Optimizing the cost-effective evaluation of gastroesophageal reflux by typical symptom phenotypes after failure of empiric acid suppression trial.

Abstract

Introduction: Concern for gastroesophageal reflux disease (GERD) is the most common reason to consult gastroenterology. We aimed to optimize routine GERD evaluation on cost-effectiveness according to the dominant typical symptom among patients with persistent symptoms failing empiric proton pump inhibitors (PPI).

Methods: We developed a decision analytic model evaluating all permutations of GERD diagnostics including empiric trials of PPI optimization or discontinuation, upper endoscopy, wireless pH-monitoring, and pH-impedance monitoring. The model was applied to patients with heartburn, regurgitation, and chest pain in general gastroenterology to identify the appropriate combination and order of testing from insurer and patient perspectives. Health outcomes were informed by systematic reviews of clinical trials. Cost outcomes were informed by Centers for Medicare and Medicaid Services and commercial datasets and national observational studies. The time horizon was one year and willingness-to-pay threshold was $100,000/quality-adjusted-life-year (QALY) gained.

Results: For patients with typical persistent GERD symptoms failing empiric PPI, routine up-front ambulatory reflux testing saved $2,500-$4,500 compared to endoscopy alone when no erosive esophagitis is found. The most cost-effective initial ambulatory reflux test was 96-hour wireless pH-monitoring for patients with heartburn and chest pain and 24-hour pH-impedance monitoring for patients with regurgitation, both performed OFF-PPI. Adding ON-PPI pH-impedance monitoring optimized cost-effectiveness for patients with documented evidence of GERD and PPI-refractory symptoms. Patient and insurer perspectives aligned on these optimal diagnostic strategies.

Discussion: Compared to a one-size-fits-all strategy, a tailored approach based on Lyon 2.0 optimizes cost-effective evaluation and management of GERD by phenotyping the appropriate diagnostics to dominant symptom.