Page 478 - WSAVA2017
P. 478

An Urban Experience
D. O’Neill1
1Royal Veterinary College,
Veterinary Epidemiology- Economics and Public Health, Hatfield, United Kingdom
Dr Dan 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, Hawkshead Lane, North Mymms, Hatfield, Hertfordshire AL9 7TA, United Kingdom
Our understanding of breed-specific prevalence and
risk is both a science and an art. The science is the generation and identification of the prevalence and
risk values, and the art is the uncertainty that inevitably accompanies these data. To optimise welfare and other gains from the application of breed-specific data, we need to be comfortable and confident in both the science and the art.
Good science requires accurate and precise
definitions. Prevalence describes the number of instances of disease (numerator) in a defined population (denominator) at/during a designated time and makes no distinction between old and new cases (1). Prevalence
is usually reported as the percentage or proportion of the denominator that was affected. Point prevalence refers to the amount of disease at a particular point in time whereas period prevalence refers to the amount
of disease that occurred during a specified period of time. Incidence the number of new cases that in a defined denominator population over a specified period of time (1).
Risk is subtly different in that it defines intentional interaction with uncertainty going forward. Risk describes the number of newly affected individuals over a specified period of time (numerator) divided by the total population at risk (denominator), although these expectations are often derived mainly from previously collected prevalence or incidence results (2). Risk can be reported in absolute terms as a percentage or proportion, or in relative terms as a relative risk or risk ratio. For example, we may say that the absolute risk (odds) for the next entire female
French Bulldog who presents at a UK emergency care clinic being a dystocia case is 20.6%. Alternatively, we can say that this risk (odds) is 15.9 times higher than if that next entire bitch was a crossbred (3). The absolute risk tells the size of the risk and the relative risk compares it to some other standard to put this risk into context.
As a creation of man’s desire, breeds constantly
evolve according to new information and fashions with consequent changing health/disorder profiles over
time. This introduces uncertainty. With an average generation interval of 4.1 years in dogs (4), it is clear
that a study performed 20 years ago is describing dogs from 5 generations ago and therefore the results may no longer be valid for the current population. So we need to consider the recency of the studies.
Disorders are also rarely a yes/no binary proposition, much as we would like to believe in this simplistic
ideal. Most disorders exist on a continuum across dimensions that include onset, severity, welfare impact and diagnostic thresholds. Indeed, many disorders such as obesity are scored as a continuous outcome which then allow researchers to select cut-offs that assign disease status to varying levels of overweight and obesity (5). This introduces uncertainty.
The VetCompass Programme at the Royal Veterinary College in London collects anonymised clinical record data on 3 million dogs from a network of 500 UK practices (6). VetCompass research has identified that dogs are subject to diagnosis of a wide array
of disorders; 430 distinct disorder were recorded in
one study (7). Disorders with the highest prevalence overall were otitis externa (10.2%, 95% CI: 9.1-11.3), periodontal disease (9.3%, 95% CI: 8.3-10.3) and anal sac impaction (7.1%, 95% CI: 6.1-8.1); certainly not the high profile disorders that often get discussed in relation to canine health. Purebred dogs had a significantly higher prevalence compared with crossbreds for just three of the twenty most-prevalent disorders: otitis externa (P = 0.001), obesity (P = 0.006) and skin mass lesion (P = 0.033) (Table 1). The prevalence of five of
the twenty most-prevalent diagnosis-level disorders differed statistically significantly between popular breeds: periodontal disease (P = 0.002), overgrown nails (P = 0.004), degenerative joint disease (P = 0.005), obesity (P = 0.001) and lipoma (P = 0.003) (Table 2). This supports a substantial impact for breed on the disorder risk for individual dogs.
Recording and collection of accurate health and demographic data are essential for effective prioritisation and monitoring of important health issues across breeds. Individual veterinarians have an important role to play here. Large-scale data initiatives like VetCompass

   476   477   478   479   480