Influenza A Virus
LEVELS: Likely to occur; Mild; Extraordinary biosecurity required; Laboratory-dependent; Moderate; Negligible; Little; Available but inconsistent
| Register id | influenza_a_virus |
| Type | virus |
| Scientific name | Influenza A virus |
| NCBI taxid | 11320 |
| Evidence | 4 document(s) |
| Assigned | 2026-09-04, against criteria version 093366e352a2 |
Overview
Influenza A virus in swine causes an abrupt respiratory illness — fever, coughing, laboured breathing, loss of appetite — that sweeps through a group in a day or two and usually resolves within a week. Mortality is low, but morbidity approaches 100%, and the growth setback, the treatment costs and the way infection opens the door to bacterial pneumonia make it a routine and substantial cost. In the United States the H1N1, H1N2 and H3N2 subtypes all circulate, in a large number of genetic lineages that change continually through mutation and through reassortment when a pig is infected with two viruses at once. That variability is the practical problem: vaccines must be matched to what is actually circulating, and many farms use autogenous products for this reason. The public health dimension is what sets influenza apart from every other endemic respiratory disease of pigs. Pigs can be infected by human, avian and swine viruses, which makes them a plausible mixing vessel for reassortment; swine-origin viruses cause sporadic human infections, most often at agricultural fairs; and the 2009 pandemic virus contained gene segments of swine lineage. Human seasonal viruses also move regularly into pigs and establish there.
C1 Zoonotic potential
Likelihood of transmission from pigs to humans
Likely to occur: Transmission from pigs can occur even when control points are intact, or is prevented only by extraordinary control measures such as respiratory protection or vector control
Basis. ch32 32.3, the Public Health section: "HUMAN INFECTIONS WITH SWINE IAV of various subtypes, lineages, and genotypes HAVE BEEN REPORTED IN THE UNITED STATES, particularly with H3N2, H1N2, and H1N1 viruses of the North American triple reassortant lineages... The number of documented infections per year is small compared to the number of people in contact with pigs worldwide. HOWEVER, LIMITED STUDIES OF IAV-S ANTIBODIES IN HUMANS SUGGEST THAT IAV-S INFECTIONS MAY BE UNDER-DIAGNOSED." 32.4.2 on the route: "aerosol detection of IAV-S during naturally occurring outbreaks in swine barns demonstrated significant amounts of virus in the air. Thus, AEROSOLIZED VIRUS IS ALSO A SOURCE OF INFECTION, AN EXPOSURE RISK TO HUMANS DURING OUTBREAKS." cfsph_2024_swine_influenza records the US pattern: variant infections at agricultural fairs, and "serological surveillance of occupationally exposed professions supports the existence of MILD OR ASYMPTOMATIC INFECTIONS".
How this level was decided
Level 4, and this entry is the framework's worked example there. The label asks whether transmission from pigs can occur even when control points are intact, and influenza is the case it was written for: the virus is in the air of an infected barn, it infects people who simply walk through it, and the US variant cases come from ordinary contact at agricultural fairs rather than from any identifiable failure. NOTE THE UNDER-DIAGNOSIS POINT, which is why the small annual case count does not argue for a lower level: serology in occupationally exposed people suggests mild and asymptomatic infections are missed, and the chapter says interpreting those studies is compromised because swine and human influenza antibodies cannot be cleanly separated. The traffic also runs the other way and more heavily -- "human-to-swine transmission of influenza virus is FAR MORE FREQUENT than swine-to-human" -- but C1 scores the direction from pigs.
C2 Zoonotic impact
Severity of human illness caused by the agent
Mild: Human illness is usually mild and self-limiting, and serious disease occurs only rarely
CONFIRMED by Eric. The assignment reached this level from the evidence, and he has reviewed and ratified it.
