Much of the research on firearms focuses on injuries and deaths that occur when people misuse firearms, and the role that firearms play in the rising violence of some societies. Injuries caused by firearms are classified as either fatal or non-fatal. Homicide, suicide and accidental death are the three types of fatal injuries caused by firearms. Non-fatal injuries are called assaultive, self-inflicted and accidental.
This chapter presents an overview of the role of firearms in deaths and injuries in Canada and introduces issues that will be further developed in subsequent chapters. It provides national statistics on firearm deaths and breaks down these data for firearm suicides and homicides, and for unintentional, or accidental, firearm deaths. It compares international statistics relating to firearms deaths. This chapter provides an overview of the limited information on firearm injuries in Canada, the difficulties researchers encounter in classifying injuries as intentional or unintentional, and the consequent inability to calculate the case-fatality rate for firearm injuries. The chapter then looks at the types of firearms involved in injuries and deaths, and the possible link between the availability of firearms in a society and levels of firearm injuries and death; it also reviews recent research on the costs of firearm injuries and deaths.
Over the past 25 years, there have been an average of 1,300 firearms deaths per year. Of the 1,125 firearm deaths in 1995, about 80.1 percent or 911 were classified as suicides; there were 145 homicides, representing 12.4 percent; and 49 unintentional deaths, for 4.3 percent of the total (Hung, 1997). These percentages have remained relatively stable over the past decade.
In 1995, there was a lower rate of firearm deaths per 100,000 population than there had been in the previous 25 years. In 1970, the rate per 100,000 population was 5.2. It increased to a peak of 7.2 in 1977, and declined steadily to a rate of 3.8 in 1995 (Hung, 1997).
Most countries have produced data on firearm deaths, allowing researchers to estimate the number of suicides, homicides and unintentional deaths by firearms, and to make international comparisons. Such comparisons must be made cautiously because of the different approaches that countries take in keeping records of public health and crime statistics. The data collected through the United Nations International Study on Firearm Regulation (1998) allow for some preliminary comparisons. Twenty-nine countries submitted statistics on the rate of firearm deaths per 100,000 population. The highest rates were reported by Columbia at 55.8, Brazil at 26.9 and the United States at 14 per 100,000 population. About 21 of the countries reported a rate of less than 5 per 100,000 population, including Canada at 4.1; Australia at 3; New Zealand at 2.9; and Sweden at 2.3. Nine countries reported a rate of one or less, including Japan at 0.07 and the United Kingdom at 0.6 (United Nations, 1998: 108).
Injuries caused by firearms that do not result in death are more difficult to determine. Canada, like most countries, does not keep track of all injuries that people suffer and, therefore, cannot point to how many of these involve firearms. However, there are some data on hospitalizations reported by Statistics Canada as well as a database maintained by the Canadian Hospital Injuries Reporting and Prevention Program. Statistics Canada hospitalization data from 1993 and 1994 show that 25 percent of firearm injuries requiring acute care were self-inflicted, such as in an attempted suicide. Nearly 43 percent of injuries were classified as accidents, 22 percent were caused by others, almost nine percent were left undetermined, and the remaining 1.7 percent resulted from legal intervention (Hung, 1997). That profile is quite different from what is known about the intent of actions causing fatal injuries.
The problem of placing firearm-related incidents into such categories as intentional or unintentional, and self-inflicted or assaultive can be complex. We may also underestimate the implications for research.
Researchers often risk misinterpreting data because of the varying number of cases classified as undetermined. Undetermined cases have declined over the last decade since the system used to classify and report incidents has improved. The proportion of undetermined cases does, however, remain significantly higher when someone is injured than when someone dies from a firearm wound.
The previous literature review noted that statistics on accidental deaths involving a firearm were likely inflated because some of them are later classified as suicides or homicides (Kleck, 1991). The argument was dismissed on the basis that the reverse could also be true (Gabor, 1994: 53). Researchers have noted that, while suicides and homicides may be misclassified as fatal accidents, the proportion of misclassified cases is likely to be small because undetermined firearm deaths tend to prompt investigations about possible homicides that are more thorough than general suicide or accidental death investigations (Dudley et al., 1996: 372). Incidents that are wrongly classified may be more misleading when the national figures are relatively small, as they are for Canada.
