Objective. For preventing nosocomial influenza infections and to facilitate prompt antiviral therapy, an accessible, rapid diagnostic method for influenzavirus is needed. We evaluated the performance of a lateral-flow immunoassay (QuickVue Influenza Test) completed at the bedside of hospitalized children during the influenza season.
Methods. All children who were evaluated at a large teaching hospital during the 1999 to 2000 influenza season were eligible if they were 1) younger than 19 years and hospitalized with respiratory symptoms or 2) younger than 3 years and hospitalized with fever. Each study child had 2 nasal swabs obtained—1 for influenzavirus culture and polymerase chain reaction (PCR) and the other for the QuickVue Influenza Test. The performance of the rapid diagnostic test was compared with the results of culture or PCR for influenza A or B.
Results. Of 303 eligible children, 233 (77%) were enrolled. In this population, 19 children had culture- and/or PCR-confirmed influenza A infection, prevalence of 8%. The QuickVue Influenza Test had a sensitivity of 74%, specificity of 98%, positive predictive value of 74%, and negative predictive value of 98%.
Conclusions. Among children hospitalized with fever/respiratory symptoms during the influenza season, negative bedside QuickVue Influenza Tests indicated very low likelihood of influenza infection, whereas positive tests greatly increased the probability of influenza-associated illness.
Influenzavirus has a significant impact on the pediatric population, with school-aged children having the highest infection rates.1,2 Retrospective cohort studies in healthy children younger than 15 years indicate that annual influenza-attributable rates of hospitalizations are 4 to 104 per 10 000 children, outpatient visits are 6 to 15 per 100 children, and antibiotic prescriptions are 3 to 9 per 100 children.3,4 Although the youngest children experience the majority of influenza-related hospitalizations, all age groups have excess outpatient visits.3 In addition, children with high-risk conditions have higher rates of hospitalization,5 outpatient visits, and antibiotics prescriptions than age-matched children without high-risk conditions.6,7 Thus, influenza infections increase health care utilization for all pediatric age groups.
Influenzavirus often circulates concurrently with respiratory syncytial virus (RSV), the most common cause of respiratory illnesses in young children.8–14 Although distinct clinical syndromes are attributed to influenzavirus and RSV, there can be considerable overlap in their manifestations. Before admitting young children during winter months, most pediatric hospitals routinely use rapid diagnostic testing for RSV to facilitate policies that minimize nosocomial infections.15 In contrast, many pediatric hospitals do not routinely perform rapid diagnostic tests for influenzavirus. Rapid influenza tests not only would help to minimize nosocomial influenzavirus infections during the winter16 but also would provide physicians the opportunity to use targeted antiviral therapy, which is reported to be effective if initiated within 48 hours of the onset of symptoms.17–23
The purpose of this prospective study was to compare the performance of a rapid diagnostic test (QuickVue Influenza Test; Quidel Corp, San Diego, CA) completed at the bedside of hospitalized children to viral culture and/or polymerase chain reaction (PCR) for influenzavirus.
All children who were admitted to Vanderbilt Children’s Hospital between January 10 and February 15, 2000, were eligible if they were 1) younger than 19 years and hospitalized with respiratory symptoms or 2) younger than 3 years and hospitalized with fever. Eligible children had 1) a primary admission diagnosis of an acute respiratory illness characterized by rhinorrhea, sore throat, cough, shortness of breath, or apnea or 2) a primary admission diagnosis consistent with a febrile illness and a temperature of >100.4°F. Broad inclusion criteria were chosen such that all children who were hospitalized with symptoms potentially related to influenza infections were eligible.
Each calendar day, potential subjects were identified by a review of all admission logs to pediatric inpatient services. A research team member approached eligible children and their parents for consent within 24 hours of admission. Discharge logs were reviewed to verify that all eligible patients had been identified. This study was reviewed and approved by the Vanderbilt University Institutional Review Board.
This prospective, cross-sectional study enrolled children who were hospitalized with fever or respiratory illnesses as discussed above. The child’s medical history was obtained by a standardized questionnaire administered to the parent/guardian. From each child, 2 nasal swabs of the turbinates were obtained—1 for influenzavirus culture and PCR and the other for the rapid diagnostic test. The rapid test results were compared with that of culture and PCR for influenzavirus.
The local influenza season, which was defined as the weeks that influenzavirus was identified in at least 2 viral cultures at Vanderbilt University Hospital’s virology laboratory or the Vanderbilt Vaccine Evaluation Clinic’s research laboratory, spanned from December 5, 1999, to March 5, 2000. The study began on January 10, after the rapid diagnostic kits became available, and ended on February 15, after the peak influenza season.