Confirmed by Eric. CONFIRMED. "c2l2, most transmission events to humans are mild, there is always a risk of a major uptick in human severity but this remains unlikely, even in the 2009 pandemic" THIS SETTLES A THREE-WAY QUESTION AND MOVES A WORKED EXAMPLE. The pass had flagged level 2, 3 or 4: level 2 on the mildness of confirmed variant cases, level 3 because criteria_levels.yml names influenza A as its C2 LEVEL 3 example, level 4 if the 2009 pandemic counts as human illness acquired from pigs. He rules level 2 and answers the pandemic argument directly -- "there is always a risk of a major uptick in human severity but this remains unlikely, EVEN IN THE 2009 PANDEMIC". So the C2 level 3 example moves to level 2, the fifth example relocated on his ruling after Cystoisospora suis, Mycoplasma hyopneumoniae, pseudorabies and Menangle.
Basis. cfsph_2024_swine_influenza, on variant virus infections in people: "Swine influenza in people is generally indistinguishable from the illnesses caused by human influenza A viruses (seasonal influenza). MOST CONFIRMED CASES HAVE BEEN MILD, THOUGH YOUNG CHILDREN WERE SOMETIMES HOSPITALIZED FOR DEHYDRATION... PNEUMONIA, SERIOUS ILLNESSES AND DEATHS HAVE BEEN REPORTED SPORADICALLY, GENERALLY, THOUGH NOT ALWAYS, IN PEOPLE WHO HAD UNDERLYING HEALTH CONDITIONS or were immunocompromised by other illnesses or pregnancy." The factsheet opens: "In people, clinical cases tend to resemble human influenza and are often mild, though there have been A FEW DEATHS. Most cases involve a single person who had" contact with pigs. ch32 32.3 records the other thing pigs do: the H1N1pdm09 virus "likely arose through... a unique intercontinental reassortment between swine IAV of North American and Eurasian lineages, followed by IN TOTO TRANSMISSION FROM PIGS TO HUMANS, human-to-human transmission, and global spread by humans".
How this level was decided
Level 2, and FLAGGED AS A LEVEL 2 OR 4 JUDGMENT that is the most consequential in this batch. LEVEL 2 IS SCORED on what actually happens to people who catch influenza from a pig: an illness indistinguishable from seasonal flu, mild in most confirmed cases, with hospitalisation mainly for dehydration in small children and deaths sporadic and mostly in people already unwell. That is "usually mild and self-limiting, and serious disease occurs only rarely" almost word for word. THE ARGUMENT FOR LEVEL 4 IS THE 2009 PANDEMIC, and it is not a small argument: that virus was assembled by reassortment in pigs, transmitted in toto from pigs to humans, and then killed on a pandemic scale. If C2 grades the worst human outcome an agent in pigs has produced, influenza A is a level 4 and nothing else in the register is close. WHY THE PASS SCORES IT LOWER: the pandemic's severity came from human-to-human spread, and the chapter says there is no evidence that infection in pigs contributed to widespread human transmission. The pig-to-human step was one seeding event. C1 already carries the fact that the step happens; making C2 carry the pandemic as well would count pandemic risk twice, which PAPRIKA cannot see because it adds the criteria. THIS IS ERIC'S CALL. AND THE CRITERIA FILE ANSWERS IT DIFFERENTLY AGAIN, which validate_criteria.py reports: influenza A virus is a worked example at C2 LEVEL 3, "clinically important illness is common and may require medical treatment or hospitalization, but death or permanent disability is uncommon". THAT IS A DEFENSIBLE THIRD ANSWER and arguably the best one: young children with variant infections have been hospitalised for dehydration, pneumonia and deaths occur sporadically, and seasonal influenza is not an illness most people would call self-limiting. SO THE CELL HAS THREE CANDIDATES, not two -- level 2 on the mildness of most confirmed variant cases, level 3 on the framework's own calibration, level 4 if the 2009 pandemic is counted as human illness acquired from pigs. If Eric rules level 3 the example stands and nothing else needs touching; if he rules 2 or 4 the example moves, as it has already done for Cystoisospora suis, Mycoplasma hyopneumoniae, pseudorabies and Menangle.
