Total Pageviews

Thursday 31 August 2017

“Dangerous breeds,” dog bite statistics, and the Merritt Clifton report

Part 1: Numbers



Which breeds bite most often? Which breeds do the most damage when they attack?

Don’t ask the Centers for Disease Control.

Breed "is no longer considered to be of discernible value" when addressing dog bite prevention, according to a CDC spokesperson.

The most important factors affecting the odds of a dog bite or attack have always been the ones any dog person can rattle off: Puppy socialization. Training. Pack behavior. The dog's health. The dog's care and condition. Something as simple as never leaving a small child unattended with any dog. Size of the dog: when big dogs bite they generally do more damage than little dogs --- although that's a moot point if you're six weeks old.



If you want facts on dog aggression, read A Community Approach to Dog Bite Prevention, the AVMA’s groundbreaking 2001 task force report. [You’ll find it in the sidebar, under More Dog Links]. Seminal quote:
Dog bite statistics are not really statistics, and they do not give an accurate picture of dogs that bite.
For the record, the AVMA task force included representatives from the American Veterinary Medical Association; the American Academy of Pediatrics; the American College of Emergency Physicians; the Professional Liability Insurance Trust; the American College of Veterinary Behaviorists; the American Medical Association; the National Animal Control Association; the Centers for Disease Control and Prevention; and the Humane Society of the United States. There are actual footnotes, too.

Is there a bite report for the footnote-averse? Given the boundless, staggering ignorance that influences so much public debate on the subject of dog bite prevention, you know there is. For those convinced that pit bulls send more people to the hospital than all other breeds combined – because "you never read about Lab attacks in the paper" -- the Clifton report is the go-to reference.

I’m embarrassed for people who cite it. It’s that bad.




Merritt Clifton’s study is actually a list of severe dog bites. The title itself ["Dog attack deaths and maimings"] is misleading, since the list is a compilation of "dog attacks doing bodily harm," including some that are fatal or disabling. Clifton’s only source is the press: specifically, press accounts of dog bites requiring “extensive hospitalization” [never defined, so this might include anything from treatment of sepsis to multiple surgeries] and caused by “clearly identified” animals. [“[T]his table covers only attacks by dogs of clearly identified breed type or ancestry, as designated by animal control officers or others with evident expertise, who have been kept as pets.”] The numbers aren’t organized by year or location, and readers have no way to access the original press accounts and follow-up articles. There is a disclaimer of sorts --- “dogs whose breed type may be uncertain” are excluded, as are police and security dogs and dogs trained to fight --- leading logical readers to assume that the list must include virtually all severe bites by dogs of identifiable breeds.




Clifton’s report never mentions that there is a huge discrepancy between actual hospital records and press accounts of dog attacks --- between relatively objective data, in other words, and highly subjective reporting and editing with an eye to selling papers. The report fails to acknowledge that a number of factors are involved whenever any dog bites. The report includes statements about dog behavior which have no basis in science, and statements about breed-specific traits which bear no relation to the actual history, behavior or modern development of the breed being discussed [in this case, the German shepherd]. Clifton’s concluding statements regarding the inevitability of attacks by certain dogs are impossible to substantiate, and as a result seem simply prejudiced and inflammatory.




***

Here’s an important CDC number to keep in mind: based on hospital records, each year some 6,000 people in the United States are hospitalized as a result of a dog bite or attack. [From the CDC’s Morbidity and Mortality Weekly Report: “Of an estimated 333,700 patients treated for dog bites in emergency departments (EDs) in 1994, approximately 6,000 were hospitalized.” I imagine that number has increased, but for the purposes of this post I’ll stick with 6,000.] 6,000 hospitalized each year: not simply treated in the ED, but requiring hospitalization due to the severity of the dog bite or attack.

According to Clifton's report [which, once again, is based entirely on press accounts], during the 24-year period covered by his study there were a total of 2,209 “[dog] attacks doing bodily harm” in the U.S. and Canada. 1,182 of those attacks were by pit bulls and pit bull mixes. (Lumping mixes together with so-called purebreds makes no sense from any standpoint, but Mr. Clifton lumps them together --- so I will, too, again for the purposes of this post.)

