20 reasons why improving breed health is so difficult

“Cognitive Bias” is a term that captures a multitude of reasons why it’s so hard to get people to see the need for improvement, let alone make the necessary changes that will improve the health of dogs. It’s also referred to as”Cognitive Dissonance”.

Breeders and exhibitors make many decisions every day about thei dogs. Everyone likes to think these are rational, but maybe things aren’t quite that simple.

Here are 20 types of Cognitive Bias which I read about in “Business Insider” and have interpreted for canine health.

1. Anchoring Bias: People rely on the first piece of information they hear. In a conversation about a particular health condition, the first person to comment on its prevalence sets the scene for everyone else’s views on the problem. “I’ve never seen it in 30 years of breeding” will anchor everyone in a mindset that it really can’t be an issue.

2. Availability Heuristic: People overestimate the importance of information that is available to them. Someone might argue that Cavaliers are not prone to heart disease because they know of a dog that lived to 15 with no disease.

3. Bandwagon Effect: The probability of one person adopting a view increases based on the number of other people who also hold that view. This is one reason why some stud dogs end up being so-called Popular Sires.

4. Blind-spot Bias: Failing to recognise your own cognitive biases, is a bias in itself. The classic manifestation is Kennel Blindness (“My dogs no longer have any faults”).

5. Choice-supportive Bias: When you choose something, you usually feel positive about it, even if you know it has problems. Choosing a stud dog to mate with your bitch often comes with this bias.

6. Clustering Illusion: This is the tendency to see patterns in random events, like the idea that red dogs of a particular breed are more likely to be aggressive.

7. Confirmation Bias: We tend to listen more to information that confirms our existing perceptions. This is perhaps one of the reasons why it is so hard to have a rational conversation about outcrossing as a means to improve genetic diversity and health.

8. Conservatism Bias: This is where people are slow to accept new evidence, for example the VetCompass data that shows, on average, crossbreed dogs live longer than pedigree dogs.

9. Information Bias: This is the tendency to seek more information rather than taking action. How much more data do people need before they get the message that high levels of inbreeding increase the risks of harmful mutations emerging and reduced levels of fertility. I’ve said before ‘if you wait for the perfect set of data, you will wait a very long time’. Sometimes, it’s easier to make a decision with less information.

10. Ostrich Effect: This is the decision to ignore dangerous or negative information by burying your head in the sand, like an ostrich. People who have invested time and effort in building a “line” of dogs are likely to be less inclined to acknowledge diseases or problems that can be traced back to their pedigrees.

11. Outcome Bias: Judging a decision based on the outcome, rather than on how the decision was made. Just because you bred a “healthy” dog with a Coefficient of Inbreeding of 30% doesn’t mean it was a smart decision.

12. Overconfidence: Some people are too confident about their abilities and this causes them to take greater risks. Experts are more prone to this bias than lay people, since they are more convinced they are right. Experienced breeders might believe they can “safely” mate two merle (dapple) dogs and not end up with deaf or blind puppies.

13. Placebo Effect: This is when simply believing something will have an effect causes it to have that effect. According to a study by Conzemius & Evans, a caregiver placebo effect by both dog owners and vets was common in the evaluation of patient response to treatment for osteoarthritis. Half the owners whose dogs received placebos stated that their dog’s lameness was improved during the study.

14. Pro-innovation Bias: People with new ideas often over-value their usefulness and under-value their limitations. Just because a new DNA test has been developed, doesn’t mean it’s important to use it, particularly if the mutation frequency is extremely low, or if the welfare impacts of the condition are minimal.

15. Recency: The latest information you receive is often weighed more heavily than older information. Conformational exaggerations seen in the ring today may be admired and rewarded, rather than remembering a breed’s original purpose and type.

16. Salience: This is the tendency to focus on the most easily recognisable features of a situation. For example, in Dachshunds it is much easier for breeders to focus on eradicating cord1 PRA where there is a DNA test than on reducing back disease which is a complex condition, with no “simple” test. The fact that, statistically, Dachshunds are more likely to suffer from back problems than to go blind, may be overlooked just because a test is available for PRA.

