Hendra Virus Infection
Queensland Health Guidelines for Public Health Units
|Full revision of guidelines
Revision post Tewantin outbreak (May 2010)
Full revision of guidelines
- Infectious agent
- Notification criteria
- Notification procedure
- Laboratory Aspects
- Reporting to NOCS
- Objectives of Surveillance
- Public health significance and occurrence
- Clinical features
- Mode of transmission
- Incubation periods
- Period of communicability
- Susceptibility and resistance
- Preventive measures
- Further actions
The agent is a paramyxovirus (previously described as equine morbillivirus) of the genus Henipavirus. The only other agent in this genus is Nipah virus.
Acute compatible illness. Clinical features in humans have included:
- self-limiting influenza-like illness (two cases)
- influenza-like illness complicated by severe pneumonic illness contributing to death (one case)
- aseptic meningitis with apparent recovery, then death from encephalitis 13 months later (one case)
- acute influenza-like illness followed by encephalitis at seroconversion, followed by recovery (one case) and death (two cases)
Exposure, within 21 days prior to onset of symptoms, to a horse with confirmed Hendra virus infection, or where heightened suspicion of Hendra virus infection exists on clinical and epidemiological grounds, as advised by Biosecurity Queensland.
Laboratory definitive evidence
Isolation of Hendra virus
Detection of Hendra virus by nucleic acid testing
Laboratory suggestive evidence
Detection of antibody to Hendra virus by microsphere immunoassay, confirmed by specific immunofluorescent assay
Detection of antibody to Hendra virus by virus neutralisation test
To notify relevant public health unit
(i) on receipt of request for examination by telephone or facsimile, and
(ii) upon confirmation of diagnosis by telephone.
To notify suspect cases in horses 2 to Biosecurity Queensland by telephone.
In turn, Biosecurity Queensland to notify confirmed animal cases, cases where heightened suspicion exists on clinical and epidemiological grounds, and any other cases where notification is appropriate (e.g. where human contacts of a suspect/exclusion case have been advised to contact their public health unit or where significant external interest is anticipated) to the Senior Director, Communicable Diseases Branch, Queensland Health by telephone.
Senior Director, Communicable Diseases Branch, Queensland Health to notify confirmed human cases to Biosecurity Queensland by telephone.
Human testing may be indicated when:
- There has been exposure to confirmed or suspected equine cases; or
- There is compatible human illness and testing is performed after consultation with an infectious diseases physician (IDP).
As of November 2011, tests available through Queensland Health Forensic and Scientific Services (FSS) include nucleic acid testing (NAT) and serology. Virus isolation is usually attempted on acute phase specimens, particularly when NAT is positive. Refer to the Queensland Health Clinical and Statewide Services intranet site for specimen collection, storage and transport advice. Advice as of November 2011 is included below.
Testing of human samples is also available, on request, from the Australian Animal Health Laboratory (AAHL) in Geelong. Tests available from AAHL include NAT, virus isolation and virus neutralisation test (VNT).
Nucleic acid testing
NAT can be performed on respiratory samples (deep nasal swab or nasopharyngeal aspirate), serum (preferable to blood), cerebrospinal fluid, urine and tissue samples.
Minimum CSF, serum, liquid respiratory or urine sample volume is 0.5ml. Tissue samples (minimum rice grain size) should be in viral transport medium, not fixed. Respiratory swab samples should be in viral transport medium. Chill all samples immediately upon collection (40C) and transport cool (40C), not frozen, within 24 hours.
FSS currently performs a screening microsphere immunoassay (MIA) test targeting antibodies against the glycoprotein, which if positive is followed by a confirmatory immunofluorescent antibody (IFA) test that can be quantified by dilution. If monoclonal antibodies have been used for prophylaxis or treatment they can be detected using a different MIA test targeting antibodies against the fusion protein. As with all serological tests, false positive results can occur. If MIA and IFA are positive in the absence of positive NAT, and there is any question as to whether this represents true infection, specimens may be referred to AAHL for VNT. Where serological testing is indicated, a baseline sample should ideally be collected as early as possible after exposure and stored for testing in parallel with subsequent samples. Minimum sample volume of CSF or serum is 0.2ml. Transport cool (40C) within 24 hours.
The earliest laboratory evidence of infection in recent human Hendra virus cases has been positive nucleic acid testing in blood, then urine, with serology becoming positive two to five days after onset of symptoms.