1,182 severe attacks by pit bulls and pit mixes in the U.S. and Canada over a 24-year period [according to the Clifton statistics] works out to an average of just over 49 severe attacks by pit bulls and pit bull mixes in North America per year.

If Clifton’s pit bull numbers are correct, and no more than 49 of the 6,000 or so hospitalizations due to severe dog bites in the U.S. each year are a result of pit bull bites or attacks, then pit bulls and pit mixes are responsible for less than one percent of those hospitalizations.




.82%. Eighty-two hundredths of a percent of hospitalizations due to dog bites in the U.S. each year are a result of pit bull bites or attacks, if the press has accurately represented the number of serious attacks by pit bulls and pit mixes.

This might be a good place to mention that the pit bull is one of the most popular breeds [or types] in the country. Using shelter numbers as a very rough means of estimating the number of pit bulls [registered and unregistered] in the general population, even low estimates end up in the millions. A board member of the California Animal Control Directors Association [CACDA] told me in 2005 that only labs and lab mixes are more common in California shelters. On sites like this, out of a total U.S. population of over 70 million dogs you’ll find estimates of 3 million to 10 million pit bulls.




Could the press be failing to report severe attacks by pit bulls?

While I struggle to get my face under control, check out the screen cap at the top of this post. “East Lubbock, Texas: Elderly man shaken by pit bull.”

Terrifying headline. But he wasn’t actually hurt, you understand, just… shaken.

What about other breeds? Let’s take a look. The only state that has attempted to track dog bites statewide by breed, using hospital records, is Texas. The Texas Severe Animal Bite Summaries were posted online by the Texas Department of Health Zoonosis Control Group from 1996 through 2002. No reports have been issued since then, for a couple reasons: the recorded information had never been complete [some counties didn’t report] and the Texas Department of Health believed the numbers were being misused and misinterpreted. (How do I know this? I phoned them and asked.)

Look at blue heelers: in one year, “blue heelers” and “heeler mixes” caused injuries that required six people to be hospitalized in Texas, according to the 2000 Severe Bite Summary provided by the Texas Department of Health.




Clifton seems unaware that the blue heeler, the Australian blue heeler, the Queensland heeler and the Australian cattle dog are all names that refer to the same breed. They are listed separately in his tabulation. In any event, according to Clifton’s list of press accounts, all of them were apparently responsible for a total of just six severe bites throughout North America over the 24 years covered by his study. (If I didn’t know the breed better – and I like heelers – I might assume that only the ones in Texas bite.)

Take chows. [Please. I still can’t process that Jean Donaldson went from border collies to Buffy ;~)] Based on the numbers provided by the Texas Department of Health, chows and chow mixes accounted for 54 hospitalizations in the state from 1996 through 2000.

According to press accounts tabulated in the Clifton report, there were 57 severe attacks by chows and chow mixes over a 24-year period throughout the entire U.S. and Canada.

Or 66 severe attacks, if you count the mixes with “chow” listed last, as in “Akita/Chow mix.”




And what determines the predominant breed in a mix for the purposes of this report? Why does the Clifton report use “akita/chow mix” rather than “chow/akita mix”? And what is a “chox mix”? What is a “Dauschund”? What is an “East Highland terrier”? I live across the river from East Highland, and I never knew they had their own terrier. “Great Pyranees,” “Weimaeaner,” “Fila Brasiero,” “Doge de Bordeaux,” “Buff Mastiff”… Stop, you’re killing me.

As proof of media bias, the Clifton report has value. The media have done a bang-up job convincing the public that only "dangerous breeds" hurt people. Editors in a shrinking market know that it's more lucrative to rail against pit bulls than talk about the importance of puppy socialization and parent supervision and how to prevent resource guarding. Clifton’s list illustrates perfectly what the AVMA Task Force on Canine Aggression calls “media-driven portrayals of a specific breed as ‘dangerous.’”

