Background:  Current Centers for Disease Control and Prevention (CDC) guidelines for rapid HIV testing recommend that all reactive rapid tests be confirmed with laboratory-based supplemental testing. Despite the high specificity of individual rapid HIV tests, in low prevalence settings false positives may represent a large proportion of all reactive HIV screening tests. Also, depending on the setting, 20 to 50% of those with a reactive screening test do not return for laboratory-based confirmatory test results. Additional rapid tests conducted at the point of care (POC) may more quickly identify persons likely to be false positive on an initial rapid test and assist in the decision to immediately refer likely infected persons for further medical evaluation. However, this may require a change in the criteria for accessing HIV care.

Methods:  We assessed the practical and public health implications of adding additional rapid tests for clients with a preliminary positive rapid screening test at HIV counseling and testing sites and emergency departments in 4 US cities. Three different rapid test algorithms (RTA) were used at implementation sites; however, all required serologic evidence of HIV infection on 2 or more rapid tests to prompt referral to HIV care. To evaluate the performance of the RTA, plasma from clients with preliminary positive screening results was also sent for confirmatory testing.

Results:  Thus far, a total of 3043 individuals have been tested at 8 counseling and testing sites, and 8570 patients were screened in 2 emergency departments. Of those tested, 101 (0.9%) had an initial reactive result, 23 of whom had non-reactive results on subsequent rapid tests and were considered most likely uninfected. Specimens from the remaining 78 were reactive on all additional rapid tests. These 78 were considered likely to be infected and immediately referred to HIV care. All laboratory-based test results were concordant with the interpretation of the RTA.

Conclusions:  The successful performance of an RTA in a variety of study settings suggests that a serologic indication of HIV infection based on a series of reactive rapid HIV tests may be an adequate criterion for patient referral into HIV care in the US. Before final recommendations can be made, data from a larger number of individuals will be required, and the studies noted here will continue. However, HIV physicians should begin to engage in discussion of implications for this broader definition of ‘serodiagnosis’ on HIV referral and care.