C3 Herd introduction risk
Effort required to keep the agent out of the herd
Extraordinary biosecurity required: Exclusion needs measures beyond the routine, such as air filtration, thermally treated feed and bedding, or repeated costly testing to find carriers, and may still fail
Basis. ch32 32.4.2: "INFLUENZA VIRUSES ARE INTRODUCED INTO HERDS WITH THE MOVEMENT OF ANIMALS. The primary route of virus transmission is direct contact with infectious oronasal secretions, with virus titers of up to 1 x 10^7 infectious particles per mL at the peak of shedding... Additionally, aerosol detection of IAV-S during naturally occurring outbreaks in swine barns demonstrated significant amounts of virus in the air. Thus, aerosolized virus is also a source of infection... and MAY CONTRIBUTE TO IAV-S INFECTIONS INCLUDING FARMS WITH HIGH BIOSECURITY." And the route no biosecurity plan covers, from 32.1: "it is now clear that HUMAN-TO-SWINE TRANSMISSION OF INFLUENZA VIRUS IS FAR MORE FREQUENT THAN SWINE-TO-HUMAN AND IS CENTRAL IN GLOBALLY SEEDING SWINE POPULATIONS WITH NEW VIRAL DIVERSITY." 32.4.2 on why it does not leave: "on farrow-to-finish farms, the virus may persist in growing pig populations due to the CONTINUOUS INTRODUCTION OF SUSCEPTIBLE YOUNG PIGS with declining levels of maternal immunity."
How this level was decided
Level 4, and this entry is the framework's worked example there alongside PRRSV and PEDV. The chapter states the level 4 test in its own words -- aerosol spread reaches farms with high biosecurity -- so quarantine and fomite control are not enough and this is not level 3. IT IS ALSO NOT LEVEL 1, which would be the easy answer for a virus most herds have: the label asks whether there is an introduction event to prevent, and there plainly is, because new lineages arrive continually and a herd's status can change with a single introduction. THE ROUTE THAT MAKES THIS DISTINCTIVE IS THE STAFF. Human-to-swine transmission is the main way new diversity enters the world pig population, and no biosecurity programme in the register is built to stop a stockperson breathing.
C4 Detection difficulty
Difficulty of recognizing and confirming infection
Laboratory-dependent: The presentation does not point to this disease specifically, but local or regional laboratories can confirm it with a validated assay once it is suspected
Basis. ch32 32.8: "THERE ARE NO PATHOGNOMONIC SIGNS and swine influenza must be differentiated from a variety of respiratory diseases of swine with similar clinical and pathological presentation. A diagnosis is only possible through isolation of virus, detection of viral proteins or nucleic acid, or by demonstration of virus-specific antibodies." On the assays: "RT-PCR assays (traditional and real-time technologies) provide HIGHLY ANALYTICALLY SENSITIVE AND SPECIFIC DETECTION of viral nucleic acid... with greater speed, reduced cost, and scalability as inherent advantages", pan-influenza A assays for screening and subtype-specific assays for characterisation; virus isolation in MDCK cells; commercial ELISAs; and "ORAL FLUIDS ARE INCREASINGLY USED FOR SCREENING POPULATIONS", along with nasal wipes, udder wipes and family oral fluids for litter-based diagnosis. The limits are about interpretation: "the sensitivity of these approaches is DEPENDENT ON THE SPECIALIZED REAGENTS used in the assay(s) AND THEIR DEGREE OF MATCH TO CIRCULATING FIELD STRAINS", and serological interpretation "is complicated by the concurrent circulation of different virus subtypes and gene lineages".
