Radio-Canada takes a bite out of pseudoscience

freedom-glassesWhen reporters are tasked to investigate stories, including dog bite stories, they are looking to include a few key things in their rush to deadline. They look for juicy quotes. They look for digestible statistics. And they try to make a nod to “both sides” of an issue, whatever they define those to be.

So when they seek information to support the position that “risk factors exist to predict aggression but breed is not one of them” there are a lot of sources to use – advocacy groups, veterinary/animal control/animal behaviour professional organizations, peer-reviewed research, and these days, pretty much anyone who works first-hand with dogs.

When they are looking for something to support the argument that “certain breeds of dogs are inherently more dangerous” they find one source on the internet. They might find a few sites, but they all lead back to two related groups. These groups manufacture their own “studies” and generate “statistics” that sound horrifying. It doesn’t take much to scratch the surface and see that the numbers don’t make sense, but few people take time to do so, and these numbers are seen in reporting and in BSL discussions everywhere. The truth is, NO reliable numbers exist for overall bite rates, serious bites, or breed and bites, because this is not tracked systematically. Anywhere.

Alas, in the wake of the proposed Montreal breed ban, one news outlet took the time to really dig into the phenomenon of these soundbite-oriented statistics. Bouchra Ouatik of CBC Radio-Canada dug into these numbers and the characters who generate them. The article has been shared widely but until now, only a Google Translate version has been available. One of our bilingual volunteers has been kind enough to translate it for us for easier reading. Please read, share, and click on the original link so the article gets the hits it deserves!

Pit bulls: Non-scientific data frequently quoted by media.
Friday September 9th, 2016
Bouchra Ouatik

Statistics on dog bites, coming from anti-pit bull groups, are often quoted in Canadian and American media as being reliable sources. These figures are however very far from reality.

The two groups in question, Animals 24-7 and DogsBite.org, often campaign openly to ban pit bulls. They regularly publish statistics on deaths and bites caused by dogs. However, their figures only represent a tiny portion of serious attacks, and those from Animals 24-7 contain many errors.

A tiny portion of serious attacks.

The author of the Animals 24-7 site, Merrit Clifton, publishes data every year on the number of attacks by dogs in Canada and the United States. The group’s most recent report claims to account for almost every serious attack having taken place between September 1982 and September 2016, a period of 34 years. M. Clifton claims to get his data solely from media reports, but he maintains that he has a comprehensive picture of the situation.

According to this report, during the 34 years studied, there would have been, all breeds included, 5756 dog attacks causing serious injury, 4194 attacks having mutilated or disfigured the victim, and 652 deaths. The author defines serious attacks as being those where the victim has been killed, mutilated, or has received wounds necessitating serious medical care. According to these same figures, pitbull type dogs were responsible for 78% of serious attacks, 70% of attacks causing mutilation or disfigurement of the victim, and 53% of deaths.

This would mean there would be on average 169 serious attacks and 19 deaths every year in Canada and the United States.

But these numbers are very far from reality.

 

According to the Agency for Healthcare Research and Quality, in 2008, in the United States, almost 9500 people have been hospitalized due to dog bites, and roughly 40 died from their injuries, or substantially more than the numbers put forth by Animals 24-7. Of this amount, there were 600 fractures, 300 surgeries requiring skin grafts, and 1100 operations on muscles or tendons. These figures however do not mention the breeds of the dogs involved.

Even so, in only one year, more people have been hospitalized for serious injuries due to dogs than Animals 24-7 has compiled in 34 years.

Deaths and serious injuries caused by deaths

Category Animals 24-7
(annual average)
Agency for Healthcare Research and Quality (for 2008)
Attacks requiring hospitalization 169 9500
Deaths 19 More than 40 (only in hospitals)

The figures from the American Society of Plastic Surgeons follow the same direction. In 2015, in the United States, there were 28079 reconstructive surgeries for dog bites, and in 2000, that number rose to 43089. We should also note that a person can require more than one surgical procedure for the same injury.

This important discrepancy between the figures from Animals 24-7 and reality can be explained in part by the fact that the majority of dog attacks are not reported in the media. The owner of Animals 24-7, Merrit Clifton, has pointed out via email that it is possible that he has missed a few cases of serious injury.

Meanwhile, the DogsBite.org site regularly publishes the numbers from Animals 24-7, in addition to maintaining its own list of incidents. Colleen Lynn, the founder of DogsBite.org, herself a bite victim, explains that her site does not claim to tabulate all serious dog bites, but that it tabulates the majority of deaths. She also claims that her site does not only rely on media reports, but also draws its figures from other sources, such as police reports, and that her site sometimes publishes cases that have not been mentioned in the media.

Several inconsistencies

The figures from Animals 24-7 underestimate the attacks perpetrated by breeds other than pit bulls. For example, according to this group, in 34 years, throughout the United States and Canada, there were 66 serious attacks by chow-chows, and 67 by Labradors. However, according to the Texas department of health, in 2000, in that state only, there were 67 severe bites from chow-chows, and 39 from Labradors.