17. Selective Perception: This is where we allow our expectations to influence how we perceive the world. If we “know” that a particular line of dogs is prone to a particular health condition, we tend to look for more examples to prove that case, rather than looking more widely across the breed. We end up with “Mrs Miggins’ dogs produce xyz disease”.

18. Stereotyping: This is where we expect a particular situation without having any real evidence. Just because one of Mrs Miggins’ dogs is aggressive, doesn’t mean all her dogs have a bad temperament. People tend to over-use and abuse the limited evidence available.

19. Survivorship Bias: This is an error that comes from focusing only on surviving examples. For example, we might think that Mrs Miggins’ dogs are long-lived, because we haven’t heard of any of her dogs that have died at an unusually young age.

20. Zero-risk Bias: Sociologists have found that we love certainty, even if it’s counter-productive. That’s why many breeders are obsessed with “health-testing” and the continual search for new tests. They should, instead, be focused on the root cause of the problem which is closed stud books and high levels of inbreeding.

All of these cognitive biases are potential stumbling blocks that affect our behaviour and they can prevent us from acting in the best interests of our dogs and our breeds.

However, if we are aware of them, we can turn some of them to our advantage. The way we present information; how we communicate good practices; how we reward and reinforce improved canine health; all these can nudge people in the right direction.

.

.

.

.

.

.

.

.

.

.

.

.

.

Advertisements

Brachycephalics: Making the move from data to improvement?

It’s been interesting following the information emerging from the various discussions on brachycephalic breeds. We’ve heard from vets calling for action to address the health issues, including via online petitions. The Kennel Club in Norway has set out its proposals for improvement and our own KC has convened a working group. There’s also the CRUFFA campaign to discourage the use of images of flat-faced animals in advertising and the media. All this follows on from the RVC’s “Building better Brachycephalics” day in 2013.

If you’ve not seen them, it’s well worth heading to vet Pete Wedderburn’s Facebook page to watch the videos he live-streamed of the various (excellent) presentations made at the first meeting chaired by Steve Dean at Clarges Street. From comments in one of the videos, it appears that it came as a surprise to some attendees that the meeting was being live-streamed by Pete. The presentations made by the scientists clearly summarised the evidence for the breadth and scale of the health problems facing brachycephalic breeds, both at individual dog level and at population level. The evidence is indisputable and the work done by David Sargan and his colleagues at Cambridge University means there are now practical ways to measure and score the health impacts in individual dogs.

The focus of that first meeting was very much on data and “the science”, with less of a discussion of the factors that have (a) led breeders to produce health-compromised dogs or (b) caused such a massive increase in demand from the puppy-buying public. The demand issue is clearly an area of focus for the CRUFFA campaign.

There was a second meeting at the KC at the end of July, but I believe Pete wasn’t present, so there are no videos to watch. In addition to the scientists, these KC meetings have included Breed Health Coordinators such as Penny Rankine-Parsons (FBs) and Vicky Collins-Nattrass (Bulldogs), both of whom have been incredibly proactive in their breed health improvement work.

At the end of the first meeting, participants were asked to go away and draw up an A4 page of actions they felt could/should be taken. Apparently, they were asked not to put “change the Breed Standards” at the top of their lists. Pinning the blame, and focusing the actions, on the KC and show communities is far too narrow a perspective if we want to improve the health of these dogs. Overall, the good news is the brachycephalic problem is moving into solution mode.

Complicated or Complex?

What interests me is how this will be managed as a Change Programme. Doing the data analysis and the science may be complicated but there are some world-class people working on these aspects. However, making change happen is complex (rather than complicated) and, the knowledge and skills needed are totally different, particularly when it comes to changing buying behaviours in the wider population.

