One of the main reasons why we have jobs as veterinarians is because animals routinely get sick from various infectious and non-infectious causes. At any given time, for example, we would expect there to be at least some patients in our population with itchy skin, irritated eyes, limping, or coughing. When we start to see a lot more of these cases than we would normally expect under the circumstances, this could mean there is something else going on with the animals, environment, and/or management that is causing them to be sick. We call these situations outbreaks and, depending on the severity and nature of the problem, these may require further investigation and control efforts. In this section we will work through the general approach for conducting outbreak investigations at an individual herd level and then in the next section, we will learn more about the principles for managing infectious disease outbreaks at the wider population level.
An outbreak is formally defined as a situation where there are more cases of disease than expected in a given area or among a specific group of individuals over a particular period of time. Usually, these cases are presumed to have a common cause or to be related to one another in some way.
The foundations of modern outbreak investigation are often traced to the work of John Snow, a London physician in the mid-1800s. During a large cholera outbreak, Snow observed a sudden increase in deaths associated with severe watery diarrhoea and systematically mapped the locations of affected cases. He identified clear spatial clustering around a public water pump on Broad Street and proposed that contaminated water was the source of infection.
This hypothesis was tested by removing the pump handle and advising residents to obtain water from alternative sources.
Within a short period, cholera cases declined sharply and the outbreak was brought under control. This occurred despite the causative agent being unknown at the time and in the absence of effective antimicrobial treatments. Snow’s approach, based on careful observation, data-driven hypothesis development, and targeted intervention, remains central to contemporary outbreak investigation.
These same principles underpin herd-level disease investigations in veterinary medicine, where veterinarians working in herd health, production systems, or shelter medicine are frequently responsible for identifying, investigating, and managing outbreaks within animal populations.
Before we can even begin to conduct an outbreak investigation, someone first has to notice that something is wrong and they have to bring it to the attention of someone who is both willing and able to do something about it. There are many different ways outbreaks can be uncovered in animal populations including:
The Mycoplasma bovis outbreak in the New Zealand cattle industry, for example, was detected by a veterinarian who noticed atypical features with the mastitis cases they treated in a dairy farm in the south island. In 2019, there was an incursion of infectious bursal disease into the New Zealand poultry industry that was detected through routine sampling and surveillance on commercial broiler farms. You will also hear the term syndromic surveillance which means that we are monitoring populations for an increase in particular clinical presentations such as respiratory signs or abortions rather than looking for a specific pathogen or other causative agent.
Once a potential outbreak has been identified and deemed worthy of further investigation, it is important to follow a robust process for conducting the investigation and documenting your findings.
You will see various different versions of the steps involved in outbreak investigations, but they all generally follow the same basic process of verifying that there is actually an outbreak to investigate, doing the grunt-work to figure out what is going on and what you are going to do about it, and then implementing interventions to see if they have the desired effect of controlling the outbreak.
Although we are presenting the steps as a very logical and linear sequence here, the real-world of outbreak investigation can be a lot messier and you will often find yourself re-visiting previous stages as you gain more information about the outbreak.
Steps for investigating outbreaks
Verifying the outbreak
Investigating the outbreak
Before launching into a full-scale outbreak investigation, it is first important to verify that you are in fact dealing with an outbreak situation. Some common scenarios where you may get false alarms:
For example, if we previously had 15 piglet deaths in a population of 150 piglets and it changes to 30 piglet deaths in a population of 300 piglets, the incidence of mortality has stayed the same at 10% despite having a doubling in the actual case counts. It’s also important not rely on the memories and perceptions of farm staff since they may not accurately remember previous statistics. On farms with chronic disease problems, farmers may develop a skewed perception of what is considered “normal” disease incidence – a “normal” year for their farm could actually be many times worse than the industry averages.
The first step in any outbreak investigation is to clearly describe the problem being observed. This initial description should include the clinical syndrome, the management groups affected, and the apparent duration and severity of disease.
