Rapid diagnostic tests suboptimal for influenza A (H1N1) detection
Click Here to Manage Email Alerts
Results from two recently published studies confirm earlier reports that rapid influenza diagnostic tests have low sensitivity for diagnosing pandemic influenza A (H1N1) in pediatric patients.
Researchers from the Texas Children’s Hospital in Houston compared the diagnostic accuracy of a rapid influenza diagnostic test (RIDT, BinaxNOW Influenza A and B, Inverness Medical) with viral culture and real-time reverse-transcriptase polymerase chain reaction (rRT-PCR), retrospectively analyzing data from 3,030 respiratory specimens collected at the hospital from May 21, 2009, to Sept. 20, 2009.
The researchers determined that the RIDT had an overall sensitivity of 45% (95% CI, 43.3-46.3) and a specificity of 98.6% (95% CI, 98.1-99) when rRT-PCR was used as a reference standard. When using viral culture as the reference standard, the RIDT had a 55.5% sensitivity (95% CI, 51.9-95.6) and specificity of 95.6% (95% CI, 95-96.1). Test sensitivity was higher among children aged younger than 2 years compared with older children (57.6% vs. 43.4%; 66.7% sensitivity in neonates). But the researchers encouraged physicians to further investigate the etiology of fever in these patients, even if a rapid test yields a positive result for influenza, given the potential risk for a serious concomitant infection, such as secondary bacterial pneumonia or an occult urinary tract infection.
“Treatment of children at high risk who experience influenza-like illness during periods of high prevalence of influenza in the community should not rely solely on the results of RIDT,” the researchers wrote.
A second study conducted by researchers from the Hospital for Sick Children in Toronto yielded similar results. From May 22, 2009, to July 25, 2009, the researchers performed RIDT on 820 children aged 17 years and younger who presented to the ED or a community pediatric clinic with influenza-like illness.
Analysis showed that RIDT had a sensitivity of 62% (95% CI, 52-70) and a specificity of 99% (95% CI, 92-100) when rRT-PCR was used as a reference standard, with higher sensitivities also observed in younger children. “A positive RIDT result indicates that influenza infection is likely present; however, a negative test result does not rule out infection,” the researchers wrote.
Physicians should continue to consider influenza as a diagnosis and initiate empiric antiviral therapy, clinical evaluation and infection-control measures when appropriate, according to researchers from both studies. – by Melissa Foster
Cruz AT. Pediatrics. 2010;125:e645-e650.
Hawkes M. Pediatrics. 2010;125:e-639-e644.
These reports confirm previously-published studies showing very poor sensitivity of the rapid influenza tests for diagnosing H1N1 influenza. This was already apparent in the Spring 2009 outbreak. While we previously relied on the in-hospital direct fluorescent antibody assay, the Toronto article demonstrates what we have recently found — that compared with rRT-PCR it misses some H1N1 cases.
It's important to consider the reasons behind doing these tests. If you are doing surveillance and you want the most accurate data, you need to use rRT-PCR. In office practice there is no pressing need to have a highly sensitive assay for most patients unless you are dealing with a high-risk patient who is likely to need antiviral medication.
– Robert Baltimore, MD
Infectious Diseases in Children Editorial Board
Follow the PediatricSuperSite.com on Twitter. |