No cases of asymptomatic infection have been identified from extensive testing of human contacts associated with Hendra virus events up to November 2011.
For suspected human cases:
- As a minimum, collect blood and urine for NAT.
- Additional tests, depending on clinical picture may include deep nasal swab, nasopharyngeal aspirates and/or CSF for NAT. Tissue samples can be collected post-mortem (or by pre-mortem biopsy where relevant) for NAT.
- Serology should be considered whenever there is a high index of suspicion (e.g. high level exposure to a confirmed equine case) or where NAT could potentially be less reliable e.g. due to duration of symptoms greater than one week.
For human contacts of an equine case:
- If high exposure level (as per exposure assessment form in Appendix), collect baseline serology and store for testing in parallel with follow up samples at 3 and 6 weeks after last exposure.
- If medium or lower exposure level, assess if symptoms develop and test as for suspected cases above where indicated.
NB: FSS will conduct testing on all specimens where considered clinically indicated by the requesting medical practitioner. However, testing will be prioritised (routine or urgent) based on advice from the relevant public health unit in accordance with this guideline.
Reporting to NOCS
Report only confirmed human cases.
|A confirmed human case requires either:|
|1||Laboratory definitive evidence,|
|2||Laboratory suggestive evidence AND epidemiological evidence AND clinical evidence.|
|A highly suspect human case requires:|
Objectives of Surveillance
To identify equine and human cases so that appropriate and public health measures, including contact tracing, can occur.
To collect information about new human cases in order to broaden understanding of this high-profile emerging infectious disease.
Public Health Significance and Occurrence
As of November 2011 there have been seven confirmed human infections. All seven cases had a high level of exposure to respiratory secretions and/or other body fluids of horses subsequently diagnosed with Hendra virus infection, or presumed to have Hendra virus infection through review of clinical evidence in the absence of samples for laboratory testing.
Three human cases were exposed to tissues during autopsies on those horses. Two were probably exposed while performing nasal cavity lavage in the days immediately before the onset of clinical signs in the horse.
In Hendra (Brisbane) in 1994, a 49 year old horse trainer died after a fulminating pneumonic illness, while a 40 year old stable worker survived an influenza-like illness and apparently remains well. These cases occurred following a sudden outbreak of an acute respiratory syndrome among thoroughbred horses in a training complex. Both cases had extensive close contact with respiratory secretions and other body fluids of several very ill horses3. The incubation periods of the two human cases were estimated to be five and six days. The causal agent was a previously undescribed paramyxovirus. Surveillance of about 90 people identified no other human cases4.
In Mackay in 1995, a 36 year old horse trainer died from severe encephalitis caused by Hendra virus. Subsequent investigations indicated he had been exposed to Hendra virus 13 months earlier whilst assisting at autopsy of two horses, and he also had direct pre-mortem exposure to their respiratory secretions5. Of note, he was hospitalised with aseptic meningitis shortly after exposure to the horses, with recovery at the time. This is considered to have been a seroconversion illness. The vet who conducted the autopsy remained well.
About 25 km south of Cairns in late 2004, a veterinary practitioner in her 20s performed an autopsy on a horse subsequently presumed, based on clinical and epidemiological evidence, to have died of Hendra virus infection. The vet had also dealt with the horse in the last few minutes of its life and was extensively exposed to respiratory secretions and blood. About seven days later she developed a mild influenza-like illness and showed evidence of seroconversion to Hendra virus. She remained well six years later. Three other people who were also exposed remained well and did not seroconvert6.
In July 2008, a veterinary hospital at Redlands (Brisbane) was quarantined after four horses with neurological symptoms tested positive to Hendra virus. Another horse developed infection while quarantined. Two of the staff members caring for the horses each developed an acute influenza-like illness followed by encephalitis at seroconversion to Hendra virus; one of them subsequently died. One case was exposed during autopsy of a horse subsequently confirmed as infected with Hendra virus (possible 16 days incubation period) and both cases were exposed while performing nasal lavage on another horse in the days immediately before it developed signs (possible incubation periods of nine and 11 days)7.
In August 2009 at an equine nursery near Rockhampton, a 55 year old veterinary practitioner performed a nasal endoscopy on a horse that was subsequently presumptively diagnosed with Hendra virus infection8. The procedure was performed without respiratory protection or gloves whilst the horse had a fever, respiratory distress, and visible petechiae and ecchymoses of the respiratory tract. The vet was given a five-day course of ribavirin from day 14 post-exposure but developed a flu-like illness 21 days post exposure followed by encephalitis and death. Three other people with high level exposure to the respiratory secretions and blood of this horse (and a second horse involved in the same outbreak also confirmed to have Hendra virus infection) remained well and did not seroconvert.