At this stage, the definition is often broad and descriptive rather than diagnostic. As additional information is collected, this working definition can be refined into a more formal case definition that better distinguishes outbreak-related cases from background disease.
Wherever possible, objective data should be used to confirm that the observed situation represents a genuine increase in disease frequency compared with what would normally be expected for that population, location, and time period. This step is critical, as not all apparent increases in case numbers reflect meaningful changes in disease risk.
What constitutes an outbreak depends on the disease in question. For rare, exotic, or notifiable diseases like foot-and-mouth disease, the presence of a single case may be sufficient to trigger an outbreak response. In contrast, for common endemic conditions, an outbreak is usually defined by an increase in the number or proportion of cases above established baseline levels rather than by isolated cases.
Verification typically involves comparing current disease frequency with historical farm records and, where available, industry benchmarks or published reference values. This process is often more straightforward in intensive production systems, such as poultry or swine operations, where detailed production and health records are routinely collected. In more extensive systems, including pastoral beef and dairy herds, historical animal health data may be limited, inconsistently recorded, or difficult to analyse, making outbreak verification more challenging.
When data are sparse, triangulating multiple sources of information can be helpful. This may include farmer records, veterinary practice data, laboratory submissions, and broader industry surveillance reports. Importantly, disease counts should be interpreted in relation to the size of the population at risk, as changes in herd size or structure can influence the absolute number of cases without reflecting a true change in disease incidence.
After confirming that a true excess of disease exists, the focus shifts to understanding why it is occurring. This stage involves systematic data collection to identify potential causes and contributing factors related to animals, environment, and management.
Many herd health problems are multifactorial, meaning that disease emerges from the interaction of multiple risk factors rather than a single cause. The goal of investigation is therefore not only to identify pathogens, but also to understand the conditions that allow disease to occur and spread.
A clear case definition is central to effective outbreak investigation. It ensures consistent identification of cases across time and personnel, supports accurate calculation of disease frequency, and allows meaningful evaluation of interventions.
Case definitions may be specific, based on confirmed diagnoses, or broader working definitions when diagnostic certainty is not yet available. Importantly, a definitive aetiological diagnosis is not always required to control an outbreak.
If you can make a formal definitive diagnosis, then it’s relatively easy to write a specific case definition (i.e. pastuerella pneumonia in recently weaned calves). However, if you can’t make a definitive diagnosis or want to cast a wider net to identify more potential cases, it’s still possible to provide a working case definition (i.e. Acute respiratory disease in recently weaned calves). You don’t actually need to have a definitive aetiological diagnosis in order to resolve an outbreak.
Enhanced surveillance is often needed to identify additional cases and track the outbreak over time. This typically involves structured, routine recording of new cases using a consistent set of variables such as date, location, animal identifiers, clinical signs, and outcomes.
Beware that enhancing surveillance can sometimes make it seem like the outbreak is initially getting worse because people are likely to be much more vigilant about identifying and reporting cases, particularly ones that may have milder clinical signs.
At the same time as we are enhancing surveillance, we also want to start collecting some more information about the animals, environment, and management so that we will be able to start formulating and testing hypotheses about why the animal got sick. For a lot of common herd health problems, you can often find worksheets like the example on the right that help you to collect and record the appropriate kinds of information during your outbreak investigation. It’s also important for us to collect information on non-cases as well as cases so that we will have a comparison group to determine if certain factors are associated with an increased frequency of cases.
This would include collecting information about things like the animal’s age, sex, breed, stage of the production cycle, and production levels, behaviour, previous health conditions, and preventative health care interventions like vaccination or deworming that can all influence their current risk of developing disease.
This would include collecting information about where the affected and unaffected animals are physically located in space as well additional environmental and management factors like temperature, humidity, hygiene, and air quality that may predispose them to getting sick. John Snow, for example, noticed spatial clustering of cholera cases centred around the Broad Street water pump
This would include collecting information on when the problem actually began and how the pattern of cases has shaped up over time. We often plot the number or proportion of individuals becoming infected over time (epidemic curves), which can help us identify the type of exposure as well as features around transmission.
