If you trained as a veterinary professional, you have spent a lot of time learning about the different diseases that affect our veterinary patients as well as the different diagnostic tests, medications, and surgical procedures that we have at our disposal to treat them. Veterinary epidemiology is the speciality field in veterinary medicine that is dedicated towards studying populations to understand what causes disease and what steps we can take as animal health decision-makers to better prevent, manage, and cure disease. In other words, epidemiology is the science underpinning the information we use to make evidence-based clinical decisions in practice.
A successful veterinary consultation involves many interconnected steps, whether the goal is to prevent disease through better routine care, manage a sick animal, or respond to an emergency. While the context and urgency may differ, all consultations follow a similar underlying structure: gathering a history, performing a physical examination, selecting and interpreting diagnostic tests, choosing appropriate treatments, and monitoring outcomes to ensure those interventions are working. We will now explore these steps again through an epidemiological and research lens to understand how science can help us improve outcomes.
Most consults start with the client ringing the clinic to book a consult because they have noticed something wrong with their animal. This requires the client to have some knowledge of what is considered normal for that animal or species and then some criteria around when the problem has become severe enough to seek advice from a veterinarian. For example, an owner might notice their dog wanting to go outside to pee 6 or 7 times per day compared to the normal 3 to 4 times and decide to call the veterinarian when the dog starts having accidents in the house.
In epidemiology and population medicine, we call this process of comparing a patient against normal values for either itself or its species to identify when things are going wrong benchmarking. If disease in a population is happening more frequently than we expect, we may be dealing with an outbreak situation that we need to investigate.
You will generally have some basic information in the practice management software including the appointment date as well as the species, breed, age, sex, and reproductive status of the animal. This information is referred to as the patient signalment. This is important information to know because certain diseases are more common in different time periods, different locations, and different patient demographics. For example, your list of differential diagnoses for a dog presenting for increased urination would be very different for an 8-week-old male puppy compared with an 8-year-old spayed female.
In epidemiology, we use the term prevalence to describe how common disease currently is in a particular population and the term incidence to describe the risk of an individual without the disease getting sick over a set time period. These are also collectively called measures of disease frequency.
At the beginning of each consult, we typically ask the client a series of questions about progression of clinical signs associated with the current problem as well as questions about the animal’s routine diet, management, environment, historical medical problems, and recent events. This helps us to identify if patients have been exposed to anything that may have increased their chances of acquiring a particular disease.
For example, if an animal has a history of recurrent urinary tract infections, we would be more suspicious of the current urinary issues being another infection. We also know that female dogs who have been spayed are at greater risk of developing urinary incontinence due to progressive loss of tone in the urethral sphincter when there is no longer oestrogen being produced by the ovaries. For routine vaccine consults, we might ask owners if their dogs will be boarded at a kennel since we know this increases the risk of getting kennel cough (Bordetella bronchiseptica) and we can preventively give a kennel cough vaccine to protect them against future infections.
In epidemiology, we use the term exposures to describe anything that could influence the likelihood of an animal having or developing disease. These can either be a risk factor like being boarded at a kennel that increases the likelihood of disease or a protective factor like vaccination that decrease the likelihood of disease. We can compare the frequency of disease in a group of animals that have risk factor and a group of animals that do not have the risk factor using measures of association to determine how strongly a risk factor influences the disease outcome.
The next step in the consult process is to conduct a nose-to-tail physical examination where we document all the observable clinical impacts of disease. We may also collect data on normal physiological parameters such as body temperature, heart rate, respiratory rate, capillary refill time, and mucous membrane colour. For example, we would not expect to see any significant systemic changes in our patient if it just has a simple urinary tract infection causing the increased frequency of urination. However, if our patient also had progressive weight loss, nausea, retinal detachment, and ulcers in the mouth, we may be more suspicious that we are seeing secondary effects from chronic kidney disease.
In epidemiology, we use the terms pathogenesis or natural history of disease to describe the biological progression of disease in animals after they first become sick. We can study groups of individuals with the disease to figure out which clinical signs are most commonly present. We can also calculate reference ranges or confidence intervals to determine the range of values for physiological measurements that we would expect in most normal healthy patients.
After integrating data we have collected from the signalment, history, and physical examination, we sometimes get lucky and can establish a definitive diagnosis based on the presence of a pathognomonic clinical sign (a clinical finding that is unique to a particular disease and therefore allows us to make a confident diagnosis). More often than not, however, we usually have some theories about what is most likely causing the problem and need to do more investigative work. For example, we may suspect that our dog has a urinary tract infection, but want to do some more testing to confirm the diagnosis and guide our treatment decisions.
In epidemiology, we call our theory about what could be causing disease a hypothesis and this often forms a basis for designing epidemiological research studies in which we use different study design frameworks to observe and collect data about individuals in a population to look for patterns and trends that provide insight about the causes for disease and what we do to better manage them. This is different experimental research studies, which are you more traditional laboratory-based studies that involve experimental manipulations.
