Are Medication Error Rates Useful as Comparative Measures of Organizational Performance?

Abstract

Although medication error rates seem a plausible indicator, it is not a foregone conclusion that error rates are good barometers of quality. This article is based on a multisite, nationwide study of one contributor to medication errors: pharmacy dispensing errors. Using questionnaires, directors of pharmacy at 157 member hospitals of a national health care management firm reported their average daily medication dispensing loads and the number of dispensing errors per 100 patient days during one quarter. Stepwise discriminant analysis was used to seek characteristics of high versus low reported error rates, as well as characteristics that could distinguish hospitals that reported no errors from those that reported errors. Results show that currently measured error rates represent a process within an organization and can range from no or few errors to substantial errors, with the variation reflecting a variety of scenarios. Hospitals that do not measure error rates obviously report no errors. Inattention or a lack of clear methods to collect and define errors can also result in no reported errors or low error rates. In contrast, a hospital that standardizes and implements a measurement scheme may have high reported error rates. Progressive experimentation with systemwide error prevention and process improvement will result in varying error rates. Logically, if hospitals are punished for reporting high error rates, they will start reporting lower error rates regardless of the true rates. Finally, successful management practices will be reflected in low error rates. The focus on measuring medication error rates is important for improving quality within organizations because drug-related errors are an important cause of adverse events. However, the variances in error-reporting rates and the variables associated with those variances documented in this study raise serious questions about the usefulness of comparing error rates between hospitals based on voluntary reports. Interorganizational comparisons of rates are not likely to be meaningful and may be counterproductive.