Rates of thromboembolic prophylaxis varied widely—from 5 to 81 percent. Rates of appropriate thromboembolic prophylaxis tended to be higher in surgical patients and in those at lower risk for thrombosis. One study also noted that thromboembolic prophylaxis was prescribed inappropriately in 38 percent of patients without risk factors for thrombosis (Aujesky et al., 2002).
Studies on the incidence of medication errors and ADEs in nursing homes use a number of different definitions, measures, and metrics. Hence, as with hospital studies, it is difficult to compare the results across studies.
Drug procurement and dispensing in the nursing home differ from hospital practice because the pharmacy is generally offsite. Handler and colleagues (2004) identified several aspects of drug delivery: (1) issues of packaging (e.g., patient-specific unit-dose packaging, patient-specific blister packages, 7-day strips of medication, color-coded drug administration devices, or medication bottles similar to usual community practice); (2) access to urgent medications, such as stock drugs in an emergency box; and (3) drug delivery when medications are added or changed, which may require hours to days (Handler et al., 2004). There is minimal research on how the approaches to addressing these issues affect medication safety.
When several pharmacies provide medications to a single nursing facility, staff must learn to use numerous systems, a practice that violates the fundamental safety principle of standardization. An evaluation of the medication-use system in one nursing home found that the facility’s 72 patients were served by seven pharmacies, and the consultant pharmacist had no relationship with any of them (Cooper, 1987). The charge nurse verifying refill needs required 8–12 hours per 100 beds per month. Qualitative data underscore the issues of time and error associated with this refill process (Vogelsmeier et al., 2005). Gupta and colleagues (1996a,b) noted that only 8.4 percent of the 19,932 Medicaid patients they studied used a single pharmacy, and the number of pharmacies used was associated with mortality rates (Gupta et al., 1996a,b).
The committee identified a few studies that measured the incidence of medication administration errors in nursing homes (see Table C-7). A well-known early study using direct observation of medication administration in