Investigation of unreported multiple causes of death finds totals & “proximate cause” numbers very close to ours
RALEIGH, North Carolina––Marilyn Goss Haskell, DVM, and Ricky Lee Langley, M.D., both also holders of master’s degrees in public health, may never have heard of ANIMALS 24-7 and our log of fatal and disfiguring dog attacks in the U.S. and Canada, maintained since 1982.
But their study “Animal-Encounter Fatalities, United States, 1999-2016: Cause of Death and Misreporting,” recently published by Public Health Reports, the peer-reviewed journal of the Association of Schools and Programs of Public Health, provides an emphatic affirmation of the accuracy of the ANIMALS 24-7 approach to logging and classifying dog attack deaths.
(See Dog attack deaths & maimings, U.S. & Canada, 1982-2020 log. Updated editions are posted annually.)
How ANIMALS 24-7 logs deaths caused by dogs
The ANIMALS 24-7 policy, from the beginning of our logging, has been to log all deaths in which dogs are a proximate cause, with footnotes explaining unusual circumstances, such as deaths by heart attack, head injury, infection, or other accident apparently occurring as result of the victim having been under attack by a dog at the time.
Examples of the latter in the ANIMALS 24-7 log include 17 fatal heart attacks, 11 deaths from antibiotic-resistant infections transmitted by bites, three fatal strokes, three attacks by rabid dogs, three cases in which dogs strangled people by tugging on leashes or scarves, two fatal head injuries suffered when the victims were knocked down by pit bulls, and another suffered when the victim was knocked down by a Rottweiler, several cases in which the victim was hit by a car while dodging a dog attack, two cases in which a pit bull dragged a victim in front of a vehicle, and three deaths occurring when a weapon discharged to try to stop a potentially fatal dog attack in progress accidentally killed a victim who was already being severely mauled.
Taking unusual cases into account
From first publication of our findings, more than 30 years ago, pit bull advocates in particular have attacked our data for including such “proximate cause” deaths, even with the asterisks that enable the data users to quickly distinguish deaths caused directly by a dog’s jaws from other deaths caused by something a dog did.
The ANIMALS 24-7 response has always been simply that the victims are dead, and would not have been dead without the actions of the dog.
Understanding the total ecology of fatal dog attacks requires taking the unusual cases into account, as well as attacks following the most familiar patterns.
How ANIMALS 24-7 gets the data
The ANIMALS 24-7 data source, collated media coverage, typically incorporates information from police reports, animal control reports, witness accounts, victim accounts in most instances in which one or more victims of a dog attack survive, and hospital reports.
Media coverage is, in short, multi-sourced, unlike reports from any single source.
Critics of the ANIMALS 24-7 log of fatal and disfiguring dog attacks, again mostly pit bull advocates, have long argued that the use of media accounts as a data source is suspect because it does not come directly and exclusively from official public health records––which are often not immediately available, not breed-specific, not multi-sourced, and in the versions released in response to public records requests, not specific in identifying victims other than fatalities because of the privacy requirements of the Health Insurance Portability & Accountability Act of 1996.
Haskell & Langley worked from public health records
Marilyn Goss Haskell, DVM, and Ricky Lee Langley, M.D., approached essentially the same set of problems addressed by ANIMALS 24-7 from an altogether different direction, working entirely from official public health records.
Explained Haskell and Langley in their abstract, “Errors and misreporting on death certificates are common, along with potential inaccuracies in cause-of-death coding,” used by public health agencies to categorize deaths of all sorts.
“We characterized and compared fatalities by animal-encounter mentions reported as underlying cause of death,” Haskell and Langley said, “with animal-encounter mentions reported as multiple cause of death to determine factors associated with misreporting underlying cause of death.
“We analyzed fatality data from 1999-2016 from the Centers for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research [WONDER],” Haskell and Langley elaborated.
Haskell and Langley then compared the WONDER data with “Multiple Cause of Death animal-encounter mentions” taken from International Classification of Diseases codes.