How this level was decided
Level 2, and this entry is the framework's worked example there. Both halves hold plainly: nothing about coughing pigs points to influenza specifically, and validated assays run everywhere -- pan-influenza RT-PCR on oral fluids is a routine US herd monitoring tool, not a specialist procedure. NOTE THE ONE THING THAT WOULD NOT BE OBVIOUS TO A READER, and that the C4 rewording of 2026-08-29 was written to catch: a subtype-specific assay is only as good as its match to what is circulating, and with the diversity this chapter describes, a negative on a subtype-specific test is not a negative for influenza. The pan-influenza A assay is the screening tool for that reason.
C5 Production cost
Financial impact on the infected farm's cost of production
Moderate: Losses measurably increase cost of production but remain manageable within normal farm operations, whether acute, chronic, or associated with endemic disease
CONFIRMED by Eric. The assignment reached this level from the evidence, and he has reviewed and ratified it.
Confirmed by Eric. CONFIRMED. "c5l3, influenza in pigs used to show up as an acute outbreak that resolved quickly but the presentation seems to be changing due to different strains, production systems, vaccine use, or something, L3 is correct for today"
Basis. ch32 32.6: "MORBIDITY IS HIGH (UP TO 100%), BUT MORTALITY IS LOW (usually less than 1%) in uncomplicated infections. Generally, RAPID RECOVERY BEGINS 5-7 DAYS AFTER ONSET." cfsph_2024_swine_influenza puts a range on it and names the cost: "mortality rates are generally low in uncomplicated cases, with case fatality rates typically ranging from <1% to 4%, and most animals recovering within 3-7 days. THE MAIN ECONOMIC IMPACT IS USUALLY FROM REDUCED WEIGHT GAIN AND A LONGER TIME TO REACH MARKET WEIGHT." Against that, 32.6 on why it is not usually uncomplicated: "SECONDARY BACTERIAL INFECTIONS with Actinobacillus pleuropneumoniae, Pasteurella multocida, Mycoplasma hyopneumoniae, Glaesserella parasuis, or Streptococcus suis type 2 MAY ENHANCE THE SEVERITY of clinical disease", and "PRRSV, M. hyopneumoniae, and IAV-S are most frequently detected in 10- to 22-week-old pigs in association with the PORCINE RESPIRATORY DISEASE COMPLEX". 32.4.2: the virus persists in growing pig populations through the continuous supply of susceptible young pigs, and 32.4.3 records repeated waves as new lineages arrive.
How this level was decided
Level 3, and FLAGGED AS A LEVEL 2 OR 3 JUDGMENT. The case for level 2 is the uncomplicated outbreak, which is exactly the level 2 label: high morbidity, under 1% mortality, recovery in five to seven days -- small and short-lived. THE CASE FOR LEVEL 3, WHICH IS SCORED, is that the uncomplicated outbreak is not the usual state of affairs in a US herd. This virus does not visit once: it persists in the growing population through the continuous supply of susceptible pigs, new lineages arrive repeatedly, most herds vaccinate twice-yearly against it, and its main clinical importance is as one of the three agents of PRDC. The label allows level 3 for losses that are "acute, chronic, or associated with endemic disease", and influenza in a US herd is all three. It is scored level with Mycoplasma hyopneumoniae and PCV2, the other two members of that complex.
C6 Market impact
Duration of material disruption to pork and pig markets
Negligible: Little or no material disruption to supply or demand when disease occurs on one or more farms
CONFIRMED by Eric. The assignment reached this level from the evidence, and he has reviewed and ratified it.
Confirmed by Eric. CONFIRMED. "c6l1, in all but rare instances, influenza in pigs has not caused a significant impact on markets. It all depends on the situation of course, but for the typical situation, influenza is not affecting markets, nor did it in the recent h5n1 dairy cattle outbreak in the US." HE SUPPLIED A PRECEDENT THE CHAPTER DOES NOT HAVE, and it is a recent one: "nor did it in the recent h5n1 dairy cattle outbreak in the US." That answers the objection the pass raised against its own level 1 -- that 2009 saw pork import bans and a demand collapse after a human pandemic was named after this species. An avian influenza event in another US food-animal species, with human cases, moved no pork market. The 2009 reaction was to a NAME, not to influenza in pigs.
