Abstract
To provide 24-h influenza diagnosis for adults presenting to the emergency department, the Directigen Flu A+B enzyme immunoassay (EIA) was performed in the chemistry laboratory during the night shift. Nasopharyngeal swabs were retested by cytospin-enhanced direct immunofluorescence (DFA; SimulFluor respiratory screen) when the virology laboratory opened. The influenza EIA detected 16 influenza A virus infections, whereas cytospin-enhanced DFA detected 31 influenza A virus infections as well as 3 respiratory syncytial virus, 2 adenovirus, and 1 parainfluenza virus infections. A positive EIA result usually correlated with 50 or more influenza virus cells positive by DFA.
Influenza is a major cause of morbidity and mortality, especially in the elderly and in those with chronic cardiac or pulmonary conditions. Rapid diagnosis in patients admitted to the hospital facilitates proper bed allocation, helps to prevent nosocomial transmission, and allows early institution of treatment with amantidine or a neuraminidase inhibitor.
The clinical virology laboratory at Yale New Haven Hospital performs cytospin-enhanced direct immunofluorescence staining (DFA) by using a respiratory screen reagent that detects multiple viruses and provides results comparable to those of cell culture (12). Results are provided within 1.5 to 2 h during laboratory operating hours. However, DFA is labor-intensive, requires significant training and expertise, and is not available when the virology laboratory is closed. In contrast, a rapid influenza enzyme-linked immunosorbent assay (EIA) can be completed within 15 min and does not require extensive technical expertise or equipment (1, 3). This test can be implemented in doctors' offices or in laboratories without virology expertise.
The recent death of an elderly woman from inhalation anthrax in a nearby Connecticut town led to recommendations in the local news media that patients with “flu-like symptoms” should report to hospital emergency rooms or their doctors' offices for a rapid influenza test. Test requests for rapid influenza testing in the 2001-2002 season were double those of previous years. In order to provide the adult emergency department with 24-h influenza diagnostic testing, a rapid influenza A and B virus EIA was performed in the chemistry laboratory between midnight and 6 a.m. during the peak influenza season. Samples were retested by DFA when the virology laboratory opened. The results are presented here.
One hundred fifty-two nasopharyngeal (NP) swabs collected in M4 medium (MicroTest, Inc., Lilburn, Ga.) from adults evaluated in the emergency department after midnight were tested in the clinical chemistry laboratory by use of the Directigen Flu A+B enzyme immunoassay (EIA) (Becton Dickinson, Cockeysville, Md.). The samples were retested by use of the cytospin-enhanced SimulFluor (Chemicon International, Temecula, Calif.) respiratory screen in the clinical virology laboratory in the morning, and the numbers of DFA-positive cells were recorded. EIA cassettes were reexamined in the morning in the virology laboratory. A number of EIA-negative, DFA-positive samples were retested by EIA in the virology laboratory to confirm EIA-negative results, including all EIA-negative samples with 50 or more cells positive by DFA. For the SimulFluor respiratory screen DFA, samples were processed, applied to slides by cytocentrifugation, stained, and examined as previously reported (12). An adequate sample required a minimum of 25 columnar epithelial cells. For the Directigen Flu A+B EIA, the manufacturer's instructions were followed.
Sixteen of 152 NP swabs were positive for influenza A virus by EIA, and 31 were positive by DFA (Table 1) (P = 0.0001, McNemar's test). No samples were EIA positive and DFA negative, and no influenza B virus infections were detected during this period. In addition, six samples were positive for other viruses by the SimulFluor respiratory screen DFA (three respiratory syncytial virus [RSV], two adenovirus, and one parainfluenza infection). In total, 16 virus infections were diagnosed by EIA and 37 were diagnosed by DFA.
TABLE 1.
Comparison of cytospin-enhanced DFA with membrane EIA for detection of influenza A virusa
Results of Directigen Flu EIA | No. of samples with indicated result by cytospin-enhanced DFA
|
||
---|---|---|---|
Positive | Negative | Total | |
Positive | 16 | 0 | 16 |
Negative | 15 | 121 | 136 |
Total | 31 | 121 | 152 |
P < 0.0001 (McNemar's test). No influenza B viruses were detected during the study period. Additional virus infections detected by the respiratory screen DFA in adult NP swabs were three RSV, two adenovirus, and 1 parainfluenza.
The correlation between the number of antigen-containing cells and a positive result by EIA was examined (Table 2). Twelve of 15 (80%) of the EIA-negative, DFA-positive samples had 20 or fewer positive cells detected by DFA. Thirteen of 16 (81%) of EIA-positive samples had 50 or more positive cells detected by DFA. Samples that had 50 or more DFA-positive cells but were EIA negative were retested by EIA, and the negative results were confirmed.
TABLE 2.
Correlation between number of DFA-positive cells and EIA result
EIA result | No. of samples with indicated no. of cells positive by DFAa
|
||||||
---|---|---|---|---|---|---|---|
<5 | 5-10 | 11-25 | 50-100 | >100 | >1,000 | Total no. of samples | |
Negative | 8 | 1 | 3 | 2 | 1 | 0 | 15 |
Positive | 0 | 1 | 2 | 2 | 5 | 6 | 16 |
Total | 8 | 2 | 5 | 4 | 6 | 6 | 31 |
No samples were in the category of 25 to 50 positive cells.
Rapid influenza testing has received increased attention in both the popular press and in medical publications due to the availability of new antiviral therapies (1, 14), the anticipation of a new influenza pandemic (15), and most recently, the concern about inhalation anthrax (2). A number of rapid influenza tests are now available (1, 7, 14). The Directigen Flu A+B EIA was chosen for this study since it does not require a unique collection device or a unique collection method, as do some of the other tests, and is among the most sensitive of the non-DFA tests. A recent study examined the Directigen Flu A+B EIA for predominantly young children and found it to be 96% sensitive for influenza A virus and 87.5% sensitive for influenza B virus (3). It also appears to be the most specific of the rapid diagnostic tests for influenza (1, 3). Specificity is of particular concern for hospitalized patients, since false-positive results could result in placing a negative patient in the same room with an infected patient or could lead to inappropriate therapy.
Unfortunately, as shown in this study, the rapid influenza EIA failed to detect almost half of the influenza cases. When performed on NP swabs from adults, the EIA detected only 16 of the 31 (52%) influenza A-positive infections detected by DFA. Surprisingly, a positive EIA result usually required 50 or more DFA-positive cells, and three samples with 50 to 100 DFA-positive cells were repeatedly EIA negative. However, some samples with few antigen-bearing cells tested positive, presumably since EIA can also detect cell-free virus. In addition, 16% of the positive adults had viruses other than influenza virus detected by DFA (i.e., three RSV, two adenovirus, and one parainfluenza virus). There were no false positives by EIA. Influenza B virus did not appear in the community until after the completion of this study. However, nine influenza B virus-positive samples (five from adults, four from children) were subsequently tested, and similar results were obtained. Four samples with 5 to 20 influenza B virus-positive cells by DFA were EIA negative, whereas four samples with 50 to 100 DFA-positive cells were EIA positive. One sample with 30 DFA-positive cells was very weakly positive by EIA. In adults, only two of five samples positive for influenza B virus by DFA were EIA positive.
For the best DFA results, an expert staff, cytospin preparation of slides, the highest quality reagents, a dark room, and a high-quality fluorescence microscope, with frequent changes of the microscope bulb to maintain brightness, are required (4, 10, 12). The SimulFluor DFA can also detect other viruses causing flu-like symptoms in adults in a single test (5, 6, 13) and can provide results within 2 h of sample receipt throughout the virology laboratory's operating hours. The rapid influenza EIA detects only influenza A and B viruses but can be performed accurately on swab specimens by personnel from the chemistry or other laboratories and can be available 24 h a day if needed.
DFA results are highly variable between laboratories, and each laboratory must determine and continually monitor its own sensitivity and specificity. Validation of DFA by comparison with culture results and continual attention to accuracy of interpretation are essential. Thus, DFA should be limited to those virology laboratories willing to invest the necessary time and effort. In our laboratory, DFA has performed as well or better than conventional culture methods for respiratory viruses, with the exception of adenovirus (12). The following data are presented to support the use of the cytospin-enhanced SimulFluor respiratory screen DFA as the “gold standard” test in the present report. From 1999 to 2002, we detected 1,008 influenza A virus positives by DFA and 613 of these positives (61%) were for adults. During this same period, 4,253 DFA-negative samples were inoculated into primary rhesus monkey kidney cell culture by using the most sensitive culture methods (8), and only 14 additional influenza A virus positives were recovered. Ten of the 14 (71.4%) positives missed by DFA were for adults. Thus, the DFA in our laboratory detected 1,008 of 1,022 (98.6%) influenza A virus-positive infections.
In the present study, we did not test samples with discrepant results by culture or reverse transcription (RT)-PCR. However, in a separate study in our laboratory in 2002, in which all samples were tested by culture, DFA, and RT-PCR, culture detected 150 (100%), cytospin-enhanced DFA detected 148 (98.6%), and real-time RT-PCR detected 143 (95.3%) of the influenza A virus positives. Since we now cancel cultures on most DFA-positive samples, our present specificity data are based on only a subset of DFA positives. However, for all lower respiratory tract samples, for samples from all immunocompromised hosts, and when there is any question about the DFA results, a culture is performed for confirmation. All DFA slides are kept for 1 month for comparison with culture results and reexamination if needed. While some false positives would be expected, in over 11,000 samples tested in the past 3 years, only one DFA result, a parainfluenza virus positive, was confirmed to be spurious due to nonspecific staining.
While the use of rapid influenza testing to allay fears of inhalation anthrax is a poor strategy and should not be recommended, rapid influenza testing has received increased attention as a potential guide to anti-influenza therapy. In a recent decision analysis paper, rapid testing as a pretreatment strategy was determined to be cost-effective only in the nonepidemic phase of influenza and only when the test sensitivity was 77% and the specificity was 95% (14). On the basis of our present and previous data (11) and those of others (9), the Directigen EIA using NP swabs from adults would not meet the sensitivity criteria stated (14). Steininger et al. (16) also found that the likelihood of negative results by EIA increased significantly with increasing age of the patient, with best results obtained in children less than 5 years of age.
Thus, clinicians should be aware that EIA test results, while excellent with high-titer samples from children (17), might be suboptimal when NP swabs from adults are tested (3, 9, 11, 15, 16). Laboratories can improve test results by intensifying efforts to educate clinicians on the importance and the mechanics of obtaining the highest quality samples. To increase the number of influenza virus-positive cells, clinicians should consider obtaining NP washes or combining two NP swabs or a throat and an NP swab and samples should be collected in the first 48 h of illness, when viral shedding is maximal (8, 9). Patients should also be counseled that a negative result does not exclude a diagnosis of influenza.
Acknowledgments
The expert technical work of the clinical virology and chemistry laboratory staff members and the data collection assistance of Maria Hernandez are gratefully acknowledged.
REFERENCES
- 1.Anonymous. 1999. Rapid diagnostic tests for influenza. Med. Lett. 41:121-122. [PubMed] [Google Scholar]
- 2.Centers for Disease Control and Prevention. 2001. Considerations for distinguishing influenza-like illness from inhalational anthrax. Morb. Mortal. Wkly. Rep. 50(44):984-986. [PubMed] [Google Scholar]
- 3.Chan, K. H., N. Maldeis, W. Pope, A. Yup, A. Ozinskas, J. Gill, W. H. Seto, K. F. Shortridge, and J. S. M. Peiris. 2002. Evaluation of the Directigen Flu A+B test for rapid diagnosis of influenza virus type A and B infections. J. Clin. Microbiol. 40:1675-1680. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Doing, K. M., M. A. Jerkofsky, E. G. Dow, and J. A. Jellison. 1998. Use of fluorescent-antibody staining of cytocentrifuge-prepared smears in combination with cell culture for direct detection of respiratory viruses. J. Clin. Microbiol. 36:2112-2114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Falsey, A. R., and E. E. Walsh. 2000. Respiratory syncytial virus infection in adults. Clin. Microbiol. Rev. 13:371-384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Glezen, W. P., S. B. Greenberg, R. L. Atmar, P. A. Piedra, and R. B. Couch. 2000. Impact of respiratory virus infections on persons with chronic underlying conditions. JAMA 283:499-505. [DOI] [PubMed] [Google Scholar]
- 7.Hindiyeh, M., C. Goulding, H. Morgan, B. Kenyon, J. Langer, L. Fox, G. Dean, D. Woolstenhulme, A. Turnbow, E. Billetdeaux, S. Shakib, C. Gordon, A. Powers, G. Vardeny, M. Johnson, L. Skodack-Jones, and K. Carroll. 2000. Evaluation of BioStar FLU OIA assay for rapid detection of influenza A and B viruses in respiratory specimens. J. Clin. Virol. 17:119-126. [DOI] [PubMed] [Google Scholar]
- 8.Hsiung, G. D. 1994. Orthomyxoviridae and paramyxoviridae, p. 141-157. In G. D. Hsiung, C. K. Y. Fong, and M. L. Landry (ed.), Hsiung's diagnostic virology, 4th ed. Yale University Press, New Haven, Conn.
- 9.Kaiser, L., M. S. Briones, and F. G. Hayden. 2001. Performance of virus isolation and Directigen Flu A to detect influenza A virus in experimental human infection. J. Clin. Virol. 14:191-197. [DOI] [PubMed] [Google Scholar]
- 10.Landry, M. L., D. Ferguson, and J. Wlochowski. 1997. Detection of herpes simplex virus in clinical specimens by cytospin-enhanced direct immunofluorescence J. Clin. Microbiol. 35:302-304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Landry, M. L., S. Cohen, and D. Ferguson. 2000. Impact of sample type on detection of influenza A virus by cytospin-enhanced immunofluorescence and membrane enzyme-linked immunosorbent assay. J. Clin. Microbiol. 38:429-430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Landry, M. L., and D. Ferguson. 2000. SimulFluor respiratory screen for rapid detection of multiple respiratory viruses in clinical specimens by immunofluorescence staining. J. Clin. Microbiol. 38:708-711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Marx, A., H. E. Greg, Jr., B. J. Marston, D. D. Erdman, R. F. Breimar, T. J. Torok, J. F. Plouffe, T. M. File, Jr., and L. J. Anderson. 1999. Parainfluenza virus infection among adults hospitalized for lower respiratory tract infection. Clin. Infect. Dis. 29:134-140. [DOI] [PubMed] [Google Scholar]
- 14.Sintchenko, V., G. L. Gilbert, E. Coiera, and D. Dwyer. 2002. Treat or test first? Decision analysis of empirical antiviral treatment of influenza virus infection versus treatment based on rapid test results. J. Clin. Virol. 25:15-21. [DOI] [PubMed] [Google Scholar]
- 15.Snacken, R. 2002. Pandemic planning. Vaccine 20(Suppl. 2):S88-S90. [DOI] [PubMed] [Google Scholar]
- 16.Steininger, C., M. Kundi, S. W. Aberle, J. H. Aberle, and T. Popow-Kraupp. 2002. Effectiveness of reverse transcription-PCR, virus isolation, and enzyme-linked immunosorbent assay for diagnosis of influenza A virus infection in different age groups. J. Clin. Microbiol. 40:2051-2056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Waner, J. L., S. J. Todd, H. Shalaby, P. Murphy, and L. V. Wall. 1991. Comparison of Directigen FLU-A with viral isolation and direct immunofluorescence for the rapid detection and identification of influenza A virus. J. Clin. Microbiol. 29:479-482. [DOI] [PMC free article] [PubMed] [Google Scholar]