Post-mortem results on the Rockhampton human case are pending, and it is likely they will be published with a review of post-mortem results from previous human Hendra virus cases.
Hendra infection in horses
In horses, Hendra virus has a clear predisposition for targeting endothelial cells of blood vessels, with clinical signs dependent on the sequence in which organs are affected. Horses to date have shown a predominance of either respiratory or neurological manifestations or a mix of these; some horses have shown evidence of multi-organ involvement. The clinical signs, particularly of early Hendra disease, may be non-specific, although progression from onset to death is typically rapid, within 24-48 hours.
Milder infections have been identified in horses being monitored in the course of an outbreak; these horses are typically second or third generation cases. From figures to date, approximately 75% of infected horses can be expected to die from Hendra virus disease, and approximately 25% will survive. Those that survive often have mild signs and seroconvert during the recovery period. The current AusVetPlan for Hendra9 prescribes euthanasia for these seropositive horses, as they may remain infectious or relapse. Biosecurity Queensland's Guidelines for veterinarians handling potential Hendra virus infection in horses10 outline current advice on the disease in horses.
The following table outlines the Hendra virus infection events that have occurred in Australia. As of November 2011 all events have occurred in Queensland or NSW. All 7 human cases have occurred in Queensland.
Summary of Hendra virus infection events with numbers of equine cases, human cases, and human deaths, 1994-2011 (to November 2011)
|Murwillumbah||New South Wales||October||2006||1||0||0|
|Wollonbar||New South Wales||July||2011||2||0||0|
|Macksville||New South Wales||July||2011||1||0||0|
|Lismore||New South Wales||July||2011||2||0||0|
|Mullumbimby||New South Wales||July||2011||1||0||0|
|Ballina||New South Wales||August||2011||1||0||0|
|Mullumbimby||New South Wales||August||2011||1||0||0|
|South Ballina||New South Wales||August||2011||2||0||0|
|North Ballina||New South Wales||August||2011||1||0||0|
|Flying foxes||Uncertain, if any.|
|Horses||Acute onset of illness with fever and rapid deterioration with respiratory and/or neurological signs. There is a high case fatality rate (75%) and death is typically rapid (within 24-48 hours).|
|Humans||The full spectrum of human clinical infection is unknown. Initial influenza-like illness, followed by recovery, or respiratory or neurological complications with recovery or death (Refer to Clinical Descriptions).|
Flying foxes (pteropid bats) are the natural host of Hendra virus. Evidence of virus presence has been demonstrated in all four species of pteropid bats in Australia.
Antibody to Hendra virus has been found in 20-50% of flying foxes in mainland Australian populations. Flying foxes with evidence of exposure to Hendra virus have been found from Darwin across to the east coast and south to Melbourne11.
Widespread testing involving 46 species of animals and arthropods since 1994 has not shown the natural presence of the virus in any species other than flying foxes.
Mode of Transmission
|Bat-to-horse||Rare (32 documented events from 1994 through to November 2011). Possibly via infected bat urine or reproductive products such as aborted bat foetuses. At the time of the original outbreak at Hendra in 1994, testing of nearly 2,000 horses across the state, including 906 within one kilometre of the quarantined properties identified no seropositive horses apart from those on the quarantined premises12.|
|Bat-to-person||No evidence to date that this occurs. A study of 128 bat carers, three-quarters of whom had daily contact with bats and three quarters of whom reported a history of bat bites, identified that none had antibodies to Hendra virus13.|
|Horse-to-horse||Has been more efficient in a stabled situation, with spread between multiple horses occurring in all stabled situations to date - Hendra (1994), Redlands (2008) and Cawarral (2009). It is possible that short-term survival of Hendra virus on environmental surfaces and inadvertent human-assisted fomite transmission may contribute to more efficient transmission in stabled situations.
