Show Me the Numbers and How to Use Them
World Small Animal Veterinary Association Congress Proceedings, 2017
Dan G. O'Neill1, MVB, BSc(hons), GPCert(SAP), GPCert(FelP), GPCert(Derm), GPCert(B&PS), MSc(VetEpi), PhD, MRCVS; Brenda N. Bonnett2, DVM, PhD
1Veterinary Epidemiology, Economics and Public Health, The Royal Veterinary College, North Mymms, Hatfield, Hertfordshire, UK; 2Epidemiologist, B Bonnett Consulting and CEO, International Partnership for Dogs (IPFD); Georgian Bluffs, ON, Canada

How Do We Get Valid Data on Diseases That Affect Breeding? And How Can We Use Such Data to Ensure Healthy Dogs and Cats?

Have you ever asked a question about how common a condition is in dogs or cats, or in a specific breed? Have you ever wondered why some breeds are more likely to get specific conditions than others? These questions seem basic to understanding and improving the health of pets and yet clear answers are often not available. It is increasingly recognised that radical improvements are needed on the quality and quantity of population­based data on dogs and cats if we are to make real gains to animal health nationally and internationally. As we try to understand similarities and differences in health issues across regions; as breeding advisors and animal breeders try to make the best decisions to improve health and welfare; as veterinarians try to understand and explain risk to their clients; as any stakeholders try to design and monitor the effect of health interventions, there are calls for more and better information quantifying the occurrence of disease across various populations and subpopulations. This presentation explores why we do not have all the answers and how we might start to get them by highlighting the basis of the evidence and the gaps underpinning a more quantitative approach to understanding health and disease, existing and developing sources of data and the need for collaborative development across stakeholder groups, internationally and the roles each of us has to play.

Historically, many belief systems in companion animal health were propagated with heavy reliance on personal experience and expert opinion. The memory and perception of, e.g., private practitioners or even the most experienced breeders cannot be expected to produce accurate estimates of disease occurrence that relate to a wide population of animals. Expert opinion is sometimes called eminence based veterinary medicine and relies on the personal opinion of recognised experts or self-appointed commentators, often with minimal explicit critical appraisal applied to the quality of this opinion. Although expert opinion has been widely accepted as highly persuasive and reliable, it too is primarily anecdotal and, therefore, among the weakest type of evidence unless it is underpinned by a solid and stated evidence base.1 This is because many cognitive biases are inevitably inherent within the belief systems of any individual; and these explain why opinions across experts or interest groups so often disagree. However, we are now seeing increasing demands to challenge existing beliefs and position statements with the call for evidence: 'Show me the numbers.'

As we embrace this new enlightenment of evidence-based veterinary medicine (EBVM),2 however, there are new and exciting challenges that include but are not limited to:

  • Better data collection and curation methods
  • Identifying and comparing data sources
  • Linking databases/Collaborative research projects/'Jigsaw' projects
  • Understanding the uses and limitations of various data collection methods and sources - from research to veterinary practice data to breed health surveys
  • Standardised terminology (e.g., SnoMed, VeNom, PETscan, Agria)
  • Knowledge of analytical methods
  • Dissemination of information
  • lnclusivity for all stakeholders
  • Awareness of opportunities missed
  • Prioritisation of data requirements
  • Understanding of the risks from Poor data versus No data

The adoption of a more quantitative, evidence-based approach requires not only knowledge but also a change in attitudes and approach. Understanding breed-specific data such as prevalence and risk is both a science and an art. The science is the generation of appropriate data of good quality from a representative population and the extraction of reliable meaning from these data. The art comes in the communication and application of these data to improve animal health given both the uncertainty that inevitably accompanies this information and the range of stakeholders who must put findings into action. As we move into the era of Big Data for companion animal health, it is critical that everyone with a serious interest in companion animal health become comfortable and confident in both the science and the art of Numbers.

