Seattle's Public Health Department has found evidence that the only rapid HIV test licensed to screen both saliva and blood (OraQuick) was not as accurate as the manufacturer claimed - it missed 8% of 133 people who were later found to have HIV. These results raise a question for all jurisdictions using oral testing: is reaching a wider range of people with testing worth it, if the results may be flawed?

Since the introduction of OraQuick in 2004, spikes in false positives have been reported at least twice, causing alarm in San Francisco in 2005 and halting use of the test in New York City. But in some ways the Seattle results are more troubling. False positives can be re-tested with rapid blood tests immediately, but false negatives represent missed opportunities to diagnose people who are infected before they become seriously ill.

Seattle is continuing to use OraSure when blood tests aren't possible, but other health departments and providers may not. One such provider that hasn't made the news is the Waianae Coast Comprehensive Health Center, on the leeward coast of Hawaii. The Waianae region has a high incidence of homeless people, many of whom camp on the beaches year-round; nationally, homelessness is linked to risk of HIV. When I interviewed a counselor there about the clinic's response to HIV, she told me that they had previously conducted outreach in schools, but had stopped that since they did not feel they had a reliable oral test any more.