Commentary
Lethal misconceptions: interpretation and bias in studies of traffic deaths

https://doi.org/10.1016/j.jclinepi.2011.09.007Get rights and content

Abstract

Clinical epidemiology studies are vulnerable to subtle confounding, leading skeptics to claim that an odds ratio below three rarely indicates a clinically important finding. We argue that such a high threshold is inappropriate when interpreting traffic death studies in clinical epidemiology research. We review 10 concepts that emphasize the value of modest effect sizes by taking into account the baseline frequency, nonfatal disability, numbers needed to treat, shared responsibility, event diversity, behavioral offsets, measurement error, indirect reinforcement, delayed progression, and economic affordability. An awareness of these concepts may help when interpreting effect sizes in studies of traffic deaths.

Introduction

A complacent attitude toward traffic deaths might be reasonable if crashes were difficult to predict, hard to prevent, or costly to avoid. In a series of articles, however, our group has shown that seemingly minor differences in behavior can lead to measurable differences in traffic deaths. For example, U.S. presidential elections lead to a 19% increase in traffic deaths during the hours of polling and the average Super Bowl leads to a 41% increase in traffic deaths during the hours following the broadcast [1], [2]. These and other studies [3], [4], [5], [6], [7], [8], [9], [10], [11] call into question community attitudes that characterize accidents as random acts, divine planning, or unavoidable fates. Instead, scientific studies prove how easily the risks of a traffic death can be increased or decreased throughout society.

Studies of traffic deaths, however, often fail to change clinical practice for at least three reasons. First, the traditional practice of medicine focuses more on relieving suffering in the aftermath of an event rather than primary prevention in otherwise asymptomatic patients [12]. Second, the failures to prevent traffic deaths can stay invisible to clinicians if reports are not relayed back to health care providers [13], [14]. Third, effective methods for preventing traffic deaths rarely take the form of prescribing a medication or a surgical procedure; instead, most interventions involve added inconvenience to the patient and clinician [15]. As a consequence, formal medical training and ongoing clinical practice can accentuate rather than mitigate complacent attitudes toward traffic deaths.

We wondered whether the inbuilt structure of traffic death studies also thwarts progress for more thoughtful attention and better patient outcomes. In accord with scientific standards, effect sizes in traffic death studies are usually quantified with numerical units such as an odds ratio or relative risk reduction [16]. In turn, the relative risk reductions almost never reach 100% and sometimes range to single-digit integers (Table 1). Well-intentioned readers, therefore, may have difficulty distinguishing trivial from substantial levels of effectiveness. The purpose of this article is to explain why small differences in observed risks may indicate important opportunities for physicians to save lives from traffic deaths.

Section snippets

Baseline frequency

Traffic deaths are a remarkably frequent problem currently ranked as the ninth most common cause of death worldwide and projected to rise to the fifth place by the year 2030 [24]. Collectively, this baseline risk amounts to about 3,000 deaths per day on a global basis [25], [26]. This burden extends throughout high- and low-income countries [27]; for example, the United States has about 3–5% of the total global population and about 3–5% of the total global traffic deaths [25], [28], [29]. A

Summary

This article has summarized 10 reasons why small differences in observed risks may sometimes represent important findings in traffic death studies. Much of the material will be familiar to investigators skilled in statistical design; however, the framework may help guide those who are either inexperienced or confronted by an unenthused audience. Each of the 10 concepts can explain why a seemingly minor odds ratio can have a major implication to human health related to traffic risks. However,

Acknowledgments

This project was supported by the Canada Research Chair in Medical Decision Sciences and the Comprehensive Research Experience for Medical Students program of the University of Toronto Faculty of Medicine. The views expressed in this article are those of the authors and do not necessarily reflect the Ontario Ministry of Health. We wish to thank the following for helpful comments: William Chan and Neil Dattani.

References (81)

  • H.D. White et al.

    Acute myocardial infarction

    Lancet

    (2008)
  • C. Laberge-Nadeau et al.

    Occupant injury severity from lateral collisions: a literature review

    J Safety Res

    (2009)
  • F. Sagberg et al.

    An investigation of behavioural adaptation to airbags and antilock brakes among taxi drivers

    Accid Anal Prev

    (1997)
  • W. Verbeke et al.

    Dietary behaviour of pregnant versus non-pregnant women

    Appetite

    (2007)
  • J.B. Edwards

    Speed adjustment of motorway commuter traffic to inclement weather

    Transp Res Part F Traffic Psychol Behav

    (1999)
  • A.F. Karr et al.

    Data quality: A statistical perspective

    Stat Methodol

    (2006)
  • N. Freemantle et al.

    Real-world effectiveness of new medicines should be evaluated by appropriately designed clinical trials

    J Clin Epidemiol

    (2010)
  • R. Foss et al.

    Effectiveness of graduated driver licensing in reducing motor vehicle crashes

    Am J Prev Med

    (1999)
  • E. Zaloshnja et al.

    Crash costs in the United States by crash geometry

    Accid Anal Prev

    (2006)
  • D.E. Clark et al.