Associated with the 1994 outbreak in Hendra, Queensland, 107 horses were tested from 13 quarantined premises with seven testing positive and 13 highly likely to be cases on clinical/epidemiological grounds. Eighteen of the 20 equine cases were from the index Hendra property; the remaining two were from the immediately adjacent property.
|Horse-to-person||All human infections have been acquired through high level exposure to the body fluids (especially respiratory secretions and blood) of Hendra virus infected horses. No cases have been documented in people with medium or lower exposure levels, including anyone appropriately using personal protective equipment, and not all people with high exposures have become infected. On the evidence available, the most likely mode of transmission is via substantial direct exposure of mucous membranes (or non-intact skin) to respiratory secretions (including large droplets) or blood from an infected horse. Indirect exposure to respiratory secretions or blood, and direct or indirect exposure to other body fluids, may contribute to overall exposure. Current evidence does not support airborne exposure as a significant mode of transmission.
Note: An AAHL experiment demonstrated that a horse can potentially excrete Hendra viral RNA in nasal secretions for at least 2 days before showing clinical signs14.Virus could not be isolated, presumably due to low levels of excretion.
|Person-to-person||No evidence to date that this occurs. Serological testing in 169 health care and four household contacts of the first three human cases found no individuals with antibodies to Hendra virus15. Testing of contacts of subsequent human cases has also produced uniformly negative results.|
|Person-to-horse||No evidence to date that this occurs.|
|Experimental||Disease has been experimentally induced in cats, ferrets, hamsters, guinea pigs, horses and pigs. Experimentally infected horses can excrete viral RNA in nasal secretions before onset of clinical signs (as above).|
|Other||A single dog developed antibodies to Hendra virus on a property where three horses developed Hendra virus infection in July 2011. Although the source of exposure for the dog cannot be definitively ascertained, horse-to-dog transmission is the most plausible scenario given the dog had potential opportunity for exposure to infected horses. There is no evidence to date that bat-to-dog or dog-to-person or dog-to-horse transmission occurs.|
Current limited evidence suggests the incubation period in humans is usually 5 – 21 days. The incubation period in horses appears to be 5 – 16 days (but could be up to 31 days)16.
Period of Communicability
Unknown, no evidence of person-to-person transmission to date.
For public health purposes, horses should be considered potentially infectious from 72 hours prior to onset of symptoms until death (by disease or euthanasia) and safe disposal of the carcase has been completed. Careful assessment is required of how closely the horse was observed for the onset of illness. It is unknown whether horses recovered from illness remain infectious, can later become infectious again, or are capable of permanently clearing the virus.
Susceptibility and Resistance
Commence investigation immediately on notification of a confirmed human or equine case, or where notified by Biosecurity Queensland of heightened suspicion of infection in a horse on clinical and epidemiological grounds.
On notification of a confirmed equine case, or where Biosecurity Queensland has heightened suspicion on clinical and epidemiological grounds:
On notification of a confirmed human case, or where a public health unit considers there is heightened suspicion on clinical and epidemiological grounds:
Biosecurity Queensland is responsible for investigating possible equine sources of human infection, if this has not already occurred. This may identify other humans requiring separate follow-up as contacts of equine cases.
When an equine case is confirmed, or if Biosecurity Queensland intends to conduct a site visit for a horse where heightened suspicion exists, the relevant public health unit (PHU) will liaise with the Communicable Diseases Branch to establish an incident management team (IMT), in accordance with the requirements for a significant public health (level 2) incident. The local public health medical officer (PHMO) will usually lead the local response. Where IDPs are involved they should be invited onto the IMT. Unless the incident is confined to a small property and there are very few or no potential human exposures, the public health actions required will be resource intensive and a second PHMO should be co-opted to support the local response and liaise with the IMT. If possible, routine communicable disease control activities at the local site should be deferred or referred to another PHU until the incident is more manageable. This may take several weeks.
Close liaison with Biosecurity Queensland will be necessary and regular joint incident management teleconferences should be held. Workplace Health and Safety Queensland should also be involved early.
- Standard, contact and droplet precautions for human cases during acute illness. Additional precautions may be ordered at the discretion of the treating IDP.
- Human cases should avoid close contact with animals during acute illness.
- Confirmed human cases should never subsequently donate blood or any other tissue, even if they fully recover.
Responsibility of the treating practitioner. Advice should be sought from an IDP.
Contact Tracing: Yes
Contact of equine case:
People who have had direct or indirect exposure of skin or mucous membranes to secretions or body fluids of a horse determined by Biosecurity Queensland to be a confirmed case of Hendra virus infection, or of a horse where heightened suspicion of infection exists on clinical/epidemiological grounds, as advised by Biosecurity Queensland.
