Return toAll Abstracts
HSR&D 2004 National Meeting Abstracts


2044. Three Rules of Thumb When Using VA Inpatient Cost Data
Todd H Wagner, PhD, VA Palo Alto Heathcare System, J Velez, VA Palo Alto Heathcare System, W Yu, VA Palo Alto Heathcare System, CS Phibb, VA Palo Alto Heathcare System

Objectives: Until recently, researchers had no inpatient cost data. Now cost data from the Decision Support System (DSS) and the Health Economics Resource Center (HERC) exist, creating questions about (1) which dataset should be used, and (2) whether the datasets can be used together to identify problems or outliers.

Methods: We compared inpatient costs from the DSS National Data Extracts and HERC average costs data from FY 2001 (n=617,503). We used multivariate regression techniques to explain discrepancies between the two datasets. We created “rules of thumb” to guide researchers wanting to conduct economic analyses with VA data.

Results: We developed three rules of thumb. First, HERC and DSS rely on different cost databases, and with the exception of pharmacy costs, researchers should not mix these databases. Second, for medical/surgical care, which accounts for 56% of VA inpatient costs, HERC and DSS use different costing methods. This is particularly important for patients who have very long lengths of stay (LOS> 35 days) and/or are in a diagnostic related group (DRG) that is associated with a large relative value weight (DRG weight >5). Third, for non-medical/surgical inpatient care (rehabilitation, mental and long term care), both datasets predominantly use daily rates to estimate patient costs. Researchers should calculate the daily rate to identify outliers.

Conclusions: Guidance is available for researchers who are conducting economic analysis with VA data.

Impact: This research provides insights on how researchers can monitor the accuracy of VA cost data.