Factors associated with the all-terrain vehicle mortality rate in the United States: An analysis of state-level data

https://doi.org/10.1016/j.aap.2007.09.021Get rights and content

Abstract

The purpose of this article is to determine and quantify the factors associated with the all-terrain vehicle (ATV) mortality rate in the United States, based on an analysis of state-level data. From 1990 through 1999, there were about 2400 reported deaths in the U.S. involving ATVs. The population-based mortality rate during this time period amounted to about 0.84 deaths per million persons per year, but varied considerably among the 50 states—ranging from a low of 0.09 deaths per million person-years in Rhode Island to a high of 6.33 deaths per million person-years in West Virginia. This variation provides an opportunity for quantifying some of the key factors associated with the ATV mortality risk. The analysis was conducted in a two-stage regression process, with the state mortality rates estimated in the second stage with a negative binomial regression model. The results indicate that the ATV mortality rate was systematically related to ATV usage rates (a measure of risk exposure) and a number of the characteristics of the potential rider pool. These included: the proportion of the state populations that was male, young, lived in rural areas, was college educated, and non-Hispanic white.

Introduction

All-terrain vehicles (ATVs) are three- and four-wheeled motorized vehicles intended for use on various types of unpaved terrain. They have low-pressure tires, seats designed to be straddled, handlebars for steering, and motorcycle-type engines. Engine sizes range from 50 to 700 cm3 of displacement, and vehicle weights range from about 100 to 700 pounds.

Since the 1980s, the U.S. Consumer Product Safety Commission (CPSC) has compiled information on more than 6000 ATV-related deaths (Ingle, 2005). Commission staff has used this fatality database, in combination with national estimates of the U.S. rider population and the number of ATVs in use, to estimate the national mortality rate and identify broad categories of risk (Ingle, 2003, Ingle, 2005, David, 1998). However, compared to studies of nonfatal injuries, the fatality database has been of limited practical value in quantifying risk patterns. This is because less is generally known about the victims of fatal accidents, their ATV usage patterns, and the accident scenarios. Fatal accidents often occur without witnesses and, since the victims cannot be interviewed, it is often difficult to gain a clear understanding of their riding patterns and how the accidents occurred. The sometimes lengthy reporting lag and the difficult process of gaining information about the victim and accident from police and coroner's reports also make it difficult to use the death data to evaluate risk patterns.

Despite these limitations, combining data on reported ATV deaths over time allows us to estimate population-based mortality rates for the states while avoiding some of the small sample problems associated with evaluating mortality rates for a shorter period of time. Moreover, since the population-based mortality rates vary considerably at the state level, they provide an opportunity for quantifying some of the key factors associated with the mortality risk.

A recent descriptive paper highlighted variations in ATV mortality rates among the states (Helmkamp, 2001). The reported mortality rate in California, for example, the state where ATVs were first introduced in the early 1970s, was about 0.4 deaths per million person-years during the 1990–1999 time period. This compared with about 2.4 deaths per million in Idaho, another Western state, and over 5 deaths per million in Alaska.

These mortality rate disparities raise a number of questions. Why, for example, are the rates so much higher in Idaho and Alaska than they are in California? Do variations in demographic and geographic characteristics explain the mortality rate differences, or are the differences related primarily to risk exposure (i.e., the amount that ATVs are used)? If the disparities are related to risk exposure, what accounts for the differences in ATV use in the various states? Additionally, how do the factors associated with the ATV mortality rate compare to those associated with the nonfatal injury risk, which has been studied more extensively? Finally, can these variations be successfully modeled in a way that provides insight into the factors associated with the mortality risk?

The purpose of this paper is to explore these issues. The data used in the analysis are described in the next section, followed by a discussion of the empirical model and the results of the analysis.

Section snippets

Data

The analysis is based on reported ATV deaths, by state, compiled by the CPSC for the years 1990 through 1999 (Ingle, 2005). The District of Columbia was excluded from the analysis because, as an urban environment, it has few (if any) areas for legal recreational riding. State mortality rates were defined as deaths per million person-years, and were constructed as the number of deaths in each state from 1990 through 1999 divided by the sum of annual state population estimates (in millions)

Descriptive results

From 1990 through 1999, there were a total of 2401 ATV-related deaths compiled by the CPSC in the 50 states. About 60% of the reported deaths were based on death certificates purchased from the states; 40% were from other sources.

