Wireless telephones and the risk of road crashes

https://doi.org/10.1016/S0001-4575(02)00043-XGet rights and content

Abstract

In light of the rapidly increasing development of the cell phone market, the use of such equipment while driving raises the question of whether it is associated with an increased accident risk; and if so, what is its magnitude. This research is an epidemiological study on two large cohorts, namely users and non-users of cell phones, with the objective of verifying whether an association exists between cell phone use and road crashes, separating those with injuries.

The Société de l’Assurance Automobile du Québec (SAAQ) mailed a questionnaire and letter of consent to 175 000 licence holders for passenger vehicles. The questionnaire asked about exposure to risk, driving habits, opinions about activities likely to be detrimental to driving and accidents within the last 24 months. For cell phone users, questions pertaining to the use of the telephone were added. We received 36 078 completed questionnaires, with a signed letter of consent. Four wireless phone companies provided the files on cell phone activity, and the SAAQ the files for 4 years of drivers’ records and police reports. The three data sources were merged using an anonymized identification number. The statistical methods include logistic-normal regression models to estimate the strength of the links between the explanatory variables and crashes.

The relative risk of all accidents and of accidents with injuries is higher for users of cell phones than for non-users. The relative risks (RR) for injury collisions and also for all collisions is 38% higher for men and women cell phone users. These risks diminish to 1.1 for men and 1.2 for women if other variables, such as the kilometres driven and driving habits are incorporated into the models. Similar results hold for several sub-groups. The most significant finding is a dose-response relationship between the frequency of cell phone use, and crash risks. The adjusted relative risks for heavy users are at least two compared to those making minimal use of cell phones; the latter show similar collision rates as do the non-users.

Introduction

The immense popularity of the wireless phone and the continued worldwide growth of the wireless phone industry (NCI, 2001) have created a new challenge for legislators. Wireless telephones save lives by allowing quicker assistance in case of traffic accident or of dangerous situation, non-road medical emergency, crimes, and being lost in the bush or being in difficulty at sea (Chapman and Schofield, 1998). The Cellular Telecommunications & Internet Association (CTIA) estimates that, in 2001, 115 millions wireless users are making over 140 000 emergency calls a day in the USA (CTIA, 2001). However, several safety concerns have been raised related to their use while driving.

Several studies have looked at the impact of the wireless phones on the driver’s attention and on car crashes. Experimental studies conducted on driving simulators or on specially equipped vehicles have shown that the use of wireless phones while driving affects the performance of the driver, particularly the reaction time, mental load and lateral control of the vehicle (Irwin et al., 2000, Lamble et al., 1999, Alm and Nilsson, 1995, McKnight and McKnight, 1993, Stein et al., 1987). However, Haigney and Westerman (2001) state that better designed studies could give us further insights on how the cell phones distract the driver in ‘real life’ situations.

Few epidemiologic studies done on real crashes have been performed to estimate the link between car crashes and using a wireless telephone.1 A case-crossover study performed in Toronto, found that wireless-phone activity while driving quadrupled the relative risk of a collision with property damage only (Redelmeier and Tibshirani, 1997). What is the real risk? It is difficult to assess it since we believe most drivers would not report to police officers that they were talking on their cell phones when involved in car crashes (James, 1991). Also, if we have access to the reported time of the accident from the wireless account of the users, it is difficult to conclude if someone was talking on the phone during the accident or right after the accident. For that reason, several researchers have raised doubts on the study results by Redelmeier and Tibshirani (1997) (NHTSA, 1997, Maclure and Mittleman, 1997). Moreover Quinlan (1997) stated that the data precision was probably inflating the real risk.

The public often thinks that it is dangerous to talk on the phone and drive (Ayres et al., 1999; Laberge-Nadeau et al., 2001a, Laberge-Nadeau et al., 2001b). For that reason, wireless companies started educating their customers to be careful when driving, and several countries have already restricted or banned the use of wireless technologies while driving. It was estimated that if one car out of 10 has a telephone, then 1.2% of car crashes will be caused by wireless technologies. This would amount to a cost to society of at least 2–4 billion per year in the US (Maclure and Mittleman, 1997). Cost-effectiveness of educational programs or legislation needs good estimates of the cost of laws and their impact on drivers (Redelmeier and Weinstein, 1999). If drivers are not educable (Peters and Peters, 2001), then tough legislation would be needed. The tendencies of the data examined, the increasing complexity of the technology, and the sensitive nature of the political and social considerations reinforce the need to gather data from studies conducted in real driving situations.

This article presents the results of an epidemiological study done on two large cohorts, namely users and non-users of mobile (cellular) phones with the objective of verifying whether an association exists between cell phone use and crash rates. A distinction is made between collisions with injuries and total crashes; the property damage collision make up 75% of the latter. We study whether licensees who are users of cell phones have more or fewer crashes than those who are non-users taking into account other variables (age, kilometres driven annually, etc.) that could predict the number of crashes. We also look at the frequency of cell phone calls for users and determine if the frequency of cell phone use is or is not correlated with collision risks. It is important to note that we do not have precise data on cell phone use while driving. Nevertheless any association between overall cell phone use and crash rates resulting from this study is good circumstantial evidence based on sound epidemiology.

Section snippets

Design

A retrospective epidemiological study was planned to obtain large samples of users and non-users of wireless phones, stratified by gender and age, in order to compare and model collision rates using pertinent explanatory variables. The sample had to be drawn from the Société de l’Assurance Automobile du Québec (SAAQ) among class 5 permit holders (passenger vehicles) since legal constraints did not allow drawing samples directly from the wireless telephone subscribers.

A pilot study by our

Results

From the 175 000 questionnaires that were mailed by the SAAQ, 36 078 were returned with the signed consent form giving access to their driving records and cell phone usage from the cell phone companies, for those with a cell phone. The overall response rate was 20.6 and 19.7% for males (n=22 942) and 22.5% for females (n=13 136). The average number of crashes per annum per 1000 drivers for the period of 1996–1999 were obtained from the SAAQ for the anonymized non-respondents and respondents by

Discussion and conclusion

Globally, results show a higher relative risk of all collisions as well as a higher relative risk of accidents with injuries for cell phone users, compared to non-users. The relative risks for injury collisions and also for all collisions is 38% higher for men and women cell phone users. When we take into account confounding variables, the adjusted relative risk for all collisions is lower, namely 1.11 among male users and 1.21 for female users compared to non-users. The adjusted relative risk

Acknowledgements

Financial assistance for this research was collectively provided by the Fonds pour la formation de Chercheurs et l’Aide à la Recherche (FCAR), the Quebec Ministry of Transport (MTQ) and the Société de l’Assurance Automobile du Québec (SAAQ), within the scope of the university research programme on road safety. Transport Canada and the SAAQ provided additional support for the mailing of the questionnaires. We are most grateful for the collaboration of the Canadian Wireless Telecommunications

References (23)

  • M. Irwin et al.

    Effect of the intensity of wireless telephone conversations on reaction time in a braking response

    Percept. Mot. Skills

    (2000)
  • Cited by (0)

    View full text