The concept of case-fatality rate refers to the proportion of cases that result in death among all firearm injury cases (Barber et al., 1996: 487). It is most often expressed as the ratio of non-fatal injuries per death. Given the limited data on non-fatal injuries in Canada, we do not know our country’s case-fatality rate.
Databases on non-fatal injuries for the United States are available, and several national or regional estimates of case-fatality rates have been produced in that country (e.g., Annest et al., 1995; Barber et al., 1996; Bretsky et al., 1996; Kellermann et al., 1996; Mercy, 1993). The estimated case-fatality ratio varied considerably from one study to another. Based on the National Electronic Injury Surveillance System, for example, the national case-fatality rate was estimated at 2.6:1; that is, 2.6 injuries for every one death (Annest et al., 1995: 1751-1752).
Most studies indicate that case-fatality rates varied markedly according to the intent of the shooter. A study of three U.S. cities showed that the case-fatality rate for unintentional injuries was 16:1; the ratio for assaults was 5.3:1; and the ratio for suicide attempts was 0.16:1 (Kellermann et al., 1996: 1443). Unintentional injuries seemed to be less serious and to lead to fatal consequences less often: they were associated with the lowest risk of serious harm. A recent Canadian study found that of those who required emergency care for firearm injuries, 47 percent sustained injuries through an accident, 32 percent had attempted suicide, and 19 percent had been assaulted (Injury Prevention Centre Edmonton, 1996). These percentages change considerably when looking at firearm deaths: 80 percent were suicides, 15 percent were homicides, and five percent were accidents. Quite likely, we are seeing this different picture emerge because, compared to self-inflicted injuries and assaults, unintentional injuries are less likely to involve vital organs.
We observed substantial regional variations in case-fatality rates, relating to the type of firearm, the type of incident and the relative availability of emergency and hospital medical care. Isolated communities and communities that lack sophisticated emergency medical services systems may be less successful at saving the lives of patients who are critically injured (Kellermann et al., 1996).
There is no national data available in Canada on the types of firearms involved in injuries, but research suggests that more people are injured by long guns than by handguns. For example, the 1993-94 Alberta study found that the majority of firearm-related visits to emergency rooms and acute hospitalizations resulted from injuries involving long guns (Injury Prevention Centre Edmonton, 1996). Conversely, in the United States, handguns are more often involved in firearm injuries than are other types of firearms (Sadowski and Muñoz, 1996: 1763; Vassar and Kizer, 1996).
Much of the research on fatal firearm injuries is concerned with the possible link between the number of available firearms and the rate of violent crime, suicide and accidental deaths in a given population. Such research is based on opportunity theory (Mayhew, 1996) and more specifically on what may be termed the general firearm availability theory, which assumes that the more firearms that are available in a society, the more injuries will occur (e.g., Leonard, 1994: 128).
Epidemiological studies of firearm availability and firearm injuries meet with methodological and conceptual problems that are difficult to resolve. For example, there is no way to measure precisely how many people own firearms (Stenning, 1996; 1996b: 10), nor is there currently a way to take into account the fact that the number of people who own firearms varies over time and between jurisdictions. It is quite likely that this issue will never be resolved because such precision is not possible. Furthermore, research has so far been unable to adequately specify, theoretically or empirically, the nature of the link between firearms and violence. Although few people would question that there must be a link between the relative availability of lethal means of violence and the actual level of violence, the exact nature of that link is not obvious. Theoretically, at least, the presence of violence can be conceptualized as either the cause or the result of the increased prevalence of firearms in some societies. In the international context, Lock concluded that widespread availability of firearms does not automatically translate into violent conflict (1996:2).
Where researchers do consider that firearm injuries may be related to the prevalence of firearms, or to specific types of firearms such as handguns, they should consider each type of incident separately (Stenning, 1996: 18). Intentional injuries may show a different pattern than unintentional ones, just as access to firearms may affect the rate of assaultive injury compared to ones that are self-inflicted. These issues will be explored further in some of the following chapters.