Which HIV Tests Are the Most Accurate?

Study Quantifies Real-World Accuracy of Approved HIV Testing Assays

A positive reading (two bars) on an OraQuick ADVANCE in-home HIV test. Photo courtesy Orasure Technologies

As the U.S. aims to increase early identification and treatment of people with HIV, a greater focus has been placed on determining the accuracy of HIV tests in real-world settings—not only to minimize the number of false positive and false negative HIV test results but to better identify individuals during the early (acute) stages of infection when transmission risk is especially high.

In order to quantify this, researchers from the University of California, San Francisco conducted a review of over 21,000 HIV tests given between the years 2003 and 2008 in some of the city's highest prevalence populations. Of four types of tests used during this period—from first generation antibody tests to recently licensed, rapid oral tests—761 people were diagnosed with HIV (3.6% prevalence), while 58 were identified during acute infection.

The study also aimed to compare the accuracy of newer testing assays—including 4th generation combination antigen/antibody tests—by retesting the blood from the 58 individuals previously diagnosed with acute infection.

Accuracy was measured both in terms of sensitivity (the percentage of test results that are correctly positive) and specificity (the percentage of test results that are correctly negative).

Test TypeBrandSensitivity from 21,234 testsSpecificity from 21,234 testsSensitivity for acute infection from 58 tests
1st generation antibody test (blood)Vironostika HIV-1 Microelisa92.3%100%0%
3rd generation antibody test (blood)Genetic Systems HIV-1/296.2%100%34.5%
3rd generation rapid antibody test (blood)OraQuick Advance91.9%100%5.2%
3rd generation rapid antibody test (saliva)OraQuick Advance86.6%99.9%--
3rd generation rapid antibody test (blood)Uni-gold Recombigen----25.9%
3rd generation rapid antibody test (blood)Multispot HIV 1/2----19.0%
3rd generation rapid antibody test (blood)Clearview Stat Pak----5.2%
4th generation rapid combination antigen/antibody test, (blood)Determine HIV 1/2 Ag/Ab Combo----54.4%
4th generation lab-based combination antigen/antibody test (lab)ARCHITECT HIV Ag/Ab Combo----87.3%

What Does All of This Tell Us?

Firstly, from the point of view of specificity, the figures confirmed that the incidence of false positives remains extremely low, even with the earlier generation tests.

By contract, the rate of false negatives varied considerably, with the OraQuick Advance rapid saliva test performing the worst, with just over one in 15 patients receiving a false negative result.

The figure only worsened when retesting bloods from the acute stage infections. Of the 58 samples tested, the 3rd generation rapid tests achieved a sensitivity of only 5.2% to 25.9%, meaning that the majority of such infections would be missed using these rapid, antibody-based technologies.

Even the 4th generation Determine rapid antigen/antibody test was able to identify only half of the acute infections, despite having an estimated sensitivity of 96.6% and specificity of 100%. According to the UCSF researchers, the Determine worked best during acute infection when the patient's viral load was over 500,000 copies/mL.

Not surprisingly, the lab-based ARCHITECT combination antigen/antibody test performed the best of possible assays. With an estimated specificity of between 99.1% and a specificity of 100%, the tests were able to identify nearly 90% of acute infections.

What Does This Mean to Me?

In terms of testing selection and performance, the following conclusions can be reasonably drawn:

  • Commercial, over-the-counter HIV tests perform least well overall, not only in terms of detecting acute infection but in returning a false negative rate of 7%.  
  • Combination HIV antigen/antibody tests are far more accurate that traditional antibody-based assays, particularly during the acute stage of infection.
  • Lab-based tests still outperform at-site, point-of-care HIV tests, most especially in cases of recent HIV exposure.

With that being said, high levels of sensitivity are only part of the reason why certain tests are preferred over others. 

For example, a significant number of people fail to return for HIV results after testing. The ability, therefore, to return test results within 20-30 minutes make rapid testing ideal, particularly if it allows clinics to link a person immediately to care. 

Similarly, people with confidentiality concerns or fears about HIV stigma may be better served by taking an in-home rapid test (pictured).

 While there little data to assess the actual number of people linked to care after receiving a positive, in-home result, it is presumed that such tests provide an entry point to many who might otherwise avoid testing centers or clinic settings.

Was this page helpful?
View Article Sources