A nontargeted, opt-out HIV screening program in a Detroit emergency department identified 176 patients with previously undiagnosed HIV from November 2019 through October 2025, with about two-thirds of diagnoses resulting from testing prompted solely by an automated electronic advisory.
The retrospective observational study, published in Open Forum Infectious Diseases, evaluated a best practice advisory-based HIV screening program implemented at Henry Ford Hospital in 2019. The program integrated emergency department and infectious diseases teams to facilitate screening, communication of positive results, linkage to care and treatment initiation. Investigators reviewed medical records from November 2019 through October 2025 for patients with newly diagnosed HIV identified during an ED visit.
The BPA prompted fourth-generation HIV testing for patients aged 18 to 65 years who had a complete blood count performed in the ED, no documented HIV diagnosis and no previous HIV test in the electronic medical record. Nurses informed patients that HIV testing had been ordered and provided an opportunity to decline. Reactive HIV screening results automatically underwent confirmatory testing with an HIV differentiation assay.
For confirmed positive HIV results, an ID nurse contacted the patient, discussed the diagnosis, provided counseling, scheduled an ID appointment and involved social work when needed to address any potential barriers to care. Patients who could not be contacted were referred to Michigan Department of Health and Human Services disease intervention specialists for further outreach.
Primary outcomes included linkage to care, time from diagnosis to the first ID appointment and antiretroviral therapy initiation, baseline viral load and CD4 counts and viral suppression at 3, 6, and 12 months. Viral suppression was defined as an HIV-1 RNA level below 30 copies/mL. Linkage to care was defined as attendance at the first ID appointment after diagnosis.
Screening identified 176 new HIV diagnoses
Among 75,914 patients screened during the study period, 176 received a new HIV diagnosis, corresponding to an overall test positivity rate of 0.23%. The highest HIV positivity rate occurred in 2021, at 0.36%, but positivity rates did not significantly differ between any year relative to the first full year of BPA system implementation in 2020. The BPA-based HIV testing uptake rate by physicians and health care providers in the ED was 72% over the course of the study.
The median age of patients with newly diagnosed HIV was 33 years. Of the 176 patients, 83% were Black and 73% were male. Reported factors included being a man who has sex with men in 35%, heterosexual contact with multiple partners in 34%, injection drug use in 9% and being unhoused in 16%. Eighteen percent had a history of a sexually transmitted infection.
Overall, 117 patients, or 67%, were diagnosed through testing prompted by the BPA system alone. Other reasons for HIV testing included high-risk behavior in 15%, concern for an STI in 9% and an opportunistic infection in 6%. Notably, 74% had a previous ED visit documented in the health system before their HIV diagnosis. At the diagnostic encounter, 39% were hospitalized and 35% had a concomitant STI diagnosis, including syphilis in 21%.
Nearly 60% linked to care within 90 days
After diagnosis, 82 patients, or 47%, attended their initially scheduled ID clinic appointment, which was scheduled within 7 days of diagnosis or hospital discharge. Overall, 104 of 176 patients, or 59%, attended an ID clinic appointment within 90 days of diagnosis. Among those patients, the median time from diagnosis to the first appointment was 13 days.
Among 107 patients who initiated ART, the median time to ART initiation from diagnosis was 9 days. Slightly more patients initiated ART than attended a postdiagnostic appointment, as some patients started ART during hospitalization.
Among 123 patients with available baseline viral load testing, the median baseline viral load was 95,953 copies/mL. Among 119 patients with available immunologic testing, the median CD4 count was 306 cells/mm³.
Among patients with available follow-up data, 51 of 77, or 66%, had documented viral suppression at 3 months; 45 of 70, or 64%, at 6 months; and 54 of 74, or 73%, at 12 months.
The investigators noted that the overall 0.23% new HIV test-positivity rate exceeded the 0.1% threshold recommended by the CDC for routine opt-out HIV screening to support cost-effectiveness. They attributed the 72% acceptance of BPA-prompted testing in part to collaboration between ED and ID teams, with the ID team assuming responsibility for communicating confirmed positive HIV results and postdischarge coordination.
Study limitations
The authors noted that the study population may not reflect the broader demographics of the HIV epidemic. The proportion of patients who were men who have sex with men was lower than the proportion of men who have sex with men accounting for incident HIV in Michigan.
Investigators also could not reliably determine why physicians opted out of BPA testing and lacked data on patients who declined HIV screening.
Follow-up data were limited to care received within Henry Ford Health. Seventy-two patients lacked information on clinic attendance, preventing investigators from determining whether they received care elsewhere or did not pursue additional care. Patients who died during the study period were included but did not have follow-up data. ART initiation also was delayed in some patients hospitalized with opportunistic infections, such as cryptococcal meningitis, which increased the median time to ART initiation.
The study period overlapped with the COVID-19 pandemic, which may have negatively affected linkage-to-care efforts because of staffing challenges. In addition, the investigators lacked preimplementation data and could not formally assess the association between ED-ID collaboration and linkage to care.
Clinical implications
“The described automated BPA-based, nontargeted, opt-out HIV screening program in the ED may be a feasible and valuable model for other health care systems to utilize to aid in identification of HIV infection,” said lead study author Brianna Hohmann, DO, in an interview with this news organization.
Close collaboration between ED and ID teams can be incorporated in other system workflows to facilitate rapid and successful linkage to care and ART initiation, particularly among populations facing barriers to health care access, noted Dr. Hohmann, an infectious diseases physician at Henry Ford Hospital and clinical assistant professor at Michigan State University College of Human Medicine and Wayne State University School of Medicine.
Next steps
According to Dr. Hohmann, next steps would be to expand to other health systems to identify new HIV infections and successfully link those newly diagnosed people to care. “Although our study was able to remove many structural barriers to improve health care access, more strategies are needed to facilitate linkage to care and ART initiation in disadvantaged populations,” she concluded.
Study author Indira Brar, MD, disclosed serving on the ViiV Speakers Bureau Advisory Board and Gilead Speakers Bureau Advisory Board. Jacob Manteuffel, MD, receives salary support from the Gilead FOCUS grant, which supports HIV screening and linkage to care. The other study authors had no relevant conflicts of interest.
