Another study, in a tertiary hospital in Florida, involving about 6,000 patients (the authors could not report precisely the number of patients involved), found an administration error rate of 5.8 per 1,000 admissions (Winterstein et al., 2004).
Similar rates to those above have been observed in intensive care unit (ICU) studies. In a study focusing on high-alert medications administered in ICUs in five U.S. tertiary care teaching hospitals, an administration error rate of 3.3 percent was found (5,744 observations) (Calabrese et al., 2001). In another study, carried out in a medical ICU in a French hospital, an administration error rate of 6.6 percent was observed (2,009 medication administration interventions by nurses) (Tissot et al., 1999).
Higher rates were seen in studies that focused exclusively on IV medications—34 percent (93 errors out of 278 observed administrations) (Wirtz et al., 2003) and 49 percent (212 preparation and administration errors out of 430 doses) (Taxis and Barber, 2003).
Two studies looking at preventable ADEs occurring during the administration stage found rates of 2.1 per 1,000 admissions (in a study of 4,031 patients at two tertiary hospitals in Boston, Massachusetts [Bates et al., 1995b]) and 17.9 per 1,000 admissions (in a study of 937 admissions at a tertiary hospital in Salt Lake City, Utah [Bates et al., 1995b; Nebeker et al., 2005]).
Rates of preventable ADEs resulting from errors in the monitoring of patients were reported in two studies as 0.6 per 1,000 admissions (Hardmeier et al., 2004) and 32 per 1,000 admissions (Hardmeier et al., 2004; Nebeker et al., 2005). (See Table C-4).
Five major studies examined the incidence of ADEs occurring during hospitalization (see Table C-5). Using hospital admissions during the period 1990–1993, investigators at LDS Hospital, Salt Lake City, Utah, found that 2,227 out of 91,574 patients experienced ADEs during hospitalization, a rate of 2.43 ADEs per 100 admissions (Classen et al., 1997). Almost 50 percent of the identified ADEs were thought to be preventable. Extrapolat-
TABLE C-4 Hospital Care: Monitoring Errors
Per 1,000 admissions—detection method
0.6 (Hardmeier et al., 2004) (Swiss study)—chart review
32 (Nebeker et al., 2005)—review of electronic medical record