***

In Part II of my look at Merritt Clifton’s dog bite study I’ll review his “Analysis.” Clifton writes that misunderstood German shepherds bite often, but only to pull children from harm’s way; that the GSD has developed “three distinctly different kinds of bite” to control sheep, and uses the same bites on people; that chows are not a common breed; that it is the “custom” to dock pit bulls’ tails; and that “temperament is not the issue, nor is it even relevant,” since the great majority of “dangerous breeds” maim or kill whenever they have “a bad moment.”

Stay tuned.



*************

As an addendum, here is an account of a severe dog bite that, like most, received no press coverage.

I heard about the attack from a student -- a friend of the girl who was injured -- and I sent my phone number to the family with a request that the girl’s mother call me if she wouldn’t mind discussing the incident. The girl’s mother gave me the details that follow.

The girl, who was twelve at the time, was staying at her aunt’s house for the night. The families were close and the children all loved the friendly, thirteen year old family dog. That night the girl went into the guest bedroom and reached down to give an affectionate pat to the dog, which was sleeping near the door to the room. The dog lunged up and bit the girl’s face, and didn’t let go. “The dog shook her like a rag doll,” the girl’s mother told me. That tired phrase doesn't sound nearly so hackneyed when it comes from a parent struggling to describe a violent attack on her child.

The girl’s upper lip and part of her nose were torn away, hanging in shreds.

The girl underwent reconstructive surgery at a local hospital where her aunt worked as a nurse. A few days after the attack, the aunt saw the plastic surgeon in a hallway. She introduced herself as a relative of the injured girl and said to the surgeon, “I bet you never saw anything like that before.” The surgeon laughed out loud. “I stitch up five dog bites a week,” he said.

The girl’s mother told me that her daughter’s reconstructive surgery was successful and that no scars would be visible after a few years, though the girl’s lip remained numb: a special concern, since she had played the flute. The dog, a yellow lab, was euthanized.

Pit bulls, dog bite statistics, and the Merritt Clifton report2


Part II:
Clifton’s “Analysis”


In Part 1 of my look at Merritt Clifton’s study of dog bites [“Dangerous breeds,” dog bite statistics, and the Merritt Clifton report: Numbers] I showed that Clifton’s tabulation of press accounts is incomplete, inaccurate, badly edited and misleading. Readers have no way to access the original news stories and follow-up articles; breeds of dogs aren’t accurately recorded; and there is a huge discrepancy between press accounts of dog attacks and hospital data.

Clifton follows his list of severe bites with a brief section of comments on selected attacks. No footnotes or links are provided: in fact, there are no citations anywhere in the report.

In the Analysis section Clifton writes:
The tallies of attacks, attacks on children, attacks on adults, fatalities, and maimings on the above data sheet must be evaluated in three different contexts. The first pertains to breed-specific characteristic behavior, the second to bite frequency as opposed to the frequency of severe injuries, and the third to degree of relative risk.
In other words, Clifton is turning his back on everything known about the cause and prevention of dog bites, and is choosing instead to evaluate severe bites and attacks on the basis of “breed–specific characteristics”: an approach with no grounds in science, and one that has been discredited and rejected by the CDC, the AVMA, veterinary behaviorists, trainers, humane organizations and ACOs.

Neither Merritt Clifton nor I nor anyone else can speak about any breed’s “bite frequency,” or compare “bite frequency” to “the frequency of severe injuries,” because no one knows how often dogs of any breed bite. It is possible, for instance, that most pit bull “bites” are nips or bruises, and that none of those bites are recorded by the press because the bites cause neither real injury nor concern. No one knows. No one knows how often dogs of a particular breed bite or what percentage of those bites are severe, and no one knows the number of dogs of each breed in the overall population. Bite frequency and relative risk are impossible to determine.

These facts are evident, or should be evident, to anyone. As the AVMA task force on dog aggression states:
Dog bite statistics are not really statistics, and they do not give an accurate picture of dogs that bite. Invariably the numbers will show that dogs from popular large breeds are a problem. This should be expected, because big dogs can physically do more damage if they do bite.
The AVMA task force needed less than a paragraph in its Community approach to dog bite prevention to kick the Clifton study to the curb, but I’m going to spend a bit more time on his Analysis section. Frankly, it’s such a train wreck I can’t look away.

