Characterization of medical errors in initiation of postexposure prophylaxis in sexual assault victims in the emergency department

Study objectives: We characterize common medical errors when postexposure prophylaxis (PEP) is initiated to prevent HIV transmission after sexual assault. Methods: This was a retrospective review of cases of sexual assault presenting to a university urban emergency department (ED) from November 2002 to November 2003. The ED has existing guidelines specifying the indications for PEP, as well as the criteria for high-risk cases in which a protease inhibitor should be added to the standard regimen of Combivir for 30 days. The guidelines also state that PEP should be initiated only within 72 hours. Using the guidelines as a standard, 2 trained abstractors reviewed each case and determined whether PEP was indicated and the correct drug regimen. Type and time of exposure, medications given, laboratory studies, and follow-up arrangements were recorded. A κ coefficient was calculated. Results: Sixty-eight cases were reviewed. Mean age of the patients was 24.2 years. Fifty-seven patients presented fewer than 72 hours after the assault; 11 patients presented outside the treatment window for PEP. According to type and time of exposure, PEP was indicated in 51 (75%) of 68 cases. PEP was offered 41 of 51 times it was indicated (80%). Four patients received PEP outside of the 72-hour window. Seven patients refused PEP. Of the remaining 38 patients receiving PEP, the correct drug regimen was given 29 times (76%), and the incorrect regimen was prescribed 9 times (24%). One patient did not have blood drawn for baseline laboratory test levels. The assessment of the HIV status of the assailant was not documented in any case. Of the patients offered PEP, only 24 (53%) of 45 were given documented follow-up instructions. The κ coefficient was 1. Conclusion: Despite department guidelines, errors in initiating PEP after sexual assault were common. Because effective PEP involves a strict adherence to a regimen and close follow-up, we are now studying the effectiveness of a real-time computerized prompt referring the physician to the PEP recommendations.