Basis. ch32 32.5, stated as a finding rather than an aside: "Pig infection studies with the H1N1pdm09 virus confirmed the ABSENCE OF VIRUS IN PORK AND MUSCLE TISSUE, SO IAV-S IS NOT A FOOD SAFETY OR PORK TRADE CONCERN." IAV-S is endemic in US herds, detected year-round with seasonal peaks, and diagnosed continuously; no trade measure or consumer response following a farm diagnosis is recorded anywhere in the chapter. 32.4.3.1 records what the 2009 virus was called and what it was not: "although the H1N1pdm09 virus was coined SWINE FLU, THERE IS NO EVIDENCE TO SUGGEST THAT INFECTION IN PIGS CONTRIBUTED TO WIDESPREAD HUMAN TRANSMISSION or maintenance as a human seasonal strain."
How this level was decided
INFERRED at level 1, AND FLAGGED, because the criterion and the history point different ways and Eric lived through the history. C6 asks what happens to the market WHEN DISEASE OCCURS ON ONE OR MORE FARMS, and the answer for influenza is nothing: it occurs on most farms most years, it is diagnosed constantly, the chapter says outright that it is not a pork trade concern, and no market has ever moved on a herd diagnosis. Under the standing note SCORING/market_impact that is a long precedent and it gives level 1. THE ARGUMENT AGAINST IT IS 2009, and it is a real one: pork import bans and a demand collapse followed the naming of a human pandemic after this species, which is a material disruption lasting well over a month. The pass does not score it here because the event was not disease occurring on farms -- it was a human pandemic that borrowed the pig's name, and the chapter is emphatic that pigs did not drive it. IF THE CRITERION SHOULD CARRY REPUTATIONAL RISK RATHER THAN DISEASE EVENTS, this is level 3 or level 4 and it is the clearest example in the register.
C7 Treatment potential
Potential for treatment to improve outcomes
Little: Effective pathogen-directed treatment is available and works, or animals recover acceptably without it
CONFIRMED by Eric. The assignment reached this level from the evidence, and he has reviewed and ratified it.
Confirmed by Eric. CONFIRMED. "c7l1, conventional strategy in people and animals is to let the infection run its course, with supportive care. The infection timeline is so short that you typically need to have antivirals on board at or before the time of infection, in order for them to have any real effect." HIS REASON IS A GENERAL ONE ABOUT ACUTE VIRAL DISEASE and is worth carrying to other entries: "the infection timeline is so short that you typically need to have antivirals on board AT OR BEFORE THE TIME OF INFECTION, in order for them to have any real effect." Where the illness resolves faster than a treatment could be started, C7 is level 1 even though no product exists -- a different route to level 1 than "the vaccine is so good" (pseudorabies, PCV2) and than "the harm is complete before birth" (the transplacental agents).
Basis. No antiviral treatment for influenza in pigs is described anywhere in chapter 32; 32.10 is titled Prevention and Control and is entirely about vaccination. 32.6: "morbidity is high (up to 100%), but mortality is low (usually less than 1%) in uncomplicated infections. GENERALLY, RAPID RECOVERY BEGINS 5-7 DAYS AFTER ONSET." 32.9: "THE IMMUNE RESPONSE TO IAV-S INFECTION IS RAPID AND EFFICIENT WITH COMPLETE ELIMINATION OF THE VIRUS FROM THE RESPIRATORY TRACT WITHIN 7 DAYS post inoculation", and a primary infection gives significant protection against reinfection with the same or a similar strain.