A questionnaire was developed by a group of experts, pretested, and modified for item clarification. The questionnaire, which was administered to parents, obtained information regarding the child’s medical history, including symptoms associated with the acute illness precipitating this hospitalization. Symptoms included fever, cough, rhinorrhea, dyspnea, and other. Number of symptom days was coded as 3 or fewer days or more than 3 days because positive cultures in children who are infected with influenzavirus are most likely during the first 3 symptom days.24
Parents were queried to determine which children had high-risk conditions for which influenza immunization was explicitly recommended in the 2000 Red Book. They included 1) asthma or other chronic pulmonary diseases, 2) hemodynamically significant cardiac disease, 3) immunosuppressive disorder or therapy, 4) human immunodeficiency virus infection, 5) sickle cell anemia and other hemoglobinopathies, 6) diseases requiring long-term aspirin therapy, 7) chronic renal dysfunction, and 8) chronic metabolic disease.23 Children with 1 or more high-risk conditions were coded as high risk, and all others were coded as low risk.
Influenza infection was defined as any sample with 1) a positive culture for influenzavirus or 2) 2 consecutive positive PCRs for influenza A or B.
One nasal specimen was collected on a Dacron applicator, transported in refrigerated Hanks’ balanced salt solution with antibiotics and 0.5% gelatin, and cultured on 2 test tubes of rhesus monkey kidney cells. Cells were observed for cytopathic effect and tested for hemadsorption with 0.1% guinea pig red blood cells 5 and 10 days after inoculation. Hemadsorbing agents were tested by indirect fluorescent antibody to differentiate between influenzavirus and parainfluenza virus and by hemagglutination inhibition assay to determine influenzavirus serotype.
The remaining nasal specimen from the Dacron applicator was stored in a −70°C freezer until PCR. Nucleic acids from 100 μL of frozen nasal washes were extracted with RNAzol B (Leedo Laboratories, Inc, Houston, TX) according to the manufacturer’s instructions. Colorimetric microtiter plate reverse transcriptase-PCR systems were performed for influenza A and B viral RNA detection as described previously.25 Each PCR assay was performed on freshly extracted RNA. A degenerate primer set (FluA01: 5′-CTT CTR ACC GAR GTC GAA ACG-3′ (R = A and G) and FluA02: 5′-GAC AAA GCG TCT ACG CTG CAG-3′) and a 5′-biotinylated capture probe (5′-TCC TGT CAC CTC TGA CTA AGG G-3′) were designed to target influenza A virus matrix gene and to create a 234-base pair product.26 The primers and probes for influenza B virus hemagglutinin gene were described previously.27 The test sensitivity is approximately 0.01 plaque-forming units per milliliter for both of these PCR assays, which were developed and validated at Vanderbilt University Medical Center under the 1988 Clinical Laboratory Improvement Amendments regulation.
Rapid Diagnostic Test
A second nasal specimen was collected on the QuickVue Influenza Test applicator and then mixed with a prepackaged extraction solution in a test tube. Next, the test strip was inserted into this test tube and read 10 minutes later. Each researcher was trained to perform and interpret the rapid diagnostic test at the bedside according to the manufacturer’s instructions. A blue line indicated a valid test. A red line of equal or less intensity than the blue line but visible in dim light was interpreted as positive. The absence of a red line or the presence of a faint red line only visible with bright light was interpreted as negative. The laboratory technician who performed the culture and PCR was masked to the rapid diagnostic test results.
The primary outcome, the performance of the QuickVue Influenza Test, was calculated by comparing its results with those of the criterion standard. The secondary outcome was the performance of the rapid test in a subset of children hypothesized to have high titers of influenzavirus in the nose: 1) children in their first influenza season or 2) those with symptoms for <4 days. Demographic characteristics were compared using Fisher’s exact test or χ2 analysis. Gender and race (white, nonwhite) were dichotomous variables. Age was classified as younger than 6 months, 6 months to younger than 5 years, and 5 years to 18 years. Stata, Version 6.0 (Stata, College Station, TX) was used for all analyses.