Just like when we were verifying the outbreak, it is important that we always consider the size of the underlying population at risk and not just the total number of cases when we are looking for evidence of clustering by individual, place, or time. We will often see the largest total number of cases in areas with the largest population and so we want to make sure that we adjust all estimates by the underlying population size (i.e. we would expect to see a lot more COVID cases reported in Auckland with a population of 1.6 million than in either Wellington with a population of 200,000 or Palmerston North with a population of 90,000.
Now that we have started collecting and recording some more information about the outbreak, we can pull this together using our critical thinking skills to come up with some hypotheses about what might be causing the outbreak and what we could potentially do to establish a definitive cause. This is where risk ratios can be particularly useful because we can look for statistical evidence of some factors being more strongly associated with disease than others. Remember from previous sections that an RR of 1 means there is no association, an RR < 1 means that something is a protective factor, and an RR > 1 means that something is a risk factor. These can often help us narrow down potential sources of the outbreak (i.e. the potato salad at the BBQ that was associated with a higher risk of people developing foodborne illness) or things we can target as preventative measures to control the outbreak (i.e. using hand sanitisers reducing the risk of illness). We can calculate the attributable fraction to determine what % of cases could potentially be reduced by targeting a potential risk factor.
The purpose of developing hypotheses is to come up with the most likely explanation for what is causing the problem. This is where subjective information from managers and other professionals can frequently be helpful. For example, you may be investigating an increase in respiratory disease in a poultry flock and the farmer conveniently tells you “Oh by the way doc, two of the five ventilation fans on the barn have been acting up over the past couple weeks”. Or you may want to talk to your local veterinary colleagues to ask them if they have seen or diagnosed similar problems in your practice area.
Even with a small outbreak on a single farm, you can still design studies to help get to the bottom of the potential causes although these may not always be needed or practical depending on the situation. Case-control studies and cohort studies can be particularly valuable for testing your hypotheses as both seek to identify associations between exposures and the disease of interest. As part of this step, you should make predictions about what you should expect to find if your hypothesis is correct (i.e. what would the test results look like or how much would the production effects change). Finding what you predict will generally support your hypothesis while not finding what you predict will generally weaken your hypothesis.
Hopefully by this stage, we have a better understanding of what might be causing the outbreak based on the diagnostic tests and additional descriptive data we have obtained. Our next job is going to be making some recommendations about what we can do to control the outbreak, getting a system in place to make sure that our recommendations are having the desired effect, and making sure that we have been clearly documenting what we have been doing. It’s not unusual to have to re-visit previous steps in the outbreak investigation as the situation progresses and new information becomes available.
When dealing with any animal health issue whether it’s for an individual patient presenting for a consult or for a herd that you are investigating for an outbreak, I like to divide my recommendations into short-term ones (i.e. what can we do today to decrease the impact of disease) and longer-term ones (i.e. what changes could we make moving forward to help reduce the impact of disease).
Shorter-term recommendations
For pretty much any outbreak situation where there is a potentially infectious aetiology, you can never go wrong with getting biosecurity and hygiene practices brought up to speed ASAP since these measures are likely to be effective across a wide range of pathogens. This could include restricting the number of people entering the barn, installing footbaths at the main entrance, changing or cleaning PPE between pens, and installing better barriers between pens to prevent nose-to-nose contact.
Please do not just tell farmers to “Improve biosecurity”. The more specific you can make the recommendations, the more likely it is that they will actually be followed (i.e. “Install a Virkon footbath outside each pen and scrub boots for at least 30 sec before and after entering the pen. Use hand sanitizer between pens after handling scouring piglets. ”
Longer-term recommendations
Once the immediate outbreak has been stabilised, attention should shift to reducing the likelihood and impact of future outbreaks. Longer-term recommendations focus on addressing underlying risk factors that predispose the herd to disease, rather than simply interrupting transmission in the short term. These measures often require more planning, investment, or behavioural change, and their effects may not be apparent for weeks or months.
