Testing for Influenza
Abstract & Commentary
By Allan J. Wilke, MD, Professor, Department of Introduction to Clinical Medicine, Ross University School of Medicine, Commonwealth of Dominica. Dr. Wilke reports no financial relationships relevant to this field of study.
Synopsis: Rapid influenza diagnostic tests have modest sensitivity and high specificity.
Source: Chartrand C, et al. Accuracy of rapid influenza diagnostic tests: A meta-analysis. Ann Intern Med 2012;156:500-511.
There are three main ways of testing for influenza: viral culture (VC), reverse transcriptionase polymerase chain reaction (RT-PCR), and immunochromatographic assays. The latter are often called rapid influenza diagnostic tests (RIDTs), which reflects their chief advantage their results are generally available in 15 to 30 minutes. VC was the original gold standard, but suffers from long delay between acquisition of samples to results, 3 to 10 days. RT-PCR has replaced VC as the gold standard because it has the highest sensitivity and specificity. However, it is also the most expensive, and results are available in hours rather than minutes. Immunofluorescence (IF) and enzyme-linked immunosorbent assays (ELISA) are other less commonly used methods. Of the various types of tests, then, RIDTs would appear to be the most clinically useful, assuming that they are accurate. Chartrand and colleagues performed a meta-analysis to determine RIDTs' accuracy.
Searching PubMed, EMBASE, BIOSIS, and Web of Science, the investigators found 286 articles that met their inclusion criteria: RIDT accuracy compared to VC, RT-PCR, or both. They excluded articles that compared their accuracy to IF or ELISA. After initial review, 119 articles remained and formed the basis for the meta-analysis.
Pooled sensitivity was 62.3% (range, 4.4% to 100%). Pooled specificity was 98.2% (range, 50.5% to 100%). In subgroup analysis, RIDTs were more sensitive in children (66.6%) vs adults (53.9%); there was no difference in specificity. RIDTs were more sensitive in detecting influenza A than B (64.6% vs 52.2%, respectively). There was no significant difference in their ability to detect influenza A (H3N2), the usual strain that circulates in North America compared to influenza A (H1N1), the 2009 pandemic strain ("swine flu"). The type of respiratory sample and the type of person collecting it (lab tech vs office personnel) did not influence results. However, the timing of collection did, with a tendency for collection on day 2 or 3 of symptoms to be more sensitive than day 1. Industry-sponsored studies reported higher sensitivities than those without that support (pooled sensitivity 73.3% vs 59.4%). When specific brands were studied, Directigen Flu A® had the highest pooled sensitivity (76.7%). QuickVue Influenza® was second at 69.0%. BinaxNOW,® Directigen Flu A+B®, and QuickVue A+B® were all less sensitive at 57.0%, 57.2%, and 48.8%, respectively. Specificities were all > 95%.
Commentary
The influenza-like illness (ILI) case definition used by the Centers for Disease Control and Prevention (CDC) for national surveillance is fever > 100° F (37.8° C) and cough and/or sore throat (in the absence of a known cause other than influenza).1 Of course, the problem with this definition is that there are many ILIs that present with a fever and cough or sore throat. Some are viral and some are bacterial. The sensitivity and specificity of fever and cough for influenza are 64% and 67%, respectively.2 The sensitivity of a symptom history then is greater than that for RIDTs. (Reminder: Sensitivity measures the true positive rate, and specificity the true negative rate.) The CDC has published a RIDT algorithm that advises their use in people who present with signs and symptoms consistent with influenza and where "the results of influenza virus testing [will] change clinical care of the patient (especially for hospitalized patients and those with high-risk conditions) or influence clinical practice for other patients."3
One of the actions you might consider is prescribing a neuraminidase inhibitor (NI), oral oseltamivir (Tamiflu®), or inhaled zanamivir (Relenza®). However, the efficacy of NIs has been called into question.4 The best offense though remains a good defense: yearly vaccination.
One could speculate why industry-sponsored studies had better results than independent studies, but I won't. The authors reason that the fact that children carry higher viral loads than adults explains why the sensitivities are higher in them. They invoke similar reasoning to explain the better results on day 2 or 3 vs day 1: we shed more viruses on those days. They caution against reading too much into the differences between brands since no head-to-head studies have been performed.
This has been a strange influenza season with a very late start, low total activity, and already a decrease in activity.5 While it appears that the season may have already peaked, the CDC does not want us to become complacent and warns, "Sporadic influenza outbreaks continue to occur."
References
1. www.cdc.gov/flu/weekly/pdf/overview.pdf. Accessed April 8, 2012.
2. Call SA, et al. JAMA 2005;293:987-997.
3. www.cdc.gov/flu/professionals/diagnosis/clinician_guidance_ridt.htm#Figure1. Accessed April 8, 2012.
4. Jefferson T, et al. Cochrane Database Syst Rev 2010;(2): CD001265.
5. www.cdc.gov/flu/weekly/summary.htm. Accessed April 8, 2012.
Rapid influenza diagnostic tests have modest sensitivity and high specificity.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content