OBJECTIVE
Barrett's esophagus (columnar metaplasia of the distal esophagus due to chronic gastroesophageal reflux) affects nearly 700,000 people in the United States, and carries a risk of esophageal adenocarcinoma that is 30-125 times that of an age-matched population. Patients who develop high grade dysplasia are at greatest risk. Current recommendations are for endoscopic surveillance to detect dysplasia and to diagnose carcinoma while it is in an early and possibly treatable stage. In addition, some authorities recommend esophagectomy for high grade dysplasia, whereas others reserve esophagectomy only for those with cancer. There are no controlled trials demonstrating that surveillance increases life expectancy in patients with Barrett's esophagus. Furthermore, endoscopic surveillance of this large group with Barrett's esophagus may be costly, and associated with considerable morbidity. Therefore, our objective was to assess the effectiveness and cost-effectiveness of endoscopic surveillance in patients with Barrett's esophagus.
METHODS
Design--Decision analysis using a computer cohort simulation (Markov). We examined 12 strategies: (A) no endoscopic surveillance. Esophagectomy is performed only if cancer is detected by biopsy. (B) no surveillance. Esophagectomy is performed if high grade dysplasia is detected by biopsy: (C1-C5) surveillance at intervals from 1 to 5 yr, with esophagectomy if cancer is diagnosed, and (D1-D5) surveillance at intervals from 1 to 5 yr with esophagectomy if high grade dysplasia is diagnosed. We measured life expectancy, quality-adjusted life expectancy, and incremental cost-effectiveness ratios for each strategy. Data Sources--Medline Search and bibliographies of retrieved articles; expert opinion when published data were not available.
RESULTS AND CONCLUSIONS
Annual surveillance with esophagectomy for high grade dysplasia prevents cancer and is the preferred strategy, if only length of life (life expectancy) is considered. For those who consider both length and quality of life, endoscopy every 2-3 yr will provide the greatest quality-adjusted life expectancy. When costs are considered, endoscopy every 5 yr also increases life expectancy and has an incremental cost-effectiveness ratio similar to common medical practices. The cumulative incidence of cancer and the quality of life with an esophagectomy had the greatest impact on the decision for surveillance and the optimal surveillance strategy.