In Hospitalized HIV Patients, ID Consult Reduces Medication Errors

Jeffrey Hertzberg, MD, MS

In hospitalized HIV patients, medication errors are common and can be mitigated by enlisting ID specialists.

Elizabeth Neuner, PharmD, spends a lot of time reviewing hospital medication lists, and what she sees isn’t pretty, especially for HIV patients. In the 30 years since the AIDS epidemic was recognized, HIV treatments have become highly effective in reducing viral load and protecting the immune system. This means that most HIV care has moved from the hospital to the outpatient setting, where care is usually provided by physicians and multidisciplinary teams experienced with anti-retroviral management. This is good news for AIDS patients, but, as Dr Neuner underscored during her presentation on October 3, 2013 at IDWeek 2013, the shift has created a brand new set of challenges in the hospitalist-dominated inpatient environment. Most general hospitalists simply aren’t familiar with the complexity of anti-retroviral therapy and monitoring. And primary care physicians who still follow their own patients into the hospital face the same challenge. Dr. Neuner, infectious disease residency director at the Cleveland Clinic, observes frequent HIV medication errors in the hospital, which persist despite the recent advent of less complex regimens.

In a retrospective pre/post observational study, Neuner’s team implemented a targeted set of interventions for hospitalized patients on highly active antiretroviral therapy (HAART), excluding those treated with lamivudine or tenofovir for hepatitis B. The interventions included:
• Pharmacy education
• Modification of electronic medication files
• Collaboration with infectious disease (ID) specialists
• Daily medication profile review by an ID clinical pharmacy specialist

Records were reviewed for 162 admissions pre-intervention (pre), and 110 admissions post-intervention (post). The rate of medication errors per admission was significantly reduced-81/162, 50%, versus 37/110, 34% (p<.001). 124 errors occurred in the pre group (1.5 per patient), compared with 43 in the post group (mean 1.2 per patient). The most common error types were major drug interactions (26%) and dosing problems (20%) in pre group, and renal dosing problems (28%) and opportunistic infection-related problems (21%) in post group. A significantly higher error resolution rate was observed in the post group (74% vs. 36%, p<.001). The median time to resolution of errors between the groups was significantly different (180 hours pre; 95% CI: 73.5-258.6 vs 23 hours post; 95% CI: 5.7-39.3; p<.001).

After adjustment for potential confounders with logistic regression, the post group was independently associated with a lower medication error rate (OR 0.4; 95% CI 0.24-0.80; p=0.006). Presence of an ID consult was associated with a significantly higher rate of error resolution (68% vs 32%, p=0.002).

It’s pretty straightforward: if you’re a primary care physician or a general hospitalist, don’t try to manage HAART in the hospital by yourself. At the very least, patients on these medications need ID consultation, but the best centers of excellence will create multidisciplinary teams and clinical systems designed to prevent and correct medication errors for these very complex patients.

Sanders J, Palotta A, Bauer S, Sekeres J, Davis R, Taege AJ, Neuner EA. Antiretroviral medication errors in hospitalized patients with HIV. Paper presented at: IDWeek 2013; October 3, 2013; San Francisco, Calif. Abstract # 176. View on IDWeek 2013 Web site.