    Rural and urban traffic fatalities, vehicle miles, and population density

    Accid Anal Prev

    (2004)
  • C.S. Carpenter et al.

    The effects of mandatory seatbelt laws on seatbelt use, motor vehicle fatalities, and crash-related injuries among youths

    J Health Econ

    (2008)
  • P. Anderson et al.

    Effectiveness and cost-effectiveness of policies and programmes to reduce the harm caused by alcohol

    Lancet

    (2009)
  • D.A. Redelmeier et al.

    Driving fatalities on Super Bowl Sunday

    N Engl J Med

    (2003)
  • D.A. Redelmeier et al.

    Driving Fatalities on US Presidential Election Days

    JAMA

    (2008)
  • D.A. Redelmeier et al.

    Association between cellular-telephone calls and motor vehicle collisions

    N Engl J Med

    (1997)
  • J. Terán-Santos et al.

    The association between sleep apnea and the risk of traffic accidents. Cooperative Group Burgos-Santander

    N Engl J Med

    (1999)
  • L.-H. Chen et al.

    Carrying Passengers as a Risk Factor for Crashes Fatal to 16- and 17-Year-Old Drivers

    JAMA

    (2000)
  • C. Owsley et al.

    Impact of Cataract Surgery on Motor Vehicle Crash Involvement by Older Adults

    JAMA

    (2002)
  • L.S. Friedman et al.

    Long-term effects of repealing the national maximum speed limit in the United States

    Am J Public Health

    (2009)
  • D.A. Redelmeier et al.

    Time lost by driving fast in the United States

    Med Decis Making

    (2010)
  • P.D. Jacobson et al.

    Reducing Distracted Driving: Regulation and Education to Avert Traffic Injuries and Fatalities

    JAMA

    (2010)
  • R.M. Stewart et al.

    Seven hundred fifty-three consecutive deaths in a level I trauma center: the argument for injury prevention

    J Trauma

    (2003)
  • F.E. Vaca et al.

    Crash injury prediction and vehicle damage reporting by paramedics

    West J Emerg Med

    (2009)
  • S. Jeffrey et al.

    An evaluation of police reporting of road casualties

    Inj Prev

    (2009)
  • R.D. Foss et al.

    Initial effects of graduated driver licensing on 16-year-old driver crashes in North Carolina

    JAMA

    (2001)
  • T.R. Miller et al.

    Cost-outcome analysis in injury prevention and control: eighty-four recent estimates for the United States

    Med Care

    (2000)
  • A. Hazen et al.

    Road traffic injuries: hidden epidemic in less developed countries

    J Natl Med Assoc

    (2006)
  • D.A. Redelmeier et al.

    Motor vehicle crashes in diabetic patients with tight glycemic control: a population-based case control analysis

    PLoS Med

    (2009)
  • T. Vaa

    Impairment, diseases, age and their relative risks of accident involvement: results from meta-analysis [Internet]

  • S. Tregear et al.

    Obstructive sleep apnea and risk of motor vehicle crash: systematic review and meta-analysis

    J Clin Sleep Med

    (2009 Dec 15)
  • Cited by (12)

    • COVID Vaccine Hesitancy and Risk of a Traffic Crash

      2023, American Journal of Medicine
      Citation Excerpt :

      A psychometric analysis of motorcycle riders found that personal temperament was the largest predictor of crash involvement.57 The weaknesses of past studies include small sample sizes, fallible self-report, cross-sectional designs, low outcome counts, and narrow generalizability.58,59 We are aware of no past study testing COVID vaccination and traffic risks.

    • Intravitreal anti-vascular endothelial growth factor treatment and the risk of thromboembolism

      2015, American Journal of Ophthalmology
      Citation Excerpt :

      Each patient's 5-year time series data were divided into consecutive 28-day segments grouped into 3 larger intervals (Supplemental Figure 1, available at AJO.com): the Baseline interval (year -4, -3, and -2 before first injection), the Induction interval (year -1 before first injection), and the Subsequent interval (year +1 after first injection).35 The main strength of this approach was in generating consecutive 1-month intervals before and after initiation of treatment for each individual patient.36 The 1-year Induction interval was defined in advance and considered a time when a patient's functioning might have potentially differed from baseline.

    • The authors reply

      2013, New England Journal of Medicine
    • Physicians' warnings for unfit drivers and the risk of trauma from road crashes

      2012, New England Journal of Medicine
      Citation Excerpt :

      Odometer data or self-report surveys would provide interesting information but were not available and can be unreliable in assessing driving risks.33,34 The mechanism of risk reduction could include a combination of altered driving behaviors such as traveling shorter distances, driving with greater care, and being more vigilant in avoiding major hazards.35 The results of this study suggest that medical warnings may help to prevent trauma from road crashes.

    View all citing articles on Scopus

    Conflicts of interest: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of data; and preparation, review, or approval of the manuscript. All authors have no financial or personal relationships or affiliations that could influence the decisions and work on this manuscript. The lead author (D.A.R.) had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the analysis.

    View full text