysplasia account for 28-61% of endoscopies in Barrett's surveillance programs. This analysis suggests that the endoscopy workload and costs associated with surveillance could be substantially reduced if patients with transient positive diagnoses of dysplasia reverted to usual surveillance after two negative examinations.