I deliberately used the words “complicated” and “complex”. It is important to understand the difference between “complicated” and “complex” situations. The complicated context calls for investigating several options where there may be multiple “right answers” and is the domain of subject matter experts, like the scientists working on brachycephalic health. One of the dangers is that innovative suggestions made by non-experts may be overlooked, or dismissed. The voices of the Breed Health Coordinators with their wealth of practical experience need to be heard. Another risk in complicated situations is “analysis paralysis”; the tendency to keep searching for the perfect set of data, or the perfect answer to a problem, which means that very little gets implemented. Decision-making in complicated situations can take lots of time and there’s always a trade-off between finding the “right answer” and simply making a decision in order to make some progress.

When it comes to implementing changes to improve brachycephalics, the situation is complex; there are no right answers. We already know from the science that the issues are not even the same in the different brachycephalic breeds. David Sargan was reported on the BBC in response to the paper published on Bulldog genetic diversity and he said “we now have pretty strong evidence that there are still multiple genetic variations between those that do and those that don’t suffer from the disease (BOAS). But, we do not know whether this is also true for other aspects of conformation and appearance related diseases.”

There are bound to be many competing ideas and what will work is likely to emerge from a range of innovative approaches. There are lots of different people who have to be engaged and whose behaviours have to change. We shouldn’t underestimate the challenge of reaching and influencing the large number of breeders outside the KC/Breed Club communities. There will be a need to encourage dissent and diversity of ideas, as well as a willingness to “just try stuff” and see what works. That’s probably going to be uncomfortable for some people, particularly if they prefer working in a world of “right answers”, predictability and hierarchical decision-making.

We need to stop reacting to individual reports and look at the whole picture. Somebody needs to be joining the dots, otherwise we just add to the doom and gloom feeding frenzy in the press.

Agile or Big Bang?

What is the strategy for change with brachycephalics? Will it be exploratory and agile, or will it be a “big bang” launch and roll-out of a “package” of solutions? If it’s the former, then it would be perfectly valid to implement a change to a Breed Standard and see what happens. It’s a simple decision to make and it will either make an impact on its own, or not!

The trouble with that one, simple decision, is that we know it will not be enough on its own. But, it could be implemented quickly and could be seen as part of what Dave Brailsford, the Team GB Cycling Director, called the concept of marginal gains. Brailsford believed that if it was possible to make a 1% improvement in a whole host of areas, the cumulative gains would end up being hugely significant. The successes of Team GB and later Team Sky clearly demonstrate the power of this approach.

There were already a few ideas being touted around on social media before the second meeting hosted by the KC. Each of these has a cost and a potential value (or impact), so their relative merits need to be evaluated. The speed with which they could be implemented also needs to be agreed. Here’s my view of what a cost-value map might look like for a few of the ideas I read about. Green ideas could probably be implemented quickly, Orange ones would take longer and Red ones would be much longer-term.

Building Better Brachycephalics 2

The good thing is that the ideas cover both the supply side and demand side of the problem. They also contain a mixture of small changes and big changes. “Change the Breed Standards” is a small change, whereas “Educate the public” is a big change. The latter cannot actually be implemented; it needs to be broken down into doable activities like “run a series of campaigns on TV”, “get celebrity owners to talk about their pets’ health issues”, or “produce posters to display in all vets’ waiting rooms”.

What struck me about the lists of ideas I saw was just how few ideas there were. That’s possibly just a reflection of the mix of big and small ideas. Linus Pauling, the American scientist said “the best way to have a good idea, is to have lots of them”. There are certainly plenty of keyboard warriors willing to share their views online; how about building that into the solution-generation stage of the Brachycephalic improvement programme? Maybe there’s an opportunity to “crowdsource” more ideas. Just a thought!

 

New systems approach needed to improve dog health

The following article is reprinted with permission from Dog World (published 14/5/15).

MANY, if not most, canine health and welfare problems are linked to people, their behaviour and attitudes. And the issues surrounding such problems are far more complex than have been argued in recent years.

  So said Philippa Robinson at the British Small Animals Veterinary Association’s recent congress, adding that ‘finger pointing’ was no longer helpful and blame counter-productive.

  She suggested a new approach to combat health problems in pedigree dogs, and said the demand for and supply of them needed to be understood for things to change.

  All the agencies and stakeholders involved needed to work together to clarify what health and welfare messages were needed, she said.