Clifton’s comments are in green --- mine are in red.
Of the breeds most often involved in incidents of sufficient severity to be listed, pit bull terriers are noteworthy for attacking adults almost as frequently as children. Numbers again: Clifton doesn’t know how often pit bulls bite children or how frequently they bite adults. He has no idea how often they bite at all. No one does. This is a very rare pattern […] What pattern? Without knowing anything about a biting dog’s health, care or condition, its quality of ownership, victim behavior and so forth, there are no patterns. Captain Arthur Haggerty, a well-known dog trainer, wrote that in cases where a pit bull bites, the owner almost always has a criminal record. That might be a pattern worth investigating. Pit bulls seem to differ behaviorally from other dogs in having far less inhibition about attacking people who are larger than they are. Which pit bulls? Out of the country’s three million or more pit bulls, which ones “differ behaviorally”? Mr. Clifton, meet Mr. GladwellOnce again, Clifton refuses to acknowledge that other factors may be involved when a pit bull bites: abuse, starvation, life on a chain. They are also notorious for attacking seemingly without warning, “Notorious?” According to whom? Barring health problems and/or abuse, all dogs exhibit a pattern of negative or abnormal behavior before a bite. a tendency exacerbated by the custom of docking pit bulls' tails so that warning signals are not easily recognized. Where on earth is it “the custom” to dock pit bulls’ tails? Seriously. Is there any substantiating evidence for this statement? Because no one docks pit bulls’ tails. Ears, sometimes --- tails, never [unless the tail is injured]. Thus the adult victim of a pit bull attack may have had little or no opportunity to read the warning signals that would avert an attack from any other dog. How many of the pit bulls involved in this list of attacks had docked tails? “We have no idea”? I thought so.

Rottweilers […] seem to show up disproportionately often in the mauling, killing, and maiming statistics simply because they are both quite popular and very powerful[...] A “large, popular breed,” in other words --- as opposed to those notorious pit bulls with their rare behavioral patterns.

German shepherds are herding dogs, bred for generations to guide and protect sheep. Please. German shepherds are show and companion dogs, and no, they have not been bred for generations to work sheep. For the last century they have been bred for conformation; for police work and protection sports; and to be pets. Most modern GSDs have as much “herding instinct” as a bulldog, and based on what I’ve seen at AKC “herding trials,” as much talent on stock. In modern society, they are among the dogs of choice for families with small children because of their extremely strong protective instinct. How does he know? They have three distinctively different kinds of bite [I can’t believe I’m reading this]: the guiding nip, which is gentle and does not break the skin; the grab-and-drag, to pull a puppy or lamb or child away from danger, which is as gentle as emergency circumstances allow; and the reactive bite, usually in defense of territory, a child, or someone else the dog is inclined to guard. The reactive bite usually comes only after many warning barks, growls, and other exhibitions intended to avert a conflict. When it does come, it is typically accompanied by a frontal leap for the wrist or throat. Unbelievable. There is not a shred of evidence for these statements, and I’m speaking 1) as a former owner of both American-bred and Schutzhund-bred German shepherds, and 2) as one with a fair amount of literature in my possession relating to this breed. If German shepherds are bred for any sort of bite, it’s the full-mouthed protection-sport bite, and that same type of bite is admired by the relatively tiny but dedicated group of GSD boundary-style herding enthusiasts here and in Germany. “Three distinctively different kinds of bite”? I thought I’d seen it all after reading a few years’ worth of AKC Gazette breed columns [“Collies must have an oblique eye set in order to scan the horizon for sheep”], and then something like this comes along. The “three different bites” concept has nothing to do with the millions of GSDs in North America and even less to do with the realities of stockwork. Seriously, where did this “three distinctive bites” business come from? Is it from the Gazette? Enlighten us. Because it sounds as if it was made up out of whole cloth.