Diagnostic tests are any tools that we use to classify an animal as being diseased or non-diseased. These could include physical examination findings, biochemical tests, serological assays, microbiology, radiographs, and other diagnostic imaging. Unfortunately, these tests are usually not completely perfect and there is a chance we could end up missing or misdiagnosing a disease.
In epidemiology, we have tools to measure diagnostic test performance including the sensitivity (probability that the test will detect disease in an animal that truly has disease) and specificity (probability that the test will fail to detect disease in an animal that truly does not have disease). We can combine this with our knowledge of disease prevalence to work out how much we should trust positive and negative test results (predictive values). We frequently use diagnostic tests in epidemiological research studies to assign animals into outcome groups – usually those that have disease (cases) and those that do not (controls) – so we can look for differences between the two groups that could explain why cases got sick.
Interventions refer to any medications, surgical procedures, or husbandry practices that we use to help manage, cure, or prevent disease in our patients. It should be noted that even without a definitive diagnosis, we can still manage disease for the our patients by providing basic supportive care such as giving intravenous fluids to correct dehydration or prescribing medications such as non-steroidal anti-inflammatories (NSAIDS) to treat pain secondary to inflammation or reduce fever while the body does the hard work of fighting off disease.
In epidemiology, we often do studies to compare the efficacy of different interventions and calculate measures of effect to determine the expected clinical impacts of implementing them in practice. Although some research findings may be statistically significant, they may not have clinically significant effects for our patients.
Our job in clinical practice does not end when the patient walks out the door. We usually need to follow-up with the client to make sure that our recommendations have been working to resolve the disease and if they haven’t been working, try to figure out why so that we can change our current course of action for this patient and make better decisions for other patients in the future.
In epidemiology, we use the term critical evaluation to describe the process of reflecting on our findings to identify errors, biases, and confounding factors that could be leading to unexpected or incorrect inferences from our research.
| Stage in consult | Definition in clinical practice | Epidemiology / population medicine equivalent |
|---|---|---|
| Presenting complaint | Owner recognises a change from normal and decides it is significant enough to seek veterinary advice. | Benchmarking against expected “normal” patterns; detecting unusual frequency of disease events that may signal an outbreak or emerging problem. |
| Signalment | Core patient details such as species, breed, age, sex, reproductive status, location, and timing that shape likely differentials. | Measures of disease frequency, including prevalence and incidence, described by time, place, and animal characteristics. |
| Clinical history | Questions on progression, environment, management, diet, prior disease, and recent events to identify likely causes and risks. | Exposures (risk and protective factors) assessed using measures of association to estimate their influence on disease outcomes. |
| Physical examination | Systematic assessment and measurement of physiological parameters to document clinical impacts of disease. | Natural history and pathogenesis; describing common clinical features and expected ranges in affected and unaffected animals. |
| Differential diagnosis | Integrating findings to generate the most likely explanations and guide further investigation. | Hypothesis generation that informs study design and investigation of disease causes in populations. |
| Diagnostic tests | Using tests to classify animals as diseased or non-diseased, recognising imperfect accuracy. | Evaluation of test performance using sensitivity, specificity, and predictive values; defining cases and controls. |
| Interventions | Treating, managing, or preventing disease using medical, surgical, and husbandry approaches. | Estimating intervention effects and interpreting both statistical and clinical significance. |
| Monitoring | Following up to assess response, detect problems, and adjust management as needed. | Critical evaluation of outcomes, considering error, bias, and confounding to inform future decisions. |
Clinical work naturally generates questions. As information is gathered during a consultation, clinicians often find themselves wondering how common different diseases really are, whether a patient has important risk factors, what other clinical signs might be expected, which diagnostic tests are most appropriate, how to interpret the results, and whether one treatment option is likely to work better than another.
Although epidemiologists and veterinarians working in practice are both concerned with disease and the control of disease, there are some important differences in mindsets for how both groups approach the problem. One illustration of these different approaches can be seen when considering an animal with diarrhoeal disease. Clinicians are focused on treating and caring for the sick individual. However, the epidemiologist will want to answer the following additional questions:
We can summarize these differences by saying that veterinarians who work as clinicians are concerned with the health of individuals, while epidemiologists are concerned with the health of the groups of individuals. This is a particularly important mindset for infectious diseases where the decisions made about an individual patient will have wider impacts on the population due to the potential for disease to spread from the infected individual to other susceptible individuals. In epidemiology, we use measures like the basic reproduction number (R0) to estimate how many susceptible individuals a single infected individual is expected to infect and then we look at how measures like infection control or the level of vaccination coverage to achieve herd immunity can slow or stop the rate of disease spread.
Infection control is also particularly important in veterinary medicine because many infectious diseases are zoonotic meaning that they can spread between animals and people. Practicing good hand hygiene and wearing appropriate personal protective equipment (PPE) can help prevent us from getting sick as well as preventing our hospitalised patients from acquiring nosocomial infections from coming into contact with other infected animals or fomites (objects that have been contaminated by infected animals). We will learn a lot more about controlling infectious diseases and outbreak investigation in further sections.
2. Measuring Disease
Advancing Animal Care through Shared Learning