The first requirement is to identify useful sources of these Numbers. Access to sources of data on large counts of cats and dogs that are reasonably typical of the general or a specific population and that provide reliable demographic and health data are critical. There are no perfect resources for such data and each resource has its opportunities and challenges.3 The optimal approach may be to combine the strengths from a number of resources to answer any one research question while mitigating the drawbacks from each resource. Examples of data resources that have proven useful to date include:

  • Insurance sources4
  • Primary-care veterinary practice data5
  • Referral veterinary practice data6
  • Breed and kennel club health surveys7
  • Cancer registries8

However, the presence of data is just the first step on a long road from data to positive health impact for companion animals. The next step is to develop sustainable systems that enable these data to be curated and processed such that they are elevated from raw primary data into high-enough quality data for research. Many of the examples cited in the previous paragraphs have developed methods for such data processing and presentation, but it is often at this step that other defunct data collection concepts have failed to deliver usable outputs. Other critical steps on this knowledge journey include a requirement for researchers with an ability and motivation to analyse these data, funding to support this work and leaders with the vision to coordinate and direct this research. And, as with all study, it is crucial the right questions are asked if there is any hope that findings will be applied to and have an impact on real world problems.

When it comes to the point where we try to use such data to ensure healthy dogs and cats, the power of effective collaboration becomes hugely influential. Researchers may be the main players in information generation, but it is breeders, owners, the media and the general public who control the decisions on which animals to breed, which breeds get bought and how these animals are cared for. In addition to determining the type of information needed, dissemination now becomes critical; the skills and effort required to achieve these goals necessitate the participation of not only researchers but a wide range of other contributors. It must be an ongoing team effort; collectively we must develop the tools that can support these efforts.

Generation of valid data on diseases that affect breeding and effective use of such data to ensure healthy dogs and cats require an overall strategic approach that focuses effort towards important and achievable goals. A piecemeal approach where multiple stakeholders carry out independent and disparate bodies of work is likely to be highly inefficient, no matter how hardworking and well-meaning these stakeholders may be. Events like this meeting that precedes WSAVA 17 and the recent IPFD 3rd International Dog Health Workshop, where representatives of various stakeholder groups come together, can promote a collaborative force that can address mutual objectives for the health and welfare of dogs and cats.


1.  Holmes MA, Ramey DW. An introduction to evidence-based veterinary medicine. Veterinary Clinics of North America: Equine Practice. 2007;23(2):191–200.

2.  Holmes M. Practice-based clinical research: an introduction. In Practice. 2009;31(0):520–3.

3.  O'Neill D, Church D, McGreevy P, Thomson P, Brodbelt D. Approaches to canine health surveillance. Canine Genetics and Epidemiology. 2014;1(1):2.

4.  Egenvall A, Nødtvedt A, Penell J, Gunnarsson L, Bonnett BN. Insurance data for research in companion animals: benefits and limitations. Acta Veterinaria Scandinavica. 2009;51:42.

5.  O’Neill DG, Church DB, McGreevy PD, Thomson PC, Brodbelt DC. Prevalence of disorders recorded in dogs attending primary-care veterinary practices in England. PLoS One. 2014;9(3):e90501(pages 1–16).

6.  Bellumori TP, Famula TR, Bannasch DL, Belanger JM, Oberbauer AM. Prevalence of inherited disorders among mixed-breed and purebred dogs: 27,254 cases (1995–2010). Journal of the American Veterinary Medical Association. 2013;242(11):1549–55.

7.  Adams VJ, Evans KM, Sampson J. Wood JL. Methods and mortality results of a health survey of purebred dogs in the UK. Journal of Small Animal Practice. 2010;51(10):512–24.

8.  Brønden LB, Nielsen SS, Toft N, Kristensen AT. Data from the Danish Veterinary Cancer Registry on the occurrence and distribution of neoplasms in dogs in Denmark. Veterinary Record. 2010;166(19):586–90.


Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Brenda N. Bonnett, DVM, PhD
B Bonnett Consulting
International Partnership for Dogs (IPFD)
Georgian Bluffs, ON, Canada

Dan G. O'Neill, MVB, BSc(hons), GPCert(SAP), GPCert(FelP), GPCert(Derm), GPCert(B&PS), MSc(VetEpi), PhD, MRCVS
Veterinary Epidemiology, Economics and Public Health
The Royal Veterinary College
North Mymms, Hatfield, Hertfordshire, UK