or 70 per 1,000 patient-days. Medication errors contributed to 27 percent of the ADEs. Of all the ADEs, 9 percent resulted in serious harm, 22 percent in additional monitoring and interventions, 32 percent in interventions alone, and 11 percent in monitoring alone; 27 percent should have resulted in additional interventions or monitoring.
Three smaller studies found similar ADE rates. A 37-day study at a Boston tertiary hospital found 27 ADEs (15 considered preventable), for a rate of 6.4 ADEs per 100 admissions or 9.1 ADEs per 1,000 patient-days (Bates et al., 1993). Another small study at the same hospital found 25 ADEs (5 considered preventable), for a rate of 6.6 ADEs per 100 admissions or 14.7 ADEs per 1,000 patient-days (Bates et al., 1995a). In a study of 157 hospitalized patients aged 70 and older, 28 probable ADEs were observed, for a rate of 17.8 ADEs per 100 admissions (Gray et al., 1998). Just over half the ADEs were considered preventable.
Errors of omission occur when a medication necessary for the appropriate care of hospitalized individuals is not prescribed. After reviewing the published literature on medication errors of omission within acute care, the committee identified three broad categories of studies: studies on treatment of acute coronary syndromes, on antibiotic prophylaxis, and on thrombosis prophylaxis (see Table C-6).
TABLE C-6 Hospital Care: Prescription and Selection Errors of Omission
Patients discharged with diagnosis of acute myocardial infarction
Percentage of patients given aspirin within 24 hours of hospitalization
84.9 (Roe et al., 2005) (NSTEMI)
88 (Roe et al., 2005) (STEMI)
92.4 (Granger et al., 2005)
93 (Sanborn et al., 2004)
Percentage of patients prescribed aspirin at discharge
53 (Krumholz et al., 2003)
76.8 (Petersen et al., 2001)
83.8 (Roe et al., 2005) (NSTEMI)
84.8 (Petersen et al., 2003)
85.6 (Alexander et al., 1998)
88.9 (Roe et al., 2005) (STEMI)
93.4 (Granger et al., 2005)