  Mrs Robinson discussed the controversies surrounding inherited disease and how she joined the ‘Pedigree Dogs Exposed campaign’ in 2007, a year before the documentary was broadcast.

  Mrs Robinson of the Karlton Index, which was launched to monitor and measure canine health, spoke of the three reports on pedigree dog which followed PDE, health including the Bateson enquiry, and the setting up of the Dog Advisory Council.

  After the Bateson report, she said, she began examining ‘how inert the Kennel Club had been on dog health’.

  “But the grip of the anti KC rhetoric began to loosen its grip on me,” she said. “An historical analysis of pedigree dog health reveals that the issues are far more complex than argued in either PDE or any of the three subsequent reports.”

  The KC had launched many initiatives regarding health and welfare including collaborative work with other parties, she said.

  Why had not other agencies taken a stand on canine health, she asked. And if they did speak out what had been the consequence?

  At that point, she said, she decided that canine health problems had not been caused by one stakeholder.

  To those present she recommended a ‘systems thinking’ approach, and looked at supply and demand of dogs and puppies, the fact they could be obtained from many different sources by people with different motives, levels of commitment and sense of duty. She talked about breed type rather than breed, and said she believed England was not a nation of dog lovers but one of dog breed lovers.

  “For whatever, complex, reason, we develop a fondness for specific types of dog,” she said. “This has resulted in each breed having very specific societal and cultural contexts.”

  She discussed the shape of the Bull Terrier’s skull, saying the breed Standard had not driven the change in shape or which dogs were awarded in the show ring.

  “Those factors may contribute to the changes of shape, undoubtedly, but the real influence I would argue is simply human preference,” she said.

  The KC could change the breed Standard to reflect the top shape of skull, she went on, and judges could begin to only award dogs with the top shape, but that would not stop people choosing the dog with the skull shape they preferred, even if it was detrimental to health.

  Mrs Robinson turned to brachycephalic breeds such as the French Bulldog saying that publicity about the breed’s health problems had had no effect on the explosion in its popularity, which was fed and encouraged by celebrity owners.

  “The French Bulldog has become a cultural icon,” she said. “It has been used to sell all sorts of merchandise, services and it is, of course, the current breed of choice for many a celebrity. This breed endorsement has not come from the KC, the show world or the breed club. Even combined, the KC, the show world, the breed club have precious few resources to counter that iconic status. They also have limited spheres of influence to change behaviours and attitudes among the wider population.”

  So how can health and welfare of the breed counter this cultural phenomenon, she asked.

  “Review the breed Standards? Introduce judicious health testing Remove untested dogs and affected dogs from the breeding programme?

  “Remove untested dogs from showing and remove untested dogs from the KC system altogether? Ensure that the public is armed with facts? Issue breeds with health warnings like cigarettes?”

  But the bigger picture needed to be considered, she said, the trends analysed and the points of intervention which would provide maximum leverage identified.

  Much had been done by the KC to improve the French Bulldog’s health, Mrs Robinson said, but most French Bulldogs were being bred away from that system by people who did not take part in health schemes.

  “Some are imported, most often illegally, to fulfil the demand created through the celebrity culture,” she said. “So without wider support of the stakeholder community, without the injection of resources from more than one stakeholder, all the valiant work will struggle to have an impact on the French Bulldog population as a whole.”

  How could the cultural phenomenon be changed, she asked. How could mind sets and contexts be changed? It had to be recognised that people make irrational choices because they are motivated by ‘a complex set of drivers’.

  The veterinary profession, welfare charities, scientists, academia, breeders, local authorities, the KC, welfare campaigners, the Associate Parliamentary Group for Animal Welfare, industry and the public need to work together, she said, to clarify what the health and welfare messages needed to be for each breed and breed type based on evidence and good data. The dynamics of the system which supplies dogs and puppies needed to be understood, as did the human behaviour which determined dog buying and acquiring decisions.

  Key messages needed to be delivered consistently across the board and by all parts of the system.

  “Finger pointing is no longer helpful and blame is counterproductive,” Mrs Robinson said.