Common longer-term strategies may include adjustments to vaccination programmes, changes to housing design or stocking density, improvements to ventilation or waste management, and refinements to nutrition, colostrum management, or breeding practices. In many cases, these interventions aim to improve baseline immunity or reduce chronic exposure to infectious agents rather than eliminate them entirely.
As with short-term measures, vague recommendations are rarely effective. Advice such as “improve immunity” or “review management practices” should be translated into clear, actionable steps. For example, rather than suggesting vaccination in general terms, a more useful recommendation would specify which animals should be vaccinated, at what stage of production, and how this fits within existing management routines. Similarly, recommendations to improve housing should outline specific changes, such as increasing airflow rates, modifying pen layouts to reduce contact between groups, or altering cleaning schedules to better match pathogen survival characteristics.
It is also important to set realistic expectations about timelines. Many longer-term interventions will not influence the current outbreak, but are intended to prevent recurrence or reduce severity in future production cycles. These measures should therefore be implemented alongside continued surveillance so that their impact can be evaluated over time and adjusted if necessary.
Finally, longer-term recommendations should be prioritised based on feasibility, cost, and likely impact. Attempting to implement too many changes at once can undermine compliance and dilute effectiveness. A small number of well-defined, achievable interventions is often more successful than an extensive list of aspirational improvements.
Similar to Step 3, we want to make sure we continue to record cases over time to make sure that the interventions we have put in place are actually having the desired effect of reducing disease cases. Again be wary of surveillance bias where it can initially look like the outbreak is getting worse because the farmer is detecting more of the milder clinical cases in the population. You should have a rough idea of when you would expect to see improvements and how much improvement there should be during that period. If things are not progressing as you would expect, you may need to re-evaluate your hypotheses or try some different interventions.
Clear, timely communication and thorough documentation are essential components of effective outbreak management. Throughout the investigation, veterinarians should maintain accurate written records of observations, diagnostic findings, hypotheses considered, recommendations made, and decisions taken. These records provide a permanent reference that supports continuity of care, enables ongoing monitoring of outcomes, and demonstrates due diligence in professional practice.
For herd-level investigations, communication with the primary decision-makers on the farm is particularly important. Written summaries help ensure that recommendations are clearly understood, reduce the risk of misinterpretation, and provide a basis for follow-up discussions as the situation evolves. Documenting both accepted and declined recommendations is also important, as it provides transparency and protects all parties if outcomes do not improve as expected.
In more complex situations, communication may need to extend beyond individual clients. Multi-site outbreaks, unusual disease patterns, or suspected notifiable diseases may require engagement with industry organisations, diagnostic laboratories, regulatory authorities, and, where there are zoonotic or food safety implications, public health agencies. Early and appropriate notification helps coordinate responses, facilitates access to additional resources, and supports consistent messaging across stakeholders.
In some circumstances, particularly where outbreaks attract public attention or have wider industry impacts, communication with the media may also be required. In these cases, messages should be factual, proportionate, and aligned with guidance from relevant authorities. Clear internal communication within veterinary teams and across collaborating organisations is equally important to ensure consistency and avoid conflicting advice.
Effective communication is not a one-off event but an ongoing process. As new information becomes available and control measures are implemented, updates should be shared with relevant stakeholders to maintain trust, support compliance, and enable timely adjustments to the outbreak management plan.
Outbreak investigations are a routine but important component of herd health practice. By applying a structured, evidence-based approach to verifying disease increases, collecting and analysing data, testing hypotheses, and monitoring outcomes, veterinarians can effectively manage outbreaks while minimising impacts on animal health, welfare, and production.
Advancing Animal Care through Shared Learning