PRM54 A Conceptual Analysis of When Medical Expenditure Panel Survey (MEPS) Meets Cost of Illness (COI) – Are They a Match?

Abstract

MEPS has been used as a data source for many direct medical expenditures (DME)/COI type studies. The objective of this conceptual analysis is to review the literature & evaluate the 4 main COI type studies with reference to patients with head and neck cancer (HNC) using the MEPS data source. The 4 main COI methods are 1) Sum_All Medical; 2) Sum_Diagnosis Specific; and 3) Matched Control (iv) Regression. Without a comparison group, analysts use a total cost approach. With a comparison group, an incremental cost approach is most often reported. ISPOR does not have standard guidelines in reporting COI studies; analysts often vary in their perspective, but should always make reference to medical costs, morbidity & mortality costs, transportation/nonmedical costs and productivity losses in their analysis. A review of recent COI type papers (n=10) that have used MEPS shows that analysts report findings based on usually 1 COI method. Only 2 studies (COPD & Diabetes) reported both total and incremental cost approaches in their analysis. Four studies merged condition, event and consolidated year files, which allows diagnosis specific estimates to be produced. As a conceptual analysis, HNC is used to illustrate that MEPS facilitates analysts to use the 4 main COI methods especially if DME is the outcome of interest. MEPS is a valuable and utilized national resource. It is possible that “good practice guidelines” can be developed (& perhaps endorsed by AHRQ) for those using MEPS to report a DME/COI type study. By using & reporting all 4 methods, an analyst is giving policy-makers a range for their cost estimates. Guidelines would ensure a level of transparency in reporting such cost estimates across conditions.