Link to article at PubMed
Frequency of HIV-related medication errors and associated risk factors in hospitalized patients.
Ann Pharmacother. 2008 Apr;42(4):491-7
Authors: Pastakia SD, Corbett AH, Raasch RH, Napravnik S, Correll TA
BACKGROUND: Retrospective studies of hospitalized HIV-infected patients have noted a high occurrence of drug-related errors, ranging from 5% to 30%. OBJECTIVE: To prospectively evaluate errors in antiretroviral (ARV) prescribing in the inpatient setting of a hospital tertiary care center and the association of risk factors with the occurrence of errors. METHODS: HIV-infected patients who received care and continued their ARVs for HIV infection on admission to a large academic teaching hospital between January and April 2006 were included in this study. The care and assessment of these patients was conducted on a daily basis by an infectious diseases/HIV specialized clinical pharmacist. All errors were documented and classified based on a severity scale. RESULTS: Among the 68 patients who met the study's eligibility criteria, at least one error in the initial HIV regimen occurred in 72% of patients, and in 56% of patients, the error had the potential to cause moderate-to-severe discomfort or clinical deterioration. Patients on atazanavir-based therapy had a statistically significant increased occurrence of errors throughout their hospitalization (RR = 1.69; 95% CI 1.03 to 2.78; p = 0.02). Receiving nonformulary (combination) HIV medications increased patients' risk of having more than one error occur in their ARV regimen on admission and during hospitalization (RR = 1.95; 95% CI 1.25 to 3.04; p = 0.02). The clinical pharmacist recommendations had 100% acceptance. CONCLUSIONS: The alarmingly high frequency of potentially harmful errors uncovered in this study necessitates further investigation using larger sample sizes. Interventions to reduce and prevent these errors must be sought to eliminate the unintended harm associated with hospitalization.
PMID: 18349307 [PubMed - indexed for MEDLINE]