|Home | About | Journals | Submit | Contact Us | Français|
The performance of two commercially available rapid test kits for influenza virus detection was compared to that of viral culture by using 356 nasal wash specimens collected during the 2001 to 2002 influenza season. Overall, the two rapid tests were easy to perform and showed comparable sensitivities (70.4 and 72.2%) and specificities (97.7 and 98.3%); for both test kit groups, most of the specimens that yielded false-negative results were found to be growing influenza B virus.
Influenza virus causes significant morbidity and mortality worldwide in both adults and children. The rapid and accurate detection of influenza virus allows for prompt administration of appropriate antiviral therapy and judicious use of antibiotics, allows hospitals and emergency departments to group or isolate influenza virus-infected patients to reduce the nosocomial spread of infection, reduces the incidence and length of hospital stays, identifies potential epidemic or pandemic strains in a timely manner, and differentiates influenza virus from other infectious and biological warfare agents such as those that cause anthrax and smallpox, diseases that may begin with flu-like symptoms (1, 3, 4, 5, 6, 9). There are currently at least six different kits that identify influenza virus in clinical specimens in 30 min or less, and their use has become widespread in laboratories and in point-of-care testing venues (2, 6, 7).
The performance of two commercially available rapid test kits for influenza virus was evaluated. One kit was a lateral-flow immunoassay (QuickVue; Quidel, San Diego, Calif.) that detected both influenza A and B viruses but did not differentiate between them. The other test was a membrane enzyme immunoassay (Directigen Flu A+B; Becton Dickinson Diagnostic Systems, Sparks, Md.) that both detected and differentiated influenza A and B viruses. The rapid test results for both assays were compared to the results of the reference standard of viral culture by using 356 fresh nasal wash specimens. These specimens were collected from children with respiratory virus symptoms who presented to Texas Children's Hospital for admission or evaluation in the emergency department between 30 January 2002 and 20 April 2002. Both of the rapid tests were performed by virology laboratory technicians during weekday, day shift hours according to the manufacturers' instructions (QuickVue [Quidel] and Directigen Flu A+B [Becton Dickinson] product package inserts).
Briefly, the QuickVue influenza virus test involved extraction of influenza A or B virus antigens from the patient specimens. Each patient's specimen was placed in a small tube containing an extraction agent, which disrupted the viral particles and exposed internal viral nucleoproteins. After extraction, a test strip was placed in the extraction reagent tube and was allowed to react in a lateral-flow chromogenic immunoassay format with mouse monoclonal antibody reagents specific for influenza A or B virus. A specimen that contained influenza virus antigens produced a pink to red test line on the reagent strip, indicating a positive test result. The absence of a pink or red test line in the presence of a positive-procedure control blue line indicated a negative test result.
The Becton Dickinson Directigen Flu A+B test is a rapid membrane enzyme immunoassay test that involves extraction of influenza A or B virus antigens from patient specimens. Each extracted specimen was expelled through a filter assembly into each of two wells of a triangular plastic test device containing a membrane surface. Viral antigens, if present in the extracted specimens, were bound to the membrane surface. Viral antigen was captured on the membrane by using enzyme-conjugated monoclonal antibodies specific for influenza A or B virus nucleoprotein, followed by a stop reagent. A positive test result was indicated by the presence of a purple triangle in well A or well B in the plastic device. The absence of a purple triangle in the presence of a positive-procedure control dot indicated a negative test result.
Both tests were performed by using the positive, negative, and procedural controls of the kits as well as external laboratory controls for each test kit run.
All specimens were also inoculated into culture monolayers of human foreskin fibroblasts, rhesus monkey kidney cells, and human lung carcinoma (A549) cells. Viral cultures were inspected daily under a light microscope for cytopathic effect, and hemadsorption with a 0.4% suspension of guinea pig red blood cells was determined on days 2, 5, and 14 of incubation of the rhesus monkey kidney cell cultures. Virus identification was confirmed by immunofluorescence assays (2). Viral cultures positive for influenza virus, type A or B, were considered true positives. Sensitivity, specificity, and positive and negative predictive values were calculated by using two-by-two contingency tables. Differences in results between tests were analyzed by using chi-square tests. Comparison of the ages of patients infected with influenza A virus, influenza B virus, noninfluenza viruses, and uninfected patients was performed by using analysis of variance with the assumption of equal variances. The Student t test was used to compare differences between the results for only two groups.
Of the 356 specimens submitted for this study, 236 (66.3%) had negative viral cultures and 120 (33.7%) grew at least one virus each. Of 124 viruses isolated, 25 (20.2%) were influenza A virus, 29 (23.4%) were influenza B virus, 19 (15.3%) were respiratory syncytial virus, 16 (12.9%) were rhinoviruses, 14 (11.3%) were adenovirus, 11 (8.9%) were cytomegalovirus, 7 (5.6%) were parainfluenza virus types 3 and 4, and 3 (2.4%) were enteroviruses. Dual viral infection was present in four patients. Two patients were infected with both cytomegalovirus and rhinovirus, one was infected with influenza A virus and adenovirus, and one was infected with influenza B virus and rhinovirus. The mean age of the patients tested was 5 years (median, 2.2 years; range, 3 months to 28 years). The mean distribution of the age of the patients with a negative viral culture was 4.84 years (median, 2.5 years), compared to 5.1 years (median, 1.6 years) for patients with a viral culture positive for influenza A virus, 10.6 years (median, 11.6 years) for patients with a viral culture positive for influenza B virus, and 3 years (median, 9 months) for patients with a viral culture positive for viruses other than influenza virus (P < 0.001). The mean age (± standard deviation) of the patients who were culture positive for influenza virus (8.02 ± 6.4 years) was significantly different from that of the patients who were culture negative for influenza virus (4.45 ± 5.5 years) (P = 0.0003). Overall, the two rapid tests performed similarly, with no significant difference between the results of the two methods (Table (Table1).1). False-positive tests were rare, but both tests had false-negative results (Table (Table2).2). Of 16 specimens that yielded false-negative results by the lateral-flow immunoassay method, 12 were actually positive for influenza B virus and 4 were actually positive for influenza A virus. Of 15 specimens that yielded false-negative results by the membrane enzyme immunoassay, 11 were actually positive for influenza B virus and only 4 were actually positive for influenza A virus. Both tests were performed easily by laboratory technicians in 15 min or less.
Both the lateral-flow and enzyme immunoassays were highly specific in detecting both influenza A and B viruses in nasal wash specimens; however, both assays were less able to reliably detect influenza B virus than influenza A virus. Furthermore, influenza B virus was more likely to be isolated and detected in older children. Most reported studies on the performance of rapid assays for the detection of influenza virus have been conducted during influenza seasons in which influenza A virus predominated (1, 3, 5, 6, 7, 8). However, in one study, a rapid neuraminidase detection assay (ZstatFlu; ZymeTx, Oklahoma City, Okla.) was also shown to be less sensitive for influenza B virus detection (4). Clinicians and laboratory personnel using rapid tests for influenza virus detection should be aware that the performance of these and other rapid tests may vary with the influenza virus type as well as with the age of the patient being tested (4, 8). Furthermore, since influenza B virus also causes morbidity and mortality and can be treated with antiviral agents such as the neuraminidase inhibitors oseltamivir and zanamivir, the ability of new rapid tests to detect both influenza A and B viruses should be carefully assessed and reported, and research to enhance the ability of these rapid tests to detect influenza B virus should continue.