Of 625 children hospitalized during the study period, 303 (48%) had qualifying admission criteria of respiratory symptoms and/or fever (Fig 1). Of 303 qualifying children, 233 (77%) were enrolled and had both nasal samples obtained. Reasons that children were not enrolled in the study were 1) failure to identify a child within 24 hours of admission (8%), 2) child or parent declined to participate (8%), and 3) unable to contact a parent (5%). Six children (2%) in whom responses to the questionnaire were obtained but permission to obtain nasal specimens was denied were enrolled. All 6 of these children had had a rapid diagnostic test for influenzavirus performed by the hospital laboratory before enrollment; none of these children was included in the analysis. Study children were more frequently male and were predominantly white, and 43% were between the ages of 6 months and 5 years (Table 1). Of the study population, 124 (53%) had high-risk conditions; the most common were asthma (32%), congenital heart disease (11%), malignancy (3%), and cystic fibrosis (2%).
Of 233 enrolled children, 222 (95%) had respiratory symptoms, 163 (73%) of which also had fever. Their admission diagnoses included apnea, asthma exacerbation, bronchiolitis, cystic fibrosis exacerbation, croup, pharyngitis with dehydration, pneumonia, and respiratory distress or failure. Cough was present in most (91%) of these children, dyspnea or apnea in 198 (89%), and rhinorrhea in 172 (72%).
Of 11 children enrolled with fever alone, 9 (82%) were younger than 1 month. Temperatures in these neonates ranged from 100.6°F to 103.8°F with an average of 101.8°F. Eight neonates (89%) were admitted with the diagnosis of febrile neonate without localizing signs and 1 with a perirectal abscess. The 2 other children with fever and without respiratory symptoms were hospitalized with the diagnoses of sepsis and fever status-post surgical procedure, respectively.
Nineteen study children (8%) had influenza infections as determined by positive culture or 2 consecutive positive PCRs for influenzavirus from 1 nasal specimen (Table 2). Eleven (58%) had positive cultures, all influenza A (H3N2), and 18 (95%) had 2 positive PCRs. Influenza B was not detected in any participant by culture or PCR. One child had a positive influenza culture and negative PCR. Of children infected with influenza A, 14 (74%) had positive rapid diagnostic tests. Of the 19 children with influenza infections, only 2 (11%) were diagnosed by a rapid diagnostic test performed by the hospital laboratory in the outpatient or inpatient setting before enrollment. The seasonal occurrence of eligible children and the number of positive tests for influenzavirus followed similar epidemic curves, although local virology surveillance data for the 1999 to 2000 season indicates that influenzavirus and RSV had similar peaks and epidemic curves. Admission diagnoses for these 19 children were febrile neonate (5), bronchiolitis (2), croup (1), apnea (1), pneumonia (1), asthma exacerbation (2), fever and neutropenia (2), seizures and febrile illness (2), fever status-post surgical procedure (1), fever and sickle cell anemia (1), and fever of unknown origin (1).
The QuickVue Influenza Test had a sensitivity of 74% and a specificity of 98%. The positive and negative predictive values were 74% and 98%, respectively (Table 3). The positive and negative likelihood ratios, measures of the discriminatory power, were 31.5 and 0.26, respectively.
We tested the hypothesis that the QuickVue Influenza Test would be more sensitive among children with the highest viral titers: those experiencing their first influenza season and those with a short duration of symptoms, defined as <4 days. The sensitivity of the rapid diagnostic test was 100% for both subgroups, and the specificity was 98% and 97%, respectively (Table 3).
In this prospective study, influenza A infections were identified in 8% of children who were 1) younger than 19 years and hospitalized with respiratory symptoms or 2) younger than 3 years and hospitalized with fever. Influenza infections were identified by positive culture and/or 2 consecutive PCRs for influenza A or B. In this study, the QuickVue Influenza Test had a sensitivity and specificity of 74% and 98%, respectively. Subgroup analysis of children who likely had the highest viral loads indicated that the sensitivity of the rapid diagnostic test increased to 100% in children who were younger than 6 months of age and those with symptoms for <4 days.
We used a combination of a specific test (culture) and a sensitive test (PCR) as our criterion standard. To minimize false-positive results, we required a positive result on 2 consecutive PCR assays to define a sample as PCR positive. To our knowledge, no other rapid diagnostic test for influenzavirus has used this methodology.28–31 PCR was included in the influenza infection definition because PCR is more sensitive than culture for influenzavirus, with an overall reported increase in the detection rate of influenzavirus by 3% to 40%.23,27,29,32–38 In this study, all influenza infections were influenza A, and PCR increased the detection of influenzavirus by 60%, from 5% detected by culture alone to 8% by culture or PCR.