NB: For public health purposes, horses should be considered potentially infectious from 72 hours prior to onset of clinical signs until death (by disease or euthanasia) and safe disposal of the carcase.
Contact of human:
People who have had close contact with a symptomatic confirmed or highly suspect human case (including household or household-like contacts, sexual partners, and anyone with direct or indirect exposure of skin or mucous membranes to their secretions or body fluids).
Investigation: Contacts of equine and/or human cases
Exposure assessments of contacts are required in relation to all confirmed equine cases of Hendra virus infection, and where heightened suspicion of infection in a horse exists on clinical/epidemiological grounds as advised by Biosecurity Queensland.
Exposure assessment of contacts of confirmed human cases is also recommended – although risk is generally likely to be negligible, assessment and reassurance will often be of benefit. Refer to the Queensland Health Hendra Virus (Confirmed Animal Case) Exposure Assessment Form.
Information about Hendra virus, and the response process should test results be positive, should be provided to the closest contacts, e.g. household/sexual contacts, of highly suspect human cases while awaiting results.
The urgent priorities are to obtain a spreadsheet of all people who may have been in contact with the infectious horse/s (it is best to refer to horses by their common name/s to minimise confusion), then to:
- Assess exposure and current health status
- If high exposure to body fluids of a confirmed equine case, liaise with an IDP as soon as practicablefor consideration of post-exposure prophylaxis
- If assessed as medium exposure, discuss as soon as practicable with other appropriately experienced public health practitioners and IDP/s to reach consensus on exposure assessment
- Refer any symptomatic people to appropriate care – their GP for an incidental illness or an IDP for an illness clinically compatible with Hendra.
- Provide information about Hendra virus
- Counsel about risk
- Provide advice about testing recommendations (see Laboratory Aspects section above)
- Provide advice about self-monitoring of their health, and to seek early medical advice if they develop fever or respiratory or neurological symptoms within three weeks of exposure
NB. If child contacts are identified, investigation and management may need to be tailored accordingly.
Follow up all identified potential contacts to determine the level and type of contact. Assessment of exposure should review activities with any equine cases from 72 hours prior to the onset of illness in the horse to death and safe disposal of the carcase, and with symptomatic human cases. Identified potential contacts, even if exposure is assessed to be nil, low or negligible, should be given PHU contact details in case their recollections change. Consideration should be given to ways of finding unidentified contacts (e.g. passersby who may have fed the horse across a fence) where significant opportunities for exposure may exist.
The urgency of response to illness in contacts depends on the assessed level of exposure and clinical picture; clinically compatible illness (respiratory or neurological symptoms) in a person with high exposure should be urgently assessed and referred to an IDP.
The public health response should include strong recommendations that people involved in the incident access their usual health care practitioner to assess need for ongoing physical and psychological support, medical or WorkCover certification and referral to other services as necessary (see below). Provide the GP with appropriate information and resources and PHU contact details.
Provide written information about the disease to contacts.
Involve a public affairs spokesperson early in the IMT and obtain advice on clear public messages.
If an equine case is confirmed, it may be appropriate for PHU officers to visit the property to provide information and support to key people (e.g. owners and managers), and to assist Biosecurity Queensland staff in providing information to the local community. Any site visit should complement and not delay the full public health response, which focuses on the timely assessment of exposure and current health status of all people who may have been exposed to infected horses, including animal health practitioners, many of whom may not be available on site.
It is likely that Biosecurity Queensland will arrange public information sessions in the area affected by the event. Assistance from Queensland Health may be expected and should be resourced appropriately. Consider the most appropriate personnel to attend these functions. Opportunities can be taken to address some of the many misconceptions surrounding Hendra virus infection.
A Hendra virus event can be extremely stressful for a horse/property owner, vet, or person in contact with an infected horse. Owners are likely to see Biosecurity Queensland staff perform elaborate infection control measures with maximum ‘space suit’ type PPE, and possibly euthanasia and post-mortem examinations of their horses and to have their horses isolated/quarantined for several weeks.
Therefore, contacts and other closely involved people, such as family members and owners, may require repeated reassurance and information even when they have had negligible exposure.
As any existing mental health issues or personal stresses might be exacerbated by such an event, contacts and closely involved persons should be promptly connected to appropriate counselling, or psychological care, preferably through their GP.
