* Update: Since this blog was written, the ASPCA has released a position statement on behaviour assessments for dogs in shelters.
There’s growing evidence that universal canine behaviour assessments in shelters are not able to accurately assess a dog’s temperament, and are certainly unable to accurately predict how that dog may behave in a home environment. Of course we want to protect our staff and community and identify dogs in need of help. But is the current approach working or should we be doing things differently?
No better than flipping a coin
A 2016 article (1) pointed out that canine behaviour assessments are, statistically speaking, “no better than flipping a coin”. In the authors’ words: “Canine behavior evaluations lack an essential component of any valid diagnostic test because key attributes of test validity (sensitivity and specificity) have not been, and likely cannot be, calculated in the context of a research situation in real shelters and adoptive homes. Furthermore, there is neither consensus nor confirmatory research on the specific behaviors elicited during a provocative test in a shelter, the relevant intensity of those behaviors, or the frequency of those behaviors in the various subtests that would be considered indicative of a potentially dangerous dog.”
A 2013 study (2) reported that only 55% of dogs identified as exhibiting food guarding behaviour in a shelter were reported to show food guarding in their homes – indeed, no better than flipping a coin! Conversely, 22% of dogs labelled as non-food-aggressive were reported to show food guarding in their new homes. Probably the most striking finding was that adopters did not consider food guarding to be a particular problem in most cases.
What if shelters just stopped assessments for food guarding?
A new study (3) asked, “What if shelters stopped including food guarding (FG) assessments in behaviour evaluations?” They compared 3,451 dogs that had FG assessments (first phase) and 6,862 dogs that did not (second phase). The study included 9 shelters.
- 8% of dogs in the assessed group were diagnosed with FG while only 3% were diagnosed in the non-assessed group. In the assessed group, 80% of the FG diagnoses were made through the assessment, while in the other group all FG diagnoses were made through history and in-shelter observations.
- Bites and injuries were low overall.
- Dogs identified with FG in the assessed group had a 0.18% bite rate in the shelter and 2.2 % in the home and those without FG had a 0.46% rate in the shelter and 0.44% in the home.
- Dogs diagnosed with FG in the NON-assessed group had a 0.48% bite rate in the shelter and 3.2 % in the home and those without FG had a 0.42% rate in the shelter and 0.47% in the home.
- In both settings, there was a statistically higher rate of bites at home for dogs identified with FG behaviour than those without, but the overall number of bites was low, and the total number of dogs that bit was higher for the non-FG dogs (because there were many more of these dogs).
- The label of FG was associated with increased length of stay, and food guarding was an important reason for euthanasia.
- Dogs diagnosed as having FG behaviour were no more likely to be returned from adoption than other dogs.
- There was a 3% higher return rate during the phase when FG assessments were not done (13% vs 10%). This was statistically significant but the authors questioned its practical significance in the field.
What do we take away from this? The authors of the new FG study recommend that shelters educate all adopters about normal canine behaviour and offer support to those who encounter problems, rather than using FG assessments for all dogs. Going further, Patronek & Bradley (1) argue that the universal behaviour assessment is not a valid diagnostic tool and should be replaced with better approaches.
It would be rational for shelters to flag dogs that show signs of behaviour problems, the same way we run medical diagnostic tests (or should!) only where there is a specific indication. A behavioural history of aggression or extreme fear, for example, should trigger further evaluation. So should in-shelter observations of aggressive behaviour, problematic resource guarding or concerns about emotional welfare. Shelter resources could be used to evaluate these animals, minimize risk to staff and the public, and help dogs to deal with the stress of the shelter and find their new homes, rather than performing unvalidated (and unvalidatable) behaviour assessments on most or all dogs. Provocative testing, in particular, may have had its day.
To quote Drs. Patronek and Bradley again, “Instead of striving to bring out the worst in dogs in the stressful and temporary environment of a shelter, and devoting scarce resources to inherently flawed and unvalidated formal evaluations, how much more productive might it be to focus our energies on giving every dog the opportunity to be at his or her best?”
- Patronek GJ, Bradley J 2016 No better than flipping a coin: Reconsidering canine behavior evaluations in animal shelters.
J Vet Behav 15: 66-77
- Marder AR et al 2013 Food-related aggression in shelter dogs: A comparison of behavior identified by a behavior evaluation in the shelter and owner reports after adoption. Appl Anim Behav Science 148:150-156
- Mohan-Gibbons H et al 2018 The impact of excluding food guarding from a standardized behavioral canine assessment in animal shelters. Animals (Basel) 8:27