Because German shepherds often use the guiding nip and the grab-and-drag with children, who sometimes misread the dogs' intentions and pull away in panic, they are involved in biting incidents at almost twice the rate that their numbers alone would predict: approximately 28% of all bite cases, according to a recent five-year compilation of Minneapolis animal control data. Yet none of the Minneapolis bites by German shepherds involved a serious injury: hurting someone is almost never the dogs' intent.

Again: I can’t believe I’m reading this.

GSDs “are involved in biting incidents at almost twice the rate that their numbers alone would predict” --- because they are trying to lead or pull children out of danger? There is nothing to support this claim: zero, zip, nada. And what’s up with the reference to Minneapolis AC data? Is this a “Look! Something shiny!” gambit? Because in Clifton’s own study, GSD attacks result in a higher percentage of maimings [60%] than pit bull attacks do [55%]. And yet -- again with no supporting evidence –Clifton states that when German shepherds bite, “hurting someone is almost never the dogs' intent.” He knows all the GSDs in the country, apparently, and reads their minds, too.
In the German shepherd mauling, killing, and maiming cases I have recorded, there have almost always been circumstances of duress: the dog was deranged from being kept alone on a chain for prolonged periods without human contract, was starving, was otherwise severely abused, was protecting puppies, or was part of a pack including other dangerous dogs. None of the German shepherd attacks have involved predatory behavior on the part of an otherwise healthy dog.
Neither Clifton nor [as far as anyone can tell] the press accounts he tabulated see fit to acknowledge “circumstances of duress” when a pit bull or other breed bites. Apparently pit bulls are never chained, never turned loose or abandoned to run in packs, never abused, always well fed and correctly socialized --- unlike those misunderstood German shepherds.

Seriously, it’s unbelievable – unconscionable – that anyone writing about dog attacks would fail to acknowledge the abusive conditions surrounding most pit bulls that bite, and yet make such a raft of excuses for another breed, claiming, “There have almost always been circumstances of duress.” Is this intentional, or simply drawn from the press accounts themselves? Either way, the lack of objectivity is remarkable.

Clifton goes on to stress “the need for some sort of strong breed-specific regulation to deal with pit bulls and Rottweilers,” and chastises the “humane community” for refusing to see that adopting a pit bull is no different from adopting a mountain lion. He states that pit bulls and Rottweilers should be strictly regulated “if they are to be kept at all.”

And that’s just stupid.

Ban all dogs over 30 lb, and children (and adults) will still be maimed and killed by dachshunds and cocker spaniels. But the statement that will make knowledgeable, informed dog people beat their heads on rocks is this one:
Temperament is not the issue, nor is it even relevant. What is relevant is actuarial risk. If almost any other dog has a bad moment, someone may get bitten, but will not be maimed for life or killed, and the actuarial risk is accordingly reasonable. If a pit bull terrier or a Rottweiler has a bad moment, often someone is maimed or killed--and that has now created off-the-chart actuarial risk, for which the dogs as well as their victims are paying the price.

I think Clifton wants very much to say, “They’ll turn on you.” But a dog with a good temperament won’t do that. If your dog is elderly and confused, suffering from a brain tumor or seizure disorder or terrible thyroid problem, or is injured and in pain, he may bite regardless of a terrific temperament. Dogs aren’t robots. But dogs with good temperaments don’t turn on people. They don’t have chainsaw-massacre “bad moments.”

Dogs with unsound temperaments? Especially if they are unsocialized, untrained and badly managed, everyone around them is always at risk.

The bulk of dogs that bite are somewhere in the middle. One may have a sound temperament but was never socialized to children. Another may be a nice dog whose resource-guarding habit was never addressed. Some have just had all they’re going to take, like the Dalmatian who bit a boy that jumped from his bed onto the sleeping dog. And because children are small, many of these injuries are terrible.

“Bad moment” isn’t a scientific term. It’s a euphemism for something only “those breeds” do. It implies that no one really has to worry about “safe breeds,” and the trouble with this particular implication is that “safe breeds” already send hundreds of thousands of children – and adults -- to hospital emergency rooms each year. The choice of phrase reflects all the misleading, inaccurate figures and statements that make the Clifton report such a hot mess: another sad measure of the general public’s “boundless, staggering ignorance about dogs.”