People with high exposures should be referred promptly to an IDP for consideration of post-exposure prophylaxis (PEP). No drug or product is of proven benefit in humans, although monoclonal antibodies show the most promise as of November 2011. Monoclonal antibodies have been shown to protect ferrets from serious disease when given 10 hours after exposure to Nipah virus17; and when given 10-12 hours after exposure to Hendra virus18. They have also been shown to protect African green monkeys from severe fatal infection when given 72 hours after exposure to a lethal dose of Hendra virus19. As the evidence base to inform the use of PEP is rapidly evolving, the IDP involved should consult with other IDPs who have experience in Hendra PEP and knowledge of the current evidence base. Wherever possible, an expert panel of PHMOs and IDP/s with appropriate experience should be convened as part of the incident management process to review all contacts identified as high and medium exposures and advise on provision of monoclonal antibody and logistic issues. For contacts receiving monoclonal antibodies, the treating IDP will be responsible for symptom and serological monitoring and communication of results until there is a formal handback to the PHU.
People who have had contact with body fluids of an infected horse should not donate blood or any other tissue until cleared by absence of illness over three weeks since last exposure (and negative test results where indicated).
Biosecurity Queensland staff involved in outbreak response:
Risk management for Biosecurity Queensland staff involved in the outbreak response is a workplace health and safety issue. As all Biosecurity Queensland staff involved in the outbreak response should have had appropriate training and use PPE accordingly, routine assessment of Biosecurity Queensland staff involved in the response is not indicated. Assessment can be undertaken if there is a significant breach of PPE when handling an infected horse or if there is a significant level of personal concern about potential exposure.
Other control measures:
- Veterinary measures: Biosecurity Queensland will require isolation and euthanasia of infected horses with appropriate disposal. Quarantine animals from infected farms.
- In human health care facilities: environmental cleaning procedures in accordance with Queensland Health Infection Control Guidelines.
- Veterinary practitioners and staff handling confirmed or suspected equine cases or conducting autopsies on the carcass of a confirmed or suspected case should wear appropriate personal protective equipment (PPE) and take appropriate preventive hygiene measures (refer to the Biosecurity Queensland Guidelines for veterinarians handling potential Hendra virus infection in horses).
- Horse owners or carers handling any sick horses should minimise unnecessary human contact and take appropriate preventive hygiene measures (refer to the Workplace Health and Safety Queensland Hendra virus - Information for horse properties and other horse related businesses)
Relevant Queensland Health staff should participate in internal debriefing and interagency debriefing with Biosecurity Queensland (and Workplace Health and Safety Queensland where appropriate) after every Hendra virus outbreak. This guideline should then be reviewed and updated.
Hendra Virus (Confirmed Animal Case) Exposure Assessment Form
|1||Hendra virus infection is not clinically notifiable under the Public Health Act and Regulation 2005, however clinical/ epidemiological criteria are required for a confirmed case where only laboratory suggestive evidence is available|
|2||Case definitions of 'suspect' and 'exclusion' equine cases are available at DPI - Guidelines for veterinarians handling potential Hendra virus infection in horses|
|3||Selvey et al. MJA 1995:162:642.|
|4||Selvey et al. MJA 1995:162:642.|
|5||Personal communication, Simon Bewg, Biosecurity Queensland, September 2006 and O'Sullivan et al. Lancet 1997;349:93|
|6||Hanna et al. MJA 2006;185:562|
|7||Playford et al. EID 2010;16(2):219-223.|
|8||The presumptive diagnosis was based on both clinical signs and an epidemiological link with a laboratory-confirmed equine case.|
|9||Available at Animal Health Australia|
|10||Available at Department of Primary Industries and Fisheries|
|11||See the Australian Biosecurity CRC for Emerging Infectious Disease for the latest review on Hendra virus.|
|12||Murray et al. EID 1995;1:31-33.|
|13||Selvey et al. CDI 1996;20:477.|
|14||Middleton. Initial experimental characterisation of HeV (Redland Bay 2008) infection in horses. CSIRO, 2009.|
|15||McCormack et al. J Infect 1999;38:22.|
|16||Personal communication, Steven Donohue, Queensland Health, and Hume Field, Biosecurity Queensland, November 2009. In the Bowen 2009 outbreak, the incubation period in the second horse may have been 31 days.|
|17||Bossart et al. PLoS Pathogens 2009:5(10);1-11.|
|18||Personal communication, Deborah Middleton, CSIRO, 20 September 2011.|
|19||Bossart et al. Sci Transl Med 2011;105:1-8.|