www.hivandhepatitis.com

Oral HIV Tests May Be Less Accurate than Blood Tests

Oral HIV antibody testing is quick and convenient, but this method is not quite as accurate as blood testing in low-prevalence settings, being more likely to produce false-positive results, according to a study review and meta-analysis described in the January 24, 2012, online edition of The Lancet Infectious Diseases.

As evidence accumulates to support early antiretroviral therapy (ART) and "treatment as prevention," there is growing emphasis on expanded HIV testing so that more people can learn their status and start treatment in a timely manner. Development of non-invasive rapid oral tests has enabled screening to be performed outside traditional healthcare settings.

Nitika Pant Pai from McGill University in Montreal and colleagues sought to compare the diagnostic accuracy of a rapid point-of-care HIV screening assay -- OraSure Technologies' OraQuick ADVANCE Rapid HIV-1/2 Antibody Test -- when used with oral fluid versus blood specimens. The oral test uses a swab to collect a sample of fluid (or "trasudate") from around the gums. The blood test can be done using either a finger-stick sample or blood drawn from a vein.

[CDC fact sheet about the OraQuick oral fluid test: http://www.cdc.gov/hiv/topics/testing/resources/qa/oralfluidqandafin.htm]

The researchers performed a systematic review and meta-analysis, searching 5 databases of published medical literature and databases of 5 key HIV/AIDS conferences. They selected eligible studies that focused on adults at risk for HIV infection; studies of children and coinfected populations, and those with incomplete reporting of key data, were excluded.

In all, the authors analyzed 45 studies, including 24 looking at comparative accuracy and 21 analyzing positive predictive value or PPV -- the proportion of people testing HIV positive who are correctly diagnosed.

Results

• In a direct head-to-head comparison of studies, the pooled sensitivity of the oral test -- the ability to accurately determine if an infected person is HIV positive -- was about 2% lower than that of the blood test (98.03% vs 99.68%, respectively).
• However the specificity of the 2 test methods -- the ability to determine that an uninfected person is HIV negative -- was similar (99.74% vs 99.91%, respectively).
• Negative likelihood ratios were small and similar for oral and blood specimens (0.019 vs 0.003, respectively).
• Positive likelihood ratios, however, were different (383.37 vs 1105.16, respectively).
• In high-prevalence settings (greater than 1% of the population), PPVs of the testing methods were similar: 98.65% for oral samples and 98.50% for blood samples.
• But in low-prevalence settings, PPVs were lower for oral samples compared with blood samples: 88.55% vs 97.65%, respectively.

Based on these findings, the researchers concluded, "Although OraQuick had a high PPV in high-prevalence settings in oral specimens, the slightly lower sensitivity and PPV in low-prevalence settings in oral specimens should be carefully reviewed when planning worldwide expanded initiatives with this popular test."

They noted that accuracy was high overall and confidence intervals (a way to express statistical likelihood of results) had "a great deal of overlap," suggesting that the clinical significance of the difference might not be very great, and acknowledged that these findings -- mostly derived from high-income countries and well-controlled studies in lower income countries -- might not reflect accuracy under real-world conditions in resource-limited settings.

Investigator affiliations: Department of Medicine and Department of Epidemiology, Biostatistics & Occupational Health, McGill University, Montreal, QC, Canada; Division of Clinical Epidemiology & Infectious Diseases, McGill University Health Center, Montreal, QC, Canada; Division of Infectious Diseases, Department of Medicine, Queen's University, Kingston, ON, Canada; Institut National de Santé Publique du Québec, Montreal, QC, Canada; Direction de Santé Publique de l'Agence de la Santé et des Services Sociaux de Montréal, Montreal, QC, Canada; Department of Clinical Research, London School of Hygiene and Tropical Medicine, London, UK.

1/24/12

Reference

N Pant Pai, B Balram, S Shivkumar, et al. Head-to-head comparison of accuracy of a rapid point-of-care HIV test with oral versus whole-blood specimens: a systematic review and meta-analysis. The Lancet Infectious Diseases. January 24, 2012 (Epub ahead of print).