A DEA methodology to evaluate multidimensional quality performance in hospitals

Abstract

Research into hospital quality performance typically considers a single dimension of quality at a time (e.g., West et al., 2002; McFadden et al., 2004). But as both hospitals and payers are aware, quality is multidimensional and needs to be measured more holistically to capture top performers. Data envelopment analysis (DEA) is a useful tool that typically looks at economic or cost data to determine the most efficient organisations in a group (with few exceptions). Using data from cardiology units in a sample of hospitals, this paper presents results from the use of DEA to study multiple quality metrics simultaneously in a geographically clustered group of hospitals to determine the best performers. This type of analysis might be useful for a hospital payer or a government agency that wants to reward hospitals for greater quality performance, but might otherwise be using a single dimension. Even those organisations that use multiple quality measures must face the problem of how to combine these different dimensions into one comprehensive quality measure. Our results highlight the usefulness of this technique in this situation and demonstrate how this technique can identify organisations that might otherwise be overlooked as high performers using traditional, single-dimension methods.