Automated EHR Algorithm More Effective Than Nurse-Driven Protocol for HCV, HIV Screening

An algorithm that uses data from electronic health records to automatically add HIV and hepatitis C testing to the laboratory orders in an emergency department is more effective than nurse-driven procedures, according to recent research published in the Annals of Emergency Medicine.

The Centers for Disease Control and Prevention recommends universal screening for HIV in people aged 13 to 64 years old and universal hepatitis C screening for people born between 1945 and 1965. The CDC also endorsed emergency departments as catchalls for vulnerable populations who might not be seen in general medical settings. However, “few emergency departments have [these] routine screening programs in place.”

Study authors evaluated two strategies for HIV and hepatitis C screening used in an urban emergency department in Oakland, California: (1) nurse-ordered testing and (2) automated orders based on data from the electronic health record.

Using a retrospective before-after comparative effectiveness cohort study, investigators set out to determine which screening protocol more effectively tested and identified new cases of HIV and hepatitis C. In the nurse-driven approach, a triage nurse initiated each screening. In the automated approach, an algorithm “linked nontargeted HIV and hepatitis C virus screening to laboratory ordering.” Each strategy ran for 5 months.

A total of 20,975 people were eligible during the nurse-driven screening period and 19,887 were eligible during the automated period. When the triage nurse ordered screening, 19.6% (n=4121) people were screened for HIV and 14.2% (n=2968) were screened for hepatitis C. During the automated ordering protocol, 33.9% (n=6736) of people were screened for HIV and 35.1% (n=6972) were screened for hepatitis C.

Opt-out screening for HIV and hepatitis C using an algorithm and automated ordering proved to be “more effective and better integrated” than nurse-driven screening protocols (difference 20.9%; 95% CI 20.1% to 21.7%). The automated ordering protocol also delivered more new cases of HIV (23 vs 17) and hepatitis C (101 vs 29) than the nurse-ordered approach.

Limitations of the study included a lack of generalizability due to the retrospective nature of the study and its site-specific design. Additionally, the yield of hepatitis C screening was biased due to the availability of RNA testing offered in the automated-order protocol but not in the nurse-ordered segment.

Study authors note that “electronic health record-based automated ordering could increase the effectiveness of [emergency department] screening for HIV and hepatitis C.”

They also report that “with widespread use of electronic health record systems, this model can be easily replicated and should be considered the standard for future programs.”

This study was supported by a grant from the Frontlines of Communities in the United States program, Gilead Sciences. Please refer to reference for a complete list of authors’ disclosures.

Reference

White DAE, Todorovic T, Petti ML, Ellis KH, Anderson ES. A comparative effectiveness study of two nontargeted HIV and hepatitis C virus screening algorithms in an urban emergency department [published on June 21, 2018]. Annals of Emergency Medicine. doi: 10.1016/j.annemergmed.2018.05.005