Some of the characteristics of the fatality database are shown in Table 1. About 87% of the victims were males and about 58% were under age 25 (26% were under age 16). The number of reported deaths declined from 234 in 1990 to 183 in 1993, but rose thereafter. The

Discussion

There are a number of limitations to this analysis. Due to a lack of available statewide data, it did not adjust for possible between-state variations in the characteristics of ATVs in use, driver use patterns (e.g., recreational versus non-recreational use), driver experience, or the use of helmets, factors that have been shown to affect ATV risks in cross-section studies of ATV drivers using case-control study designs (Rodgers, 1990, Rubinfeld and Rodgers, 1992, Rodgers and Adler, 2001).

Conclusions

The results of this analysis suggest that variations in the population-based estimates of the state ATV mortality rates can be successfully modeled, and that the mortality rate variations are in large part related to observable factors such as the ATV usage rate and various demographic characteristics, including the proportion of the state population that was male, young, lived in rural areas, was college educated, and was non-Hispanic white. Where comparable, the results of the present study

References (29)

  • J.C. Helmkamp

    A comparison of state-specific all-terrain vehicle-related death rates, 1990–1999

    Am. J. Public Health

    (2001)
  • R. Ingle

    Analysis of ATV-Related Fatality Data for CPSC Petition CP 02-4/HP 02-1

    (2003)
  • Ingle, R., 2005. 2004 annual report of ATV deaths and injuries. U.S. Consumer Product Safety Commission, Bethesda, MD....
  • P. Kennedy

    A Guide to Econometrics

    (1998)
  • Cited by (37)

    • Understanding risk factor patterns in ATV fatalities: A recursive partitioning approach

      2016, Journal of Safety Research
      Citation Excerpt :

      While our findings support previous research on risk factors for ATV injuries and fatalities, what makes this study unique is the potential for practical applications based on associations between risk factors. Demographic risk factors that have been identified in previous studies include being male (Breslau et al., 2009; Goldcamp et al., 2006; Helmkamp, 2012; Helmkamp et al., 2009; O'Connor et al., 2009; Rechnitzer et al., 2013; Rodgers, 2008; Rodgers & Adler, 2001), young or old ages (Balthrop et al., 2009; Helmkamp, 2012; Helmkamp & Carter, 2009; Helmkamp et al., 2011; O'Connor et al., 2009; Rodgers, 2008; Rodgers & Adler, 2001), inexperience (Goldcamp et al., 2006; O'Connor et al., 2009; Rodgers & Adler, 2001), and use of unsafe practices including lack of personal protective equipment (PPE) and using drugs or alcohol (Fleming, 2010). The high ratio of male to female deaths is not unique to the data contained in this report.

    • Larger ATV engine size correlates with an increased rate of traumatic brain injury

      2015, Injury
      Citation Excerpt :

      Two possible solutions to decreased awareness may include increased education or increased legislation. While one study notes that increased levels of general education likely correlates with increased use of helmets [21], several other studies have concluded that ATV-specific education and the risks of operating an ATV would make a difference [7,20–23]. Meanwhile, legislation has been effective when comparing states with and without helmet laws in decreasing rates of injuries and death, as well as severity of injuries [9,24,25]; however there is a host of research that points to a relative ineffectiveness and inconsistent enforcement of helmet laws [3,26,27].

    • Off-highway vehicle parks: Combining environment, knowledge, and enforcement for all-terrain vehicle injury prevention

      2013, Accident Analysis and Prevention
      Citation Excerpt :

      However, the vast majority of ATV use and the greatest cause of injuries is recreational riding (Fisher et al., 2009; Hendricks et al., 2005; Kute et al., 2007; Rodgers, 1993). A number of risk factors for ATV crashes and injuries have been identified, among them being male, under 16 years of age, inappropriate vehicle size for age and maturity of operator, lack of helmet use, multiple riders on single-person machines, and riding on the road (Aitken et al., 2004; Gittelman et al., 2006; Hall et al., 2009; Rodgers, 2008; Shulruf and Balemi, 2010). Laws for ATV use vary from state to state, but rarely address all known risk factors.

    View all citing articles on Scopus

    The opinions expressed in this article are those of the author and do not necessarily represent the views of the U.S. Consumer Product Safety Commission or other members of its staff. Because this article was written in the author's official capacity, it is in the public domain and may be freely copied.

    View full text