Objectives: Refractory gastroesophageal reflux disease (rGERD–defined as persistent GERD symptoms despite PPI therapy) constitutes ~30% of treated GERD patients. Consumer Price Index (CPI) is often used in health economics to adjust past cost data to present or future dollar values. This study evaluates CPI adjustment as a predictive tool for estimating present/future costs using data from a previous study of rGERD (Gerson et al. 2011 data collected 2004-2008) and a nearly identical current study (data collected 2011-2017). Methods: Average annual costs incurred by refractory and treatment-responsive GERD patients extracted from the previous and current studies were grouped into three categories: all-cause, GERD-related and GI-related. Costs from the previous study were inflated using CPIs of medical care and subcomponents (inpatient, outpatient [OP], emergency room [ER] and OP-pharmacy), using three adjustment models – end-to-end of the study periods (2008à2017), middle-to-middle of the study periods (2006à2014) and the ratio of average annual CPIs for the study periods. After CPI adjustments, model mean absolute deviations (MAD), standard absolute deviations (SAD) and root-mean-square deviations (RMSD) were calculated. Results: All three models exhibited similar accuracies. Middle-to-middle had the lowest MAD and SAD but highest RMSD for all costs, and the lowest MADs in OP, ER and OP-pharmacy costs. Middle-to-middle predicted all-cause OP and total direct medical costs with <7% and <11% deviations, respectively. It also predicted GERD-related rGERD total cost within <2% deviation. The model, however, underestimated ER costs by >20%. When determining GERD-related costs, middle-to-middle overestimated inpatient and OP-pharmacy costs by >110% and >200%, respectively and underestimated ER and OP costs by >30%. All GI-related costs were underestimated (>10%) but inpatient costs were overestimated (>30%). Conclusion: CPI-adjustment inflation models accurately predicted costs of large bundles of costs (e.g., total, outpatient) but less accurately when the costs of interest were defined more specifically (e.g., disease-related or line-items of healthcare).