In Animals 24-7’s report, the same breed of dogs can be found in many categories, or several different breeds can be grouped in a single category. For example, the author considers the Presa Canario and the bull-mastiff as a single breed which he counts in the same category. In reality, these two dogs are of different origins, and even have a different physical appearance. The Presa Canario saw the light of day in the 15th century in the Canary Islands, whereas the bull-mastiff appeared at the end of the 19th century in England.

Conversely, we find a category for the “Blue Heeler”, one for the “Australian Blue Heeler”, and another for the “Queensland Heeler”, while these all refer to a single breed of dog, the Australian Cattle Dog. We can also see breeds of dogs that do not exist, such as the East Highland terrier.

Merrit Clifton also claims on his site that the Cane Corso is a cross between pit-bull and mastiff, which is impossible, since the Cane Corso is an Italian dog which has existed since Ancient Rome, whereas the pit bull saw the light of day in England in the 19th century.

Additionally, when many dogs are involved in an attack, it is possible that the incident is classified solely in the pit bull category. For example, in 2007, an 18 year old man was attacked by 13 dogs, including one pit bull and 12 of unidentified breed, but this attack is regardless counted in the pit bull category.

Statistics from classified ads

In his report on dog attacks from 1982 to 2016, Merritt Clifton indicates what proportion each dog breed makes up in the total population of dogs. However, there is no dog census throughout the United States and Canada, and the mandatory registration imposed by municipalities is not always observed by owners, which makes it impossible to know the population of each dog breed in North America.

Hence, to estimate the dog population, Merrit Clifton explains that he checked the dog sales classified ads on various web sites for the month of July 2016. He concludes that pit bulls represent 5% of dogs and he considers that this number is representative of dog populations throughout Canada and the United States for a period of 34 years.

Merrit Clifton replies that he has proceeded this way because there is no other database of dog populations. “In Canada and the United States, only 10% of dogs were registered in 1982 and it’s roughly 25% today. No American city has a registration rate above 40%”, he explains.

Deaths indirectly attributed to dogs

The Animals 24-7 site sometimes considers that a person has been killed by a pit bull even if the dog was only indirectly involved. These cases are counted in the same category as those where the dog has bitten the victim, even when the coroner maintains that the dog was not responsible for the death.

Here are some cases that Animals 24-7 considers as deaths caused by pit bulls:

  • In 2009, in Wisconsin, Louanne Okapal, a 55 year old woman, died after having been struck in the face by her horse. The horse had been frightened by a pit bull.
  • In 2009, Teresa Foss, a 48 year old woman, died from a head wound after having been knocked over by a pit bull. The dog had not bitten her.
  • In 2010, Richard Martratt, a 64 year old Texas man, stabbed a pit bull and slaughtered a Catahoula because the two dogs had attacked a border collie on his property. The man was not attacked by the dogs, but when the authorities arrived, he collapsed and died from a heart attack.
  • In 2010, in Georgia, Miracle Parham, a 14 year old teenage girl, fled after having been frightened by a dog that witnesses described as a pit bull. She was fatally struck by a car.
  • In 2013, James Harding, a 63 year old man, was struck by a car after having attempted to get away from two pit bulls.
  • A 6 year old girl was strangled by a chain to which a pit bull was tethered. The year and location were not specified.

Another case mentioned on DogsBite.org deals with James Chapple, a 57 year old Tennessee man who was seriously injured by pit bulls in 2007. Four months later, he died of atherosclerosis and complications from alcoholism. Despite that, DogsBite.org counts him as a death caused by pit bulls.

At odds with the scientific community

Both groups are very critical of scientific experts. The DogsBite.org site even goes as far as using the term “science whore” to describe some experts. Colleen Lynn, the site founder, defends herself by saying that the term did not originate from her and that it was only used on three occasions since the creation of the site in 2007.

In addition, as opposed to studies published in scientific journals, the statistics from these groups are not reviewed by independent experts in order to verify their validity.

Karen Overall, researcher in veterinary medicine from the University of Pennsylvania, has analyzed all the studies of dog bite statistics published between 1950 and 2000 throughout the world. Her research shows that the dog breeds responsible for the largest number of attacks vary based on the year and the region studied.

She is critical of the methodology used by the groups who depend mainly on media stories. “The reports by the media and by the police are almost always incomplete, she says, and there is no independent confirmation of the breed involved. These publications use these reports as if they were infallible.”

CORRECTION:
A previous version of this article erroneously reported that Animals 24-7 had tabulated over 34 years, 7045 dog attacks causing serious injury, 4424 attacks having mutilated or disfigured the victim, and 657 deaths and that pit bulls would be responsible for 64% of serious attacks, 66% of attacks having mutilated or disfigured the victim, and 51% of deaths. Regardless, this in no way changes anything in the analysis of it which was done nor in the conclusions we have reached.

Translation by Phil Boutros.

For more information on this topic, visit the HugABull website.
Statistics and soundbites
– Breed and bites

This entry was posted in Uncategorized. Bookmark the permalink.