An estimated 1 in 5 HIV‐positive Americans is unaware of their HIV status. Experts warn that this subgroup notably contributes in propagating the U.S. HIV epidemic. Universal opt‐out screening may mitigate this problem but remains haphazardly implemented. Inpatient‐based programs are a potential pathway toward efficient HIV screening. However, supportive studies are sparse and focus mostly on medicine patients. Data evaluating screening in surgical inpatients are lacking.


The aim of this study was to evaluate HIV screening practices in “medical” versus “surgical” patients admitted to an academic urban hospital during a pilot of universal opt‐out HIV screening and linkage to outpatient care. This was a prospective cohort study design. For the intervention, from October 2011 through Oct 2012, attendings and house officers were asked to perform opt‐out HIV screening with all admissions and inpatient consultations. Periodic education, reminders, and feedback were used to encourage compliance. Anticipated barriers to screening were addressed, including removal of separate written HIV testing consent. All adult inpatients were eligible for screening. Obstetrics‐gynecology patients were prospectively excluded from analysis. Patients were categorized as medicine or surgery based on the admitting attendings' specialty. Screening rate assessed the proportion of patients admitted who were tested for HIV. Positive rate assessed the proportion of patients tested who were confirmed as HIV(+). Rates were prospectively collected and reviewed monthly.


A total of 3831 medicine and 4489 surgery patients were admitted. On average, the monthly screening rate was higher in medicine patients compared with surgery patients (38.0% vs. 13.5%; P = 0.001). The mean monthly positive rate was not statistically different between the 2 groups (medicine, 2.49%, surgery, 2.66%; P = 0.88). All patients identified as HIV(+) were successfully linked to outpatient HIV follow‐up. In addition, qualitative assessments found screening rates were notably diminished by providers' preconceptions of patients as low risk, fears of harming doctor–patient rapport, misinformation about current HIV screening guidelines, and concerns about medicolegal responsibility for HIV follow‐up care.


Inpatient opt‐out HIV screening appears to be a viable tool for identifying HIV(+) patients and linking them to appropriate care. However, the efficacy of the intervention was limited by low screening rates. Moreover, despite similar positive rates in both groups, surgical inpatients were significantly less likely to undergo screening. Approaches to systematically engage all providers are needed. Strategies to specifically enhance HIV screening among nonmedicine inpatients such as hospitalist‐led screening programs for surgical comanagement services and interventions to target providers' biases may meaningfully improve inpatient‐based HIV patient detection and linkage to care.