  “Meaningful dialogue and courageous and creative action are the things we should be working on jointly.

  “The good news on that is courageous action and creative solutions such as VetCompass, estimated breeding values, genetics and breeder education can be really exciting projects in which to get involved. So let’s stop sniping and start sharing.”

Modelling the UK Dog Population – #OR56 conference presentation

Alessandro Arbib (DECC) made this presentation (see Slideshare below) to the Operational Research Society’s 2014 conference OR56.  Abstract:

2014-09-09 14.46.17-1 (Custom)The breeding, ownership and welfare of dogs in the UK is a complex social area. Although there has been research into the size of the dog population, nobody has pulled all this together into a single model that everyone can use to help focus priority issues. A consensus understanding of the population and how it is stratified is crucial to allow proposing meaningful welfare improvement policies. From November 2013 to May 2014 a group of 3 OR analysts and an engineer from DECC worked with the RSPCA (the UK’s leading animal welfare charity) and Dog-ED (a Social Enterprise applying Systems Thinking to canine welfare) to provide analytical evidences about the number of dogs currently present in UK and how they move through the system. The project involved a significant literature review to collect the data necessary to produce a snapshot of the UK dog population; designing and building a “stocks and flows” model to investigate the flows of dogs from the different categories; and developing recommendations for possible uses and future development of the model. Lack of consensus amongst the data sources, and considerable variation in data quality and definitions used made it difficult to provide accurate answers to the customer’s problem. We will describe our main outputs including estimated upper and lower bounds for the dog population, a “stocks and flows” model developed in Excel, and a list of the main data gaps and issues we met in our work. Last but not least, we will focus on the valuable experience of working for the Third Sector, summarising the main lessons learnt and the value that OR was able to add in this area.

.

.

..

.

.

.

.

.

.

 

Modelling the UK Dog Population – summary report

The breeding, ownership and welfare of dogs in the UK is a complex social area. Although there has been research into the size of the dog population, nobody has pulled all this together into a single model that everyone can use to help focus priority issues.  As a consequence, different stakeholders have varying, and sometimes conflicting, views of how many dogs there are and their needs.  Without a consensus understanding of the population and how it is stratified, it is difficult to propose meaningful welfare improvement policies.

Earlier this year the RSPCA and Dog-ED worked with a group of Operational Research analysts from the Department for Environment and Climate Change (DECC) to review the literature and establish a baseline of data on the UK dog population.  The DECC team did this for us as a pro bono project coordinated by the Operational Research Society, of which Ian Seath is a member.
The project case study (download pdf) summarises the results of the literature survey and the challenges the DECC team faced when trying to build a population stocks and flows model (example below).  The DECC team will be presenting a paper at this year’s OR Conference, based on this work.
Stocks and Flows

Conclusions:

  1. Top-down and bottom-up calculations of the UK dog population do not agree, resulting in a significant range (8.5 – 11+ million)
  2. There is insufficient data from publicly available sources to quantify the origins and populations of non-KC registered pure-bred dogs (e.g. “hobby breeder”, “commercial” or “puppy farm”)
  3. There is insufficient data to understand the reasons why dogs are relinquished and go into welfare, or to identify the extent to which dogs in welfare may be moving in and out of “revolving doors”
  4. The lack of data makes it too difficult to identify additional areas (over currently known points) where interventions could occur to improve the welfare of dogs
  5. Forecasts of the potential impacts of different interventions are dependent on external factors (economic and societal) which are themselves difficult to predict
  6. The DECC OR team has brought a rigorous and disciplined approach to this project and highlighted the data and evidence challenges that exist in this complex social policy area

 

Download: Understanding the UK Dog Population

.

.

.

..

.

.

.

.

.

.

 

What would an effective Breed Health Improvement Strategy look like?

This presentation is based on work we have done to help develop Breed Health Improvement Strategies. It provides a framework for strategy development and gives examples of the elements required within a comprehensive strategy. The framework can be used to help pull together existing approaches and to ensure they are aligned.

.

.

.

.

.

..

.

.