Background and Assumptions

Seriously injurious dog bites are very uncommon, and dog bite-related fatalities (DBRFs) extremely rare, but they tend to invoke a strong emotional response and because research indicates that they are largely preventable and may disproportionately affect children (Raghavan, 2008; Sacks, Sattin, & Bonzo, 1989; Sacks, Sinclair, Gilchrist, Golab, & Lockwood, 2000; Shields, Bernstein, Hunsaker, & Stewart, 2009), there has been a push to better understand the circumstances and variables that contribute to such incidents. The first factor that investigators and journalists seek to identify is very often the dog’s breed, with the intention of identifying causal factors in order to reduce future harm. One problem with this approach is that it is extremely difficult to accurately identify a dog’s breed without extensive knowledge of its lineage or a DNA test. Unfortunately, this does not stop people from assigning a breed to dogs based on their visual appearance. Visual breed identifications are insufficient and unreliable, with experts often disagreeing. Therefore, studies that make claims about specific breeds based on visual breed identifications should be scrutinized. In fact, since studies have discredited visual breed identification (Hoffman, Harrison, Wolff, & Westgarth, 2014; Simpson, Simpson, & VanKavage, 2012; Voith, Ingram, Mitsouras, & Irizarry, 2009; Voith et al., 2013), any studies that rely on visual breed identification, including those that link DBRFs and breed, can no longer be responsibly cited in the developing literature.

A quote from Sacks, Sattin, and Bonzo (1989) concisely summarizes the fundamental weakness of visual breed identification, “Any short-haired stocky dog is likely to be called a pit bull.” This should be considered when reading research articles that make claims about “pit bulls” wherein the source of breed identification is either unknown or retrieved from visual breed identification or media reports. Despite this early awareness of potentially unreliable data, decades of research still relied upon and made claims based on media-sourced visual breed identifications. The conclusions cannot be validated and should not be perpetuated. Moreover, the most comprehensive research suggests that dog demographics are not relevant to DBRFs, but rather preventable human decisions regarding care, husbandry, and control of their canine charges may be the most important variables.

Typical Methodology

Over the last several decades, researchers across the globe have attempted to gather and disseminate pertinent information on DBRFs. Though each study differs, the most common approach has been to retroactively study canine bite incidents using media reports as the primary source for data collection. Common factors that have been examined include dog breed, sex of dog and victims, number of dogs involved, relationship between dog and victim, location of attack, the dog’s neuter status, any noted previous aggression, and characteristics of the wounds. Early reports (1965 to the early 2000s) are typically case studies that serve to describe specific attacks and garner scientific interest in dog bite fatalities. As the literature progresses (early 1990s to present), the datasets typically become larger and the researchers try to put the attacks into context and determine correlating factors. Despite a progression in the field, some researchers still rely on outdated methods for gathering data and grouping subjects. This obscures the bigger picture and contributes to a misunderstanding of the true causes of severe dog bites and fatal incidents. One objective of the current review is that going forward each study’s results will be considered in the context of its strong or weak methodology when readers cite the literature, plan future research, report facts to the media, or design public policy. Fortunately, research now exists that does not suffer from the weaknesses of media-sourced data. The use of media reports for data collection is perpetuated because it is well documented in the literature, relatively accessible, and cost effective. However, this standard sacrifices reliability and completeness. It is both unproductive and unethical to knowingly cite and use flawed data. Here we have summarized and analyzed all known studies relating to the DBRF literature, including the scope, data collection methods, and interpretations. This collation may benefit academics, researchers, journalists, and policy makers.


In sum, there is no reliable evidence that demonstrates a link between breed and fatal dog bites. Children and the elderly seem to be more at risk than able-bodied adults, but this can be alleviated by an adult’s presence and supervision. There does appear to be a higher rate of fatal dog bites from sexually unaltered male dogs, though this falls under the bigger umbrella of human-controlled factors; that is to say, owners can choose to spay and neuter their dogs. In fact, many of the factors found to correlate with DBRFs are preventable and in the owners’ control. Improper or insufficient husbandry, lack of positive human interaction, abuse, neglect, failure to spay or neuter dogs, and not supervising individuals who cannot appropriately interact with dogs are all easily preventable factors that have been shown to correlate with DBRFs. Educating the public on responsible dog ownership may be an effective method for reducing dog bite-related fatalities.

The following papers are the most commonly cited, beginning with the most comprehensive study done to date

Most Comprehensive to Date

Patronek, G. J., Sacks, J. J., Delise, K. M., Cleary, D. V., & Marder, A. R. (2013). Co-occurrence of potentially preventable factors in 256 dog bite-related fatalities in the United States (2000-2009). Journal Of The American Veterinary Medical Association243(12), 1726-1736. doi:10.2460/javma.243.12.1726

Early Epidemiological Studies

Sacks, J. J., Lockwood, R., Hornreich, J., & Sattin, R. W. (1996). Fatal dog attacks, 1989-1994. Pediatrics97(6 Pt 1), 891-895.

Sacks, J. J., Sinclair, L., Gilchrist, J., Golab, G. C., & Lockwood, R. (2000). Breeds of dogs involved in fatal human attacks in the United States between 1979 and 1998. Journal Of The American Veterinary Medical Association217(6), 836-840.

Sacks, J. J., Sattin, R. W., & Bonzo, S. E. (1989). Dog bite-related fatalities from 1979 through 1988. JAMA: Journal Of The American Medical Association262(11), 1489-1492.

Pinckney, L. E., & Kennedy, L. A. (1982). Traumatic deaths from dog attacks in the United States. Pediatrics69(2), 193.

Raghavan, M. (2008). Fatal dog attacks in Canada, 1990-2007. The Canadian Veterinary Journal. La Revue Vétérinaire Canadienne49(6), 577-581.

Small, Local, Individual Case Studies

De Munnynck, K., & Van de Voorde, W. (2002). Forensic approach of fatal dog attacks: a case report and literature review. International Journal Of Legal Medicine116(5), 295-300.

Winkler, W. G. (1977). Human deaths induced by dog bites, United States, 1974-75. Public Health Reports (Washington, D.C.: 1974)92(5), 425-429.

Borchelt, P. L., Lockwood, R., Beck, A. M., & Voith, V. L. (1983). Attacks by packs of dogs involving predation on human beings. Public Health Reports (Washington, D.C.: 1974)98(1), 57-66.

Clark, M. A., Sandusky, G. E., Hawley, D. A., Pless, J. E., Fardal, P. M., & Tate, L. R. (1991). Fatal and near-fatal animal bite injuries. Journal Of Forensic Sciences36(4), 1256-1261.

Shields, L. E., Bernstein, M. L., Hunsaker, J. 3., & Stewart, D. M. (2009). Dog bite-related fatalities: a 15-year review of Kentucky medical examiner cases. The American Journal Of Forensic Medicine And Pathology30(3), 223-230. doi:10.1097/PAF.0b013e3181a5e558

Tong, G. T., & Pang, T. C. (1965). Unusual injuries: savaged to death by dogs. Medicine, Science, And The Law5(3), 158-160.

Loewe, C. L., Diaz, F. J., & Bechinski, J. (2007). Pitbull mauling deaths in Detroit. The American Journal Of Forensic Medicine And Pathology28(4), 356-360.