• 0.01 preventable ADE per resident-month (Gurwitz et al., 2000).

  • 0.04 preventable ADE per resident-month (Gurwitz et al., 2005).

In the first study, carried out in 18 community-based nursing homes in Massachusetts, ADEs were identified by voluntary reporting and review of the record of each nursing home resident by two nurses and one pharmacist, performed every 6 weeks. In the second study, carried out in two large academic long-term care facilities, one in Connecticut and one in Ontario, Canada, ADEs were identified by a pharmacist’s monthly review of patient records. Medical records were also targeted for review using computer-generated signals (for example, abnormal serum levels), and administrative incident reports were reviewed as well for any indication of an ADE. This second study identified a much higher rate of ADEs than the first study. The authors suggested this difference could be attributed to the enhanced approach to identification of ADEs in the second study, although they thought the estimates from this study were still conservative since the study relied solely on information in medical records; there was no direct assessment of residents, which likely would have led to the identification of additional events.

The committee believes the second Gurwitz et al. study provides a better estimate of preventable ADE rates in the long-term care population. Applying the findings of this study to an average nursing home in the United States (bed size 105), 50 preventable ADEs (Gurwitz et al., 2005) would occur annually in the nursing home setting; applying the findings to the entire 1.6 million nursing home population in the United States, 800,000 (Gurwitz et al., 2005) preventable ADEs would occur each year in these settings. These figures are likely conservative, however, given the much higher ADE incident rates published in two other studies—0.44 ADEs per patient-month or 115 ADEs per 100 admissions (Gerety et al., 1993) and 134 ADEs per 100 admissions (Cooper, 1999). (Neither of these studies quoted the proportion of ADEs considered preventable.)

In one of the two nursing home studies by Gurwitz and colleagues (2000), of the 464 preventable ADEs and potential ADEs identified, 315 occurred in the ordering stage. Among those 315 errors, wrong dose (for example, excessive dose for an elderly patient) occurred in 63 percent of cases, followed by prescription of a drug for which there was a well-established interaction with another drug, which occurred in 22 percent of cases. The other Gurwitz et al. (2005) nursing home study found similar results. Among the 338 preventable ADEs identified, 198 occurred in the ordering stage. Of these prescribing errors, the most common were wrong dose (48 percent), wrong drug choice (38 percent), and known interaction (12 percent).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement