Consensus Date: 28 June 2010
1 PHLN Summary laboratory definition
1.1.1 Definitive Criteria
- Detection of influenza virus by nucleic acid testing (NAT) from appropriate respiratory tract specimens; or
- Isolation of influenza virus by culture from appropriate respiratory tract specimens; or
- Detection of influenza antigen using detection by a properly validated influenza virus antigen assay from appropriate respiratory tract specimens; or
- Seroconversion or a fourfold or greater rise in antibody titre to influenza virus
1.1.2 Suggestive Criteria
- a single high influenza virus-specific antibody titre
- detection of influenza virus-specific IgM by immunofluorescence
1.1.3 Special Considerations / Guide for Use
- Results of ‘Point of care’ (POC) tests for influenza antigens should be treated with caution due to their currently relatively low sensitivity. Further testing should be sought if influenza is suspected in the presence of a negative POC test result.
- All isolates of influenza should be typed (influenza A/B) and subtyped (A/H1, A/H3, pandemic (H1N1) 2009, etc).
1.1.4 Links to related documents Pandemic influenza
Top of Page
Influenza in humans is an acute, usually self-limited, febrile respiratory illness caused by infection with influenza virus type A or B that in temperate climates occurs in outbreaks of varying severity almost every winter. Influenza infection may occur year round in tropical climates, typically with no epidemic focus. In pandemics, outbreaks may be seen outside the usual influenza season.
Influenza viruses are enveloped viruses of the family Orthomyxoviridae and contain segmented negative sense single-stranded RNA genomes. The influenza viruses A, B and C can be distinguished on the basis of antigenic differences between their nucleocapsid and matrix proteins; influenza A viruses can be further subtyped on the antigenic nature of their haemagglutinin and neuraminidase glycoproteins. They replicate in the columnar epithelial cells of the respiratory tract, and the primary mode of transmission is via respiratory droplets.
Influenza C causes very mild illness without the clinical or public health implications of influenza A & B. It also does not undergo antigenic shift or drift. Standard laboratory tests do not detect influenza C. As only a few reference laboratories carry out influenza C testing, it is not included in this document.
In Australia, seasonal influenza is a notifiable disease by laboratory confirmation only. Pandemic influenza is notifiable in Australia and New Zealand. Human cases of highly pathogenic avian influenza (HPAI) are subject to quarantine regulations Australia-wide. Influenza laboratory data are currently gathered from Lab-VISE, a number of GP surveillance systems, some jurisdictional Departments of Health and laboratory reporting to the WHO Collaborating Centre (WHOCC) for Reference and Research on Influenza in Melbourne. Surveillance programs for influenza are based upon laboratory, clinical or a combination of laboratory and clinical reporting schemes. Clinically based surveillance programs eg. Australian Sentinel Practice Research Network (ASPREN), WHO, the NZ national surveillance program use various definitions of influenza-like illness (ILI),. An example of a case definition of ILI would be the sudden onset (<24 hours) of three of fever, rigors or chills, myalgia, cough, and prostration and weakness (1,2).
Most laboratories performing influenza virus culture send isolates and clinical samples to the WHOCC in Melbourne either directly, or via one of the three WHO National Influenza Centres (NIC). These isolates from clinical samples are used for the selection and manufacture of the following year's influenza vaccine, as well as monitoring antigenic and genetic changes and antiviral resistance.
This document outlines methods for the laboratory diagnosis of influenza to confirm the diagnosis of clinical ILI. Accurate laboratory and clinical diagnosis (or surveillance) of influenza is required for the annual winter epidemics (influenza activity is concentrated between May 1 and September 30 in southern Australia and New Zealand, and throughout the year in northern Australia), occasional imported or other cases of influenza, and potential pandemic or other novel influenza viruses. The role of laboratory surveillance in influenza pandemic planning is discussed in the Australian Health Management Plan for Pandemic Influenza (AHMPPI, released June 2008 at http://www.health.gov.au
Top of Page
3 Laboratory Diagnosis
3.1 Clinical specimens
3.1.1 Antigen detection
Nasopharyngeal washes or aspirates (NPA) are the best sample because they contain the highest number of respiratory epithelial cells. They are usually only available from children because of the difficulty in collecting NPA from adults. Alternately, swabs collected from the throat and from the anterior ends of the nasal turbinates can be combined. This is better than separate swabs from either site. Isolated throat swabs or throat gargles are less useful as the majority of cells captured are squamous epithelial cells. Nasal washes can also be used. Sputum is a poor sample for antigen detection or virus isolation due to bacterial contamination and the presence of mucous. In practice, NPA specimens are usually available from young children, and combined nasal and throat swabs are available from adults. Lower respiratory tract specimens (eg. bronchoalveolar lavage fluid) are indicated where possible if lower respiratory involvement is clinically apparent. Appropriate personal protective equipment should be worn during sample collection. Ideally, all samples should be collected within 96 hours of disease onset.
Swabs should be cotton, rayon or dacron-tipped, plastic-coated or aluminium shafted. They either contain their own viral transport media (VTM), Universal Transport Medium (UTM) or can be placed into a vial of VTM/UTM immediately after collection. Other specimens should be placed in a sterile container. Specimens should be stored and transported at 4°C (according to the appropriate IATA standards) or, if they cannot be processed within 72 hours, they should be frozen at –70°C. Samples should never be frozen at –20°C.
If antigen detection tests are positive, then the original clinical sample, if suitable, should be sent for virus culture (at least in a proportion of cases). Alternatively, a further sample should be obtained which is suitable for culture.
3.1.2 Virus Culture
This may be performed on the same samples as those used for antigen or NAT detection. Samples that have dried out or leaked are not suitable for viral culture. Other respiratory specimens suitable for culture include endotracheal aspirates, bronchoalveolar lavage fluid and lung biopsy tissue.
Laboratories undertaking influenza cultures should keep an aliquot of the original clinical sample at –70°C to allow re-inoculation should novel strains of influenza be isolated, or if strains are needed for vaccine development. They should send any influenza isolates to their local WHO laboratory or public health laboratory.
3.1.3 Nucleic acid testing (NAT)
This may be performed on the same samples as those used for antigen detection and virus isolation. It may also be performed on sputum, although this is not a preferred sample. Ideally samples for NAT should also be collected within 96 hours of onset of illness, but later samples may still contain detectable nucleic acid. Samples should be stored and transported in the same way as those for isolation as culture from samples may also be required. However, if samples have not been properly transported and stored, they may still be suitable for NAT. Samples which are positive by NAT should be cultured if possible (at least in a proportion of cases).
3.1.4 Serum samples
Acute (within 7-10 days of ILI onset) and convalescent phase serum (14–21 days after ILI onset) should be collected. The sensitivity of laboratory diagnosis is increased by using a combination of virus detection early in the illness and serology on acute and convalescent blood samples.
Top of Page
3.2 Detection of influenza virus antigen or genome
Antigen detection assays are typically directed towards circulating strains. They are less sensitive in the detection of novel strains, eg. new pandemic strains or those of different H (haemagglutinin) subtypes such as A/H5N1 (HPAI).
3.2.1 Antigen detection
This is currently most often done using monoclonal antibody (Mab)-based assays, with detection of the product by direct fluorescence (DFA) or enzyme immunoassay (EIA), using fluorescein-labelled secondary antibodies. These can be done on the same clinical specimens used for culture.
These assays have a sensitivity of 50–90% compared to isolation (Table 22.3, 2,4), take from 2–20 hours to perform, and are usually combined with other Mabs to the other common viral respiratory pathogens, such as parainfluenza types 1–3, adenovirus, RSV and human metapneumovirus.
3.2.2 Nucleic acid testing (NAT)
Influenza RNA in samples can be detected by NAT such as reverse transcriptase PCR (RT-PCR) or real time RT-PCR assays, or other related assays. They are the most sensitive test for influenza detection and also have extremely high specificity and a more rapid turnaround time than virus culture (3,4). NAT increases detection rates of influenza in clinical specimens compared with virus culture. They are more tolerant of samples that have been improperly stored, transported, or excessively delayed, as NAT may detect non-viable virus. The sensitivity (depending on the primers used) is ~80-100% compared with other direct detection tests, and the specificity approaches 100%. Primers can be combined in a multiplex PCR to detect other respiratory pathogens. NAT primers should be reviewed annually to ensure that influenza virus genetic drift has not affected assay sensitivity.
There are two major classes of influenza NAT
- Those directed at targets that are common to a specific influenza type, i.e. targets (usually matrix or nucleoprotein genes) common to all influenza A strains or to all influenza B strains.
- Those that are specific to the influenza A subtypes, i.e. primers are directed at targets in the haemagglutinin gene that are specific for H1, H3, pandemic (H1N1) 2009, H5 etc.
Nucleic acid testing is the test of choice for definitive identification of human infection with influenza A H5N1 strains (HPAI) currently circulating overseas and is the preferred test for pandemic (H1N1) 2009 (5).
Virus isolation for antigenic characterisation should be also be performed, where possible, on suitable NAT positive samples.
Laboratories performing NAT detection of influenza are encouraged to participate in recognised testing quality assurance (QA) Programs (e.g. RCPA, WHO) for A/H5N1, pandemic (H1N1) 2009 and seasonal influenza viruses.
Top of Page
3.3 Detection of the organism - virus isolation
Isolation of influenza A or B virus in cell lines such as Madin-Darby canine kidney (MDCK) or primary monkey kidney cells provides a definitive diagnosis of influenza infection, though it is generally less sensitive than NAT and requires the virus to remain viable during storage and transport. However, cell culture does provide virus for the more detailed antigenic analysis needed for strain identification and to assist vaccine selection, as well as the potential capacity to detect new influenza types that may be missed by other methods. Therefore, a reasonable proportion of influenza isolates need to be obtained each year from each jurisdiction (100-200, depending on the jurisdiction), and cell cultures should be used more widely when new pandemic or seasonal strains are beginning to circulate. It is important that appropriate biosafety guidelines are followed.
3.3.1 Conventional culture and typing
Influenza viruses are usually isolated using trypsinised MDCK cells. Primary monkey kidney cells are difficult to obtain and problematic to work with, so they are now rarely used. Influenza virus growth is suggested by haemadsorption or observation of typical cytopathic effect at 4–5 days. Confirmation is usually achieved by IFA using influenza type or subtype-specific MAbs or by NAT. Further strain typing can be carried out using the more time-consuming technique of haemagglutination inhibition (HAI).
All isolates should have preliminary typing and subtyping as influenza A/H1 or H3 (or other), or influenza B, as quickly as possible, either in the laboratory carrying out the culture or in a reference laboratory. All influenza isolates (including those that fail to type) should also be referred to the WHOCC in Melbourne directly, via a local WHO NIC or local reference laboratory. Where a new or significantly different strain is suspected, then isolates should be referred urgently to the WHOCC.
Traditional influenza isolation is undertaken by inoculating cells lining the allantoic cavity of embryonated chicken eggs, and this technique is still used for vaccine generation. This procedure is routinely performed using original clinical samples at the WHOCC in Melbourne.
Drug susceptibility testing using genotypic or phenotypic techniques is available at the WHOCC and in some reference laboratories, but is not currently part of routine testing. Antiviral susceptibility is monitored internationally via the WHO network (6).
3.3.2 Rapid culture assays
The time required for cell culture identification of influenza virus can be reduced to 1–3 days using shell-vial or multiwell plates (typically using MDCK cells) that can be stained after 48 hours culture using commercially available MAbs. Sensitivity varies from 56–100% compared with conventional culture (3,4), and is ~85% compared to PCR. Specificity of culture is 100%. Rapid culture positive samples should be subsequently recultured in the testing or a reference laboratory to provide an isolate for antigenic analyses as part of surveillance and vaccine development.
There is currently no quality assurance program specifically for influenza isolation or rapid antigen detection.
3.4 Rapid antigen tests or "point of care" tests
Rapid antigen detection, also called "point of care" (POC) tests, are mostly based on immunochromatography using antibodies directed to conserved influenza proteins. They have been available for a number of years and may assist in the early management of suspected influenza cases. These may have a role in laboratories unable to perform virus isolation, in outbreaks, for doctors without reasonable laboratory access, or to guide the rapid use of antiviral agents (4). They are intended for use at the bedside and the specificity has been shown to be generally high (90-95%). They are consistently less sensitive than NAT, or culture and IFA antigen detection tests. If a reliable negative result is required, or if the accuracy of a positive result is critical, then these samples should also be tested by NAT or another method. The sensitivity is lower for pandemic (H1N1) 2009 (or A/H5N1) influenza (5).
The rapid antigen tests have sensitivities between 60–85% and specificities between 60–95%, though most of the current POC tests perform at the upper end of these ranges, at least for specificity.
Top of Page
Detection of influenza-specific antibodies allows a retrospective diagnosis of influenza infection and is also a useful surveillance tool. As influenza is often a reinfection with pre-existing partial immunity present, recent infection can only be reliably diagnosed by demonstrating a significant (four-fold) rise in influenza-specific antibody titres. Single samples with a high titre are less reliable as they may reflect past infection. However, a single high titre is suggestive of probable recent infection when the patient has had a consistent ILI occurring during the influenza season (3,4). This is because neither vaccination nor infection in the previous season produces high antibody titres. Serological methods include complement fixation (CF), haemagglutination inhibition (HAI), single radial haemolysis (SRH), neutralisation (Nt) and enzyme immunoassays (EIA). Sensitivity varies from 80–50% (EIA>HAI>CF). Enzyme immunoassay-based IgM (and IgA) assays have not proved useful in routine laboratory diagnosis as most influenza infections are reinfections.
CF tests detect antibodies directed at the influenza nucleoprotein and can distinguish between influenza A (but not between subtypes) and B. A fourfold rise is definitive evidence of acute infection. HAI and Nt are the gold standard assays that allow subtype determination. There are generally accepted criteria that correlate with postvaccine "immunity". The technical difficulties associated with these tests mean that they are not in widespread diagnostic use. Also, they may not detect new influenza A subtypes due to antibody specificity. Like CF, a four-fold rise in influenza-specific antibodies is the most reliable indicator of recent infection. Single high titres are suggestive of recent infection in the appropriate clinical situation. Due to differences in how these tests are performed in different laboratories, the definition of "positive titre" will vary between laboratories.
SRH is rarely used although technically easier than HAI, as it requires high antigen concentrations, and the correlation between zone sizes and antibody titres measured by other methods is not well established. EIA tests have not been widely used and the criteria for immunity and for significant antibody levels are not established.
Laboratories participate in the RCPA Serology QA Program (http://www.rcpaqap.com.au/serology
3.6 Quality Assurance
Test-specific QA considerations are described under the various testing procedures.
3.7 Diagnosis of Human Cases of Influenza A H5N1 (Avian Influenza)
Where there is a suspected case of human infection with a highly pathogenic avian influenza (such as influenza A/H5N1 or another novel strain) or in the event of pandemic influenza, the laboratory case definition will be similar to that for seasonal influenza, as well as including the appropriate exposure history as defined in the AHMPPI (http://www.health.gov.au
Specimen collection for suspected A/H5N1 infection will be similar to that for seasonal influenza (combined nose and throat swabs are the recommended sample) except that additional non-respiratory specimens (e.g. serum, faeces, cerebrospinal fluid) can be useful in diagnosing influenza A/H5N1. Where possible, safely collected lower respiratory tract samples are best for A/H5N1 diagnosis. At a minimum, laboratories must meet basic PC2 standards and use PC3 work practices to handle specimens that are suspected to contain pandemic influenza.
