Estimates are available for time from HIV infection to first detection with FDA-approved rapid tests using plasma seroconversion panels. However, most point-of-care (POC) HIV tests that are critical for HIV screening programs and entry to HIV and PrEP care are performed on unprocessed whole blood, fingerstick blood, or oral fluid. We assessed seroconversion sensitivity for tests when using these unprocessed specimen types.
Participants were at high risk for HIV infection and seeking HIV testing or were referred to the study after diagnosis with early infection. Participants were tested with a panel of POC tests on up to 3 specimen types (Table). Those with discordant results were tested repeatedly with the test panel through seroconversion. Through 8/25/2017, there were 1,211 participants, including 43 newly identified as HIV infected; this analysis is limited to the 12 HIV-infected participants with discordant results and/or documented dates of last negative HIV test. All 12 initiated treatment a median of 2 days after enrollment (range: -10 to 37 days). We estimated the infection date based on changes in viral load, reported symptoms of acute HIV infection, and HIV risk history, and describe the distribution of time from infection to first detection with each test and specimen type combination.
Estimated dates of infection ranged from 21 to 60 days before enrollment. Median time to first detection ranged between 33 and 43 days, and most tests were reactive in all participants by 90 days after estimated date of infection (Table). For 3 individuals, the time of first detection was delayed 1 study visit (3 to 7 days) for the same test performed on fingerstick compared to whole blood. For the 2 tests performed on oral fluid, there was a median delay of 2 and 4 days for oral fluid compared to whole blood. However, 3 participants who initiated treatment prior to peak viremia remained negative on ≥1 oral fluid test through day 90 of follow-up.
These are the first longitudinal data documenting seroconversion sensitivity of FDA-approved HIV tests performed on unprocessed specimens as intended for use in POC settings. These tests show a delay in earliest detection of 1 to 3 weeks when performed on these specimen types compared to previously published estimates using plasma. Because of the importance of POC tests, further improvements in seroconversion sensitivity and evaluations of new point-of-care tests using unprocessed specimens are warranted.