Results of this study need to be interpreted with several caveats. First, only influenza A was identified in the study samples. Because we did not detect influenza B in our study population, the rapid test characteristics for influenza B infections could not be assessed. Second, the study was conducted in the hospital setting. Rapid diagnostic tests need evaluation in the ambulatory setting where the viral titers or time between onset of illness and presentation to a health care provider might differ from patients in this study. Third, we trained researchers to perform this test according to the manufacturer’s instructions. Modifying the methods by which the sample was collected or tested may alter the performance of this rapid diagnostic test. Fourth, using acute and convalescent sera as the criterion standard could potentially identify other children with influenzavirus infection. However, sera were not obtained in our study because some parents would not have consented to phlebotomy and would not have returned for the convalescent sample. Finally, the power to detect potentially significant characteristics associated with influenza infection was limited by the fact that only 19 children had influenza infections.
The QuickVue Influenza Test can be used to diagnose influenza infections at the bedside of children with a compatible clinical syndrome during the influenza season. A negative test made the diagnosis of influenza infection unlikely. A positive test should be interpreted as probable influenza infection only if influenzavirus is circulating in the community. During periods in which 10% to 15% of children who are hospitalized with respiratory symptoms have influenzavirus, 78% to 85% of children with positive rapid diagnostic tests would be expected to have influenza infection. Hence, during the influenza season, a positive result could assist clinicians in identifying the need for respiratory isolation for influenzavirus and possibly deciding whether to initiate antiviral agents directed against influenzavirus. Caution is warranted in interpreting positive tests during periods with little or no influenzavirus circulating in the community because false-positive results would predominate.
In our study population, a positive rapid diagnostic test for influenzavirus changed the isolation status of some children but did not change the likelihood of receiving antibiotic therapy. Similarly, another study reported that children who were hospitalized with a positive rapid diagnostic test obtained in the emergency department were as likely to receive antibiotics as those with a negative rapid diagnostic test.39 The lack of association of a positive test and likelihood of receiving antibiotic therapy in our study may reflect either the perceived need to initiate antibiotics in hospitalized children with fever or respiratory symptoms or that the decision to initiate antibiotics was frequently made concurrently with the decision to admit, which was before study enrollment. Of 7 enrolled children with a positive rapid test for influenzavirus within 48 hours of the onset of symptoms, none received antiviral therapy for influenzavirus. Two children had a positive test within 24 hours (both younger than 1 month), and 5 children had a positive test between 24 and 48 hours, 2 of which only were older than 1 year (1.9 and 2.2 years, respectively).
Currently, 5 rapid diagnostic tests for influenzavirus are available for clinical use (Table 4). FLU OIA (Thermo BioStar, Inc, Boulder, CO), QuickVue Influenza, and ZstatFLU (ZymeTX, Inc, Oklahoma City, OK) identify influenza A or B but do not distinguish between them. Only Directagen FLU A+B (Directagen BD Diagnostic Systems, Sparks, MD) identifies and differentiates influenza A and B. Directagen FLU A identifies influenza A only. Table 4 outlines the performance of these tests according to the products’ package inserts40–43 or the article from which the data used in the packet insert is described.30
Additional studies are needed to confirm these findings, to evaluate the performance of the rapid diagnostic test with influenza B infections, to determine the results in other clinical settings, and to identify patients with a higher likelihood of influenza infection in whom the test results, if positive, would be considered confirmatory. In addition, this test should be compared with the other available rapid diagnostic tests for influenzavirus. Ideally, all of the rapid diagnostic tests should be tested using the same methodology for obtaining specimens and criterion standard for determining an influenza infection.
Potentially, the most important aspect of this rapid test is that it can provide timely, accurate, and useful information at the bedside. The information can be provided in real time when the diagnostic, isolation, and therapeutic questions need to be addressed, not hours later. Because the QuickVue Influenza Test received a Clinical Laboratory Improvement Amendments Waiver, the rapid diagnostic test can be performed in clinical settings with a Certificate of Waiver from the Centers for Medicare and Medicaid Services.44 In this study, only 2 (11%) of 19 children with influenza infections were diagnosed before enrollment. Routine use of rapid diagnostic testing for influenzavirus during the influenza season would have increased the number of children diagnosed with influenzavirus and changed the isolation status of some children. Our data suggest that the QuickVue Influenza Test may be useful for children who are hospitalized with febrile or respiratory illnesses during the influenza season.
