Receipt of previous diagnoses and endoscopy and outcome from esophageal adenocarcinoma: a population-based study with temporal trends.

OBJECTIVES: Endoscopic screening of the at-risk population with chronic gastroesophageal reflux disease (GERD) for the presence of intestinal metaplasia or Barrett's esophagus has been suggested as a method to reduce mortality from esophageal adenocarcinoma. There are only limited data regarding the use, effectiveness, and temporal trends of this strategy.
METHODS: All patients aged 68 years and older with a new diagnosis of adenocarcinoma of the esophagus from 1994 to 2002 were identified from a linked tumor registry health claims database. Using claims from 1991 to 2002, the use of endoscopy as well as a diagnosis of GERD or Barrett's esophagus during the time interval from 3 years through 6 months prior to diagnosis were measured. The association of these measures with early-stage cancer (in situ, local) and long-term survival was determined, as well as changes over time.
RESULTS: We identified 2,754 patients, including 30.8% at early stage. Previous endoscopy had been performed in 11.5% of patients, and GERD and Barrett's esophagus diagnoses were recorded in 22.4 and 8.1% of patients, respectively. Barrett's esophagus diagnosis was strongly associated with both early-stage cancer (odds ratio 3.68, confidence interval (CI) 1.30-10.40) and survival (hazard ratio 0.45, CI 0.25-0.80). A GERD diagnosis was associated only with early stage, and endoscopy was associated only with survival. There was no association of year of diagnosis with stage or survival.
CONCLUSIONS: Although only a minority of at-risk patients were identified, the findings are consistent with the beneficial effects of recognition. Despite the development of practice guidelines, we were unable to demonstrate any temporal increases in diagnostic frequency or endoscopic utilization, which highlights the challenges that clinicians face.