Improving cost-effectiveness of endoscopic surveillance for Barrett's esophagus by reducing low-value care: a review of economic evaluations.

Background: Individuals with Barrett's esophagus are believed to be at 30-120× risk of developing esophageal adenocarcinoma (EAC). Early detection and endoscopic treatment of dysplasia/early cancer confers a significant advantage to patients under surveillance; however, most do not progress past the non-dysplastic state of Barrett's esophagus (NDBE), which is potentially an inefficient distribution of health care resources.

Objectives: This article aimed to review the outcomes of cost-effectiveness studies reducing low-value care in the context of endoscopic surveillance for non-dysplastic Barrett's esophagus (NDBE).

Methods: A systematic search was conducted by two reviewers in accordance with PRISMA guidelines.

Inclusion Criteria: cost-utility analyses of endoscopic surveillance of NDBE patients with at least one treatment strategy focused on reduction of surveillance. A narrative synthesis of economic evaluations was undertaken, along with an in-depth analysis of input parameters contributing to stated Incremental cost-effectiveness ratios (ICER). Study appraisal was performed using the consolidated health economic evaluation reporting standards (CHEERS) tool.

Results: 10 Studies met inclusion criteria. There was significant variation in cost-model structures, input parameters, ICER values, and willingness-to-pay thresholds between studies. All studies except one concluded guideline-specified endoscopic surveillance for NDBE patients was not cost-effective. Studies that explored a modified surveillance by deselection of low-risk NDBE patients found it to be a cost-effective strategy.

Conclusion: Guideline specified endoscopic surveillance for NDBE was not found to be cost-effective in the studies examined. A modified endoscopic surveillance strategy removing individuals with the lowest risk for progression from NDBE to adenocarcinoma is likely to be cost-effective but is dependent on risk profile of patients excluded from surveillance.