The diagnostic test of choice is NAT with both influenza A-specific and novel strain-specific primers. Appropriately equipped laboratories may also perform virus isolation under PC3 conditions. The reliability of POC tests for detection of A/H5N1 is not well established.
3.8 Diagnosis of Pandemic Influenza (H1N1) 2009
In the event of pandemic influenza, such as with the pandemic (H1N1) 2009 influenza A virus, the laboratory case definition will be similar to that for seasonal influenza, as well as including the appropriate exposure history as defined in the AHMPPI (http://www.health.gov.au
Specimen collection for human pandemic infection will be similar to that for seasonal influenza (combined nose and throat swabs are the recommended sample). Lower respiratory tract samples are better for diagnosis of pandemic (A H1N1) 2009 virus so that where possible, and especially in patients with lower respiratory tract infection, they should be collected safely. At a minimum, laboratories must meet basic PC2 standards and work practices to handle specimens that may contain pandemic influenza.
The diagnostic test of choice is NAT targeted at both influenza A-specific sequences and at pandemic strain-specific sequences as primers/probes. Appropriately equipped laboratories may also perform small scale virus isolation under PC2 with PC3 techniques, or PC3 conditions. Most current POC tests are not regarded as reliable for detection of pandemic (H1N1) 2009, as they are less sensitive than other methods and mostly do not differentiate pandemic (H1N1) from other influenza A viruses (5). Testing strategies will vary depending on the phase of the pandemic and the severity of the clinical illness. Further details are available on the AHMPPI and WHO websites (7,8,9).
Top of Page
4 Agreed Typing & Subtyping Methods
4.1 Laboratory Nomenclature for National Database Dictionary
4.1.1 Organism Name(s) List
Orthomyxoviridae, Influenza A, Influenza B, Influenza C
4.1.2 Typing/Subtyping Nomenclature List(s)
Types: influenza A, or influenza B (influenza C is rarely tested diagnostically).
Influenza A Subtypes: Currently there are 3 influenza A viruses that are or have recently circulated in humans. These are A/H3N2, A/H1N1 and pandemic A/(H1N1) 2009 also commonly referred to as "swine influenza". The H refers to the haemagglutinin type of which there are 16, and the N refers to the neuraminidase type of which there are 9.
4.2 SNOMED CT concepts
|SNOMED CT concept||SNOMED CT Code|
|Influenza-like illness (clinical finding)||95891005|
|Influenza A virus (organism)||407479009|
|Influenza B virus (organism)||407480007|
|Influenza C virus (organism)||4007482004|
- Joseph C. (1995) Virological surveillance of influenza in England and Wales: results of a two year pilot study 1993/4 and 1994/5. Communicable Disease Review 5(10): R141-R145.
- Thursky K et al. (1993) Working towards a simple case definition for influenza surveillance. Journal of Clinical Virology
- 27 (2): 170-179.
- Zambon M. (1998) Laboratory Diagnosis of Influenza. In: Textbook of Influenza. Eds. Nicholson K.G, Webster R.G, Hay A.J. Ch.22: 291-313.
- Playford EG and Dwyer DE. (2002) Laboratory diagnosis of influenza virus infection. Pathology 34:115-125.
- Kok, J. et al. (2010) Comparison of a rapid antigen test with nucleic acid testing during cocirculation of pandemic influenza A/H1N1 09 and seasonal influenza A/H3N2. Journal of Clinical Microbiology 48:290-291.
- Hayden F.G. (2006) Antiviral resistance in influenza viruses – implications for management and pandemic response. New England Journal of Medicine 2006; 354 :785-788.
- Australian Health Management Plan for Pandemic Influenza (released June 2006, www.health.gov.au)