Funding was provided by the Quidel Corporation, a cooperative agreement (#U50/CCU41398) from the Centers for Disease Control and Prevention, the Veterans Affairs Tennessee Valley Healthcare System Quality Scholars Program, and an unrestricted educational grant from the Pfizer Foundation. The Quidel Corporation provided all of the QuickVue Influenza Test kits and gave an unrestricted gift, which covered the cost of half of the PCRs.
We thank Gay Waddling, RN; Ayesha Khan, MPH; Brent Frisbee, BS; and Brian Emerson, BS, for their contributions to the data collection and Lisa Rush for data entry.
- Received November 2, 2001.
- Accepted February 20, 2002.
- Reprint requests to (K.A.P.) Vanderbilt Children’s Hospital, Division of General Pediatrics, D-5028 Medical Center East, Nashville, TN 37232-8555. E-mail:
This work was presented at the Options for the Control of Influenza, Crete, Greece, September 2000 and at the Pediatric Academies Societies meeting, Baltimore, MD, April 29, 2001.
- Tsai HP, Kuo PH, Liu CC, Wang JR. Respiratory viral infections among pediatric inpatients and outpatients in Taiwan from 1997 to 1999. J Clin Microbiol.2001;39 :111– 118
- ↵Langley JM, LeBlanc JC, Wang EE, et al. Nosocomial respiratory syncytial virus infection in Canadian pediatric hospitals: a Pediatric Investigators Collaborative Network on Infections in Canada Study. Pediatrics.1997;100 :943– 946
- ↵Pickering LE. Influenza. In: 2000 Red Book: Report of the Committee on Infectious Diseases. 25th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2000:351–359
- ↵Cherian T, Bobo L, Steinhoff MC, Karron RA, Yolken RH. Use of PCR-enzyme immunoassay for identification of influenza A virus matrix RNA in clinical samples negative for cultivable virus. J Clin Microbiol.1994;32 :623– 628
- ↵Schweiger B, Zadow I, Heckler R, Timm H, Pauli G. Application of a fluorogenic PCR assay for typing and subtyping of influenza viruses in respiratory samples. J Clin Microbiol.2000;38 :1552– 1558
- ↵Boivin G, Hardy I, Kress A. Evaluation of a rapid optical immunoassay for influenza viruses (FLU OIA test) in comparison with cell culture and reverse transcription-PCR. J Clin Microbiol.2001;39 :730– 732
- ↵Herrmann B, Larsson C, Zweygberg BW. Simultaneous detection and typing of influenza viruses A and B by a nested reverse transcription-PCR: comparison to virus isolation and antigen detection by immunofluorescence and optical immunoassay (FLU OIA). J Clin Microbiol.2001;39 :134– 138
- ↵Covalciuc KA, Webb KH, Carlson CA. Comparison of four clinical specimen types for detection of influenza A and B viruses by optical immunoassay (FLU OIA test) and cell culture methods. J Clin Microbiol.1999;37 :3971– 3974
- ↵Kehl SC, Henrickson KJ, Hua W, Fan J. Evaluation of the Hexaplex assay for detection of respiratory viruses in children. J Clin Microbiol.2001;39 :1696– 1701
- Carman WF, Wallace LA, Walker J, et al. Rapid virological surveillance of community influenza infection in general practice. BMJ.2000;321 :736– 737
- ↵van Elden LJ, Nijhuis M, Schipper P, Schuurman R, van Loon AM. Simultaneous detection of influenza viruses A and B using real-time quantitative PCR. J Clin Microbiol.2001;39 :196– 200
- ↵QuickVue Influenza Test Product Package Insert. Quidel Corporation, San Diego, CA. Available at: http://www.quidel.com/libraries/pkginserts/RD/QuickVue_influenza.pdf. Accessed October 8, 2001
- ZstatFlu Product Package Insert. ZymeTx, Inc, Oklahoma City, OK. Available at: http://www.zymetx.com/physician/zstatflu_package_insert.pdf. Accessed October 8, 2001
- Laboratory Procedure Directigen Flu A for the Direct Detection of Influenza A Antigen. Sparks, MD: Becton, Dickinson and Company; 2001
- ↵Laboratory Procedure Directigen Flu A+B for the Differentiated, Direct Detection of Influenza A and B Antigens. Sparks, MD: Becton, Dickinson and Company; 2001
- ↵PR Newswire Association. Quidel Rapid Flu Test Now Available for Widespread Use in the US. Available at: http://www.findarticles.com/cf_0/m4PRN/2000_Oct_12/66000937/print.html. Accessed October 12, 2000
- Copyright © 2002 by the American Academy of Pediatrics