Deaths “may be under-reported”
Suggested Haskell and Langley, “Animal-encounter fatalities, analyzed by underlying cause of death alone, may be under-reported. An initiating animal injury, complicated by co-morbidities and fatality, may obscure the causal chain, resulting in misreporting underlying cause of death.”
Haskell and Langley looked at the entire spectrum of animal-encounter fatalities, not just fatal dog attacks, but dog attacks are by far the most common cause of deaths inflicted by mammals.
“The Centers for Disease Control and Prevention and the World Health Organization estimate that each year in the United States, dogs bite approximately 4.5 to 4.7 million people,” Haskell and Langley continued.
“Of these, nearly 885,000 [people] seek medical care, resulting in an average of 323,000 emergency department visits, 30,000 reconstructive procedures, and 10-20 fatalities,” Haskell and Langley wrote; but this fatality count, they discovered, is far low.
Multiple causes of death ignored
“One weakness of previously published studies of animal-encounter deaths in the United States,” Haskell and Langley observed, “is that only the underlying cause of death was reported and analyzed,” without attention to multiple causes contributing to deaths.
Thus, Haskell and Langley explained, “if a death certificate reported sepsis as the underlying cause of death when the injury that initiated the events resulting in death was an animal bite, the animal bite would not have been reported accurately with underlying cause of death analysis alone.”
This proved to be no mere hypothetical speculation.
“Of 77 animal-encounter mentions of dogs” in multiple cause of death reports, Haskell and Langley found, “18 (23%) decedents died of potential zoonotic infection, including 12 [deaths from] septicemia, three from pasteurellosis, and one from rabies.”
These numbers are all close to those discovered by ANIMALS 24-7 from combing news reports.
Over the years for which Haskell and Langley reviewed the data, 1999-2016, ANIMALS 24-7 logged 546 total U.S. deaths caused by dogs, 328 of them (60%) caused by pit bulls.
Haskell and Langley found 553 total U.S. deaths caused by dogs during that time frame in underlying cause of death reports, plus 42 more when multiple causes of death were taken into account, for a combined total of 595: 8% more than ANIMALS 24-7, which is to say that ANIMALS 24-7 found 92%, without having access to official cause of death reports not available until several years later.
ANIMALS 24-7 found that proximate causes accounted for 6% of U.S. dog-caused deaths from 1999 through 2016.
Haskell and Langley found that proximate cause deaths accounted for 7%.
“Under-reporting of risk & fatalities”
Wrote Haskell and Langley, “Our findings suggest that analyzing fatalities by underlying cause alone in CDC WONDER may result in under-reporting of risk and animal- encounter fatalities. We speculate that an initiating animal injury complicated by coexisting conditions (eg, heart disease, cancer, or infection) may trigger a cascade of morbid events leading to hospitalization and fatality, over-shadowing the true underlying cause of death and causal chain of events.”
Most of the fatalities that eluded ANIMALS 24-7 but were discovered by Haskell and Langley involved infections leading to deaths occurring days or weeks after the bite wounds that produced the infections.
“Of all animal bites, 3% to 18% of dog bites and 28% to 80% of cat bites become infected,” Haskell and Langley mentioned, footnoting studies that reported widely varying results.
Deaths from infection most likely to be overlooked
Overall, looking at all animal-related deaths, Haskell and Langley concluded, “Our study found that among 642 additional multiple cause of death animal-encounter mentions” in CDC WONDER data, “infections ranked second for underlying cause of death,” with the data pointing toward pasteurellosis, rabies, streptobacillus, and tularemia.
Thus, Haskell and Langley came to believe, deaths categorized as resulting from a zoonotic infectious disease are especially likely to be misreported in a manner that overlooks an animal bite as a multiple cause of death.
Cautioned Haskell and Langley in conclusion, “Our analysis reemphasizes the risk of fatal dog encounters among children aged less than four years old.
“Fatal dog encounters among children aged four or less are particularly concerning,” Haskell and Langley wrote, noting also the increased propensity of adults over 65 years old to have fatal encounters with reptiles, dogs, and other mammals.
“These findings support an exploration of targeted age group–based preventive education using a One Health approach,” Haskell and Langley suggested.