How this level was decided
Level 1 on the second clause -- animals recover acceptably without treatment -- and the chapter supports it directly rather than by silence: the pig clears the virus in a week by its own immune response, mortality is under one percent, and recovery is described as rapid. FLAGGED AS A LEVEL 1 OR 3 JUDGMENT for one reason, which is that the usual escape route is closed here. Pseudorabies and PCV2 are both scored level 1 partly because their vaccines are so good that a treatment would add little; influenza's vaccines are NOT good, which is why C8 sits at level 2, so that argument is unavailable. What carries level 1 instead is the recovery itself. If the view is that shortening the illness in the face of secondary bacterial invasion and PRDC would meaningfully improve outcomes -- and Eric's 2026-09-02 rule says reducing shedding duration counts as an outcome in its own right -- this is level 3.
C8 Vaccine availability
Availability of effective vaccines or bacterins
Available but inconsistent: Commercial or autogenous vaccines exist in the US but protection may be inconsistent
Basis. ch32 32.10: "VACCINATION IS THE PRIMARY MEANS OF PREVENTING INFLUENZA IN PIGS. Many commercial IAV-S vaccines are inactivated whole virus vaccines with an adjuvant", with two North American commercial products covering multiple H1 and H3 clades, a monovalent H1N1pdm09 product, and two replicon-particle vaccines licensed since 2012 and 2022. AND THE LIMITS, all in the same section: "THE EFFICACY OF INACTIVATED IAV-S VACCINES IS CONFOUNDED BY THE CURRENT DIVERSITY IN IAV-S STRAINS. IT IS HIGHLY UNLIKELY THAT ONE STRAIN PER SUBTYPE WOULD BE EFFICACIOUS GLOBALLY. In addition, a single H1 or H3 vaccine strain WILL FREQUENTLY FAIL TO CROSS-PROTECT against all the co-circulating virus lineages within a given region." "IN CONTRAST WITH HUMAN INFLUENZA VACCINES, THERE IS NO FORMAL SYSTEM FOR RECOMMENDING IAV-S VACCINE STRAINS", and US manufacturers "are not required to disclose the names of the vaccine strains in commercial vaccines, WHICH MAKES FORECASTING THEIR POTENTIAL COVERAGE LESS CERTAIN". "CUSTOM-MADE, INACTIVATED, AUTOGENOUS VACCINES ARE POPULAR IN THE UNITED STATES AND MAY ACCOUNT FOR MOST OF THE VACCINES USED." 32.7 records the failure mode that is worse than no protection: vaccine-associated enhanced respiratory disease, "significantly increased gross and microscopic pathology in vaccinated pigs versus unvaccinated controls following challenge" with a drifted strain of the same subtype.
How this level was decided
Level 2, and this entry is the framework's worked example there. The label is "commercial or autogenous vaccines exist in the US but protection may be inconsistent", and every clause is documented: commercial products exist, autogenous products probably account for most doses given, and protection depends on a strain match that nobody coordinates -- there is no equivalent of the human vaccine strain selection system, and US manufacturers do not have to say what is in the bottle. NOTE THE ONE FACT THAT PUTS THIS BELOW EVERY OTHER LEVEL 2 IN THE REGISTER: with a sufficiently drifted strain of the same subtype, vaccination can make the disease WORSE, which is not merely inconsistent protection but negative protection. That is also why the C7 argument used for pseudorabies and PCV2 -- the vaccine is so good that no treatment is needed -- cannot be used for influenza.
Levels and evidence are generated from data/assignments/influenza_a_virus.yml; the overview is authored in data/overviews/influenza_a_virus.md. Do not hand-edit this page — corrections go in data/assignments/CORRECTIONS.yml.
Quoted material marked Basis is from Zimmerman JJ, Karriker LA, Ramirez A, Schwartz KJ, Stevenson GW, Zhang J, eds. Diseases of Swine, 12th edition. Hoboken: Wiley Blackwell, 2025 — except where another source is named in the quotation itself.