|
|
| |||||||||||||||||||||||||||||||||||||||||
Background Because of a belief that the use of cellular telephones while driving may cause collisions, several countries have restricted their use in motor vehicles, and others are considering such regulations. We used an epidemiologic method, the case–crossover design, to study whether using a cellular telephone while driving increases the risk of a motor vehicle collision.
Methods We studied 699 drivers who had cellular telephones and who were involved in motor vehicle collisions resulting in substantial property damage but no personal injury. Each person's cellular-telephone calls on the day of the collision and during the previous week were analyzed through the use of detailed billing records.
Results A total of 26,798 cellular-telephone calls were made during the 14-month study period. The risk of a collision when using a cellular telephone was four times higher than the risk when a cellular telephone was not being used (relative risk, 4.3; 95 percent confidence interval, 3.0 to 6.5). The relative risk was similar for drivers who differed in personal characteristics such as age and driving experience; calls close to the time of the collision were particularly hazardous (relative risk, 4.8 for calls placed within 5 minutes of the collision, as compared with 1.3 for calls placed more than 15 minutes before the collision; P<0.001); and units that allowed the hands to be free (relative risk, 5.9) offered no safety advantage over hand-held units (relative risk, 3.9; P not significant). Thirty-nine percent of the drivers called emergency services after the collision, suggesting that having a cellular telephone may have had advantages in the aftermath of an event.
Conclusions The use of cellular telephones in motor vehicles is associated with a quadrupling of the risk of a collision during the brief period of a call. Decisions about regulation of such telephones, however, need to take into account the benefits of the technology and the role of individual responsibility.
Cellular telephones can be used for placing and receiving telephone calls while in a motor vehicle. North American sales are enormous; for example, in 1995 the number of new subscribers in the United States exceeded the birth rate.7,8 Many believe that telephones may contribute to collisions by distracting drivers,9 and a few countries (such as Brazil, Israel, and Australia) have laws against using a cellular telephone while driving. Research with simulators suggests that use of the telephone can impair some aspects of driving performance.10,11,12,13,14 However, industry-sponsored surveys have found no increased risk associated with car telephones.15,16
The most rigorous experimental method for testing the effects of cellular telephones on motor vehicle collisions is to assess outcomes for persons randomly assigned to use or not use the devices, but such a study would be very difficult to perform and possibly unethical. Instead, we used an epidemiologic method, the case–crossover design, to evaluate potential associations between the use of a cellular telephone and the risk of a motor vehicle collision in real-world circumstances.
Methods
The study was conducted in Toronto, an urban region of 3 million people with no regulations against using a cellular telephone while driving. Persons who came to the North York Collision Reporting Centre between July 1, 1994, and August 31, 1995, during peak hours (10 a.m. to 6 p.m.) on Monday through Friday were included in the study if they had been in a collision with substantial property damage (as judged by the police). Drivers do not report to the center if the collisions involve injury, criminal activity, or the transport of dangerous goods. Drivers were excluded if they said they did not have a cellular telephone or if their billing records could not be located by May 1, 1996.
Use of Cellular Telephones
Consenting subjects completed a brief questionnaire about their personal characteristics and the features of the collision. We collected telephone records through each person's cellular-telephone number and verified each invoice by checking the subject's full name, mailing address, and calls made to his or her home telephone number. For each record, we analyzed all telephone activity on both the day of collision and the preceding seven days, with particular attention to the time, duration, and direction (incoming or outgoing) of each call. Special note was made of contact with ambulance personnel, police, or other emergency services.
Time of the Motor Vehicle Collision
The time of each collision was estimated from the subject's statement, police records, and telephone listings of calls to emergency services. We classified the times of collisions as "exact" when information from all three sources was available and consistent or when one source supplied no data but the remaining two agreed. Otherwise, we classified the times as "inexact" and used the earliest of the available two or three times to avoid misclassifying calls made after the collision as contributing to the event. Selecting the earliest listed time reduced the chance of finding spurious associations between telephone use and collisions. However, selecting an excessively early time could lead to the underestimation of the magnitude of any association.
Analytic Method
We used case–crossover analysis, a technique for assessing the brief change in risk associated with a transient exposure. According to this method, each person serves as his or her own control; confounding due to age, sex, visual acuity, training, personality, driving record, and other fixed characteristics is thereby eliminated.17 We used the pair-matched analytic approach to contrast a time period on the day of the collision with a comparable period on a day preceding the collision.18 In this instance, case–crossover analysis would identify an increase in risk if there were more telephone calls immediately before the collision than would be expected solely as a result of chance.
Definitions of Time Periods
We defined the hazard interval to include any telephone calls occurring during the 10 minutes before the estimated time of the collision, and tested the robustness of our results by analyzing intervals of 1, 5, and 15 minutes.19 In the primary analysis, we compared each person's telephone activity immediately before the collision (case) to his or her activity during a control period at the same time as the hazard interval on the day before the collision (crossover). In supplementary analyses we evaluated alternative comparison days and considered intervals of an hour leading up to the collision.
Alternative Comparison Days
We checked our estimates by repeating the calculations using four other control intervals. In the workday comparison we selected the day of the workweek preceding the collision; for example, the period just before a collision on Monday was compared with the same period on the preceding Friday. In the weekday comparison, we selected the same day one week before the collision; for example, Monday was compared with the preceding Monday. In the matching-day comparison, we selected the nearest day of the preceding week on which there was cellular-telephone activity in the predefined lead-up period before the collision. For the maximal-use-day comparison, we used the control interval from the preceding three days in which there was the greatest amount of cellular-telephone activity.
Accounting for Intermittency of Driving
Evaluating telephone activity on the day before a collision is appropriate only if driving occurred during the control interval on that day. A pilot survey involving 100 subjects indicated that 35 percent of them did not drive during the selected period; the rules of conditional probability suggested that this degree of intermittency of driving would inflate the apparent relation between cellular-telephone use and motor vehicle collisions by a factor of 1.5 (1 ÷ 0.65).20,21 Our estimates of relative risk were therefore divided by this factor as one way of adjusting for the intermittency of driving.
To examine the robustness of our analysis, we also tested a different adjustment that relied on individual driving patterns. To do so, between October 25 and November 28, 1996, we attempted to contact all subjects who had used their cellular telephones in the 10 minutes before the collision or the 10-minute control period. We asked each person to remember his or her driving pattern on both the day of the collision and the day before the collision. We then recalculated relative risks by limiting the analysis to subjects who were confident that they had driven a motor vehicle during both periods on both days.
Ethical Issues
The protocol was approved by the University of Toronto Human Ethics Committee, and all participants provided informed consent. Private industry supplied telephone records but otherwise had no involvement in data collection or analysis or funding the study. Individual billing records were obtained directly from cellular-telephone carriers who provided records for 100 consecutive days of telephone use for each person and who were not told which particular date was the day of the collision. Police reports were obtained directly from police departments; they, in turn, were not provided copies of the drivers' cellular-telephone records.
Statistical Analysis
The sample size was calculated to provide an 80 percent chance of detecting a doubling or halving of collision rates. Relative risks were estimated with methods for matched-pairs studies on the basis of exact binomial tests and conditional logistic-regression analyses.22 Confidence intervals for the relative risks were derived with the bootstrap bias-corrected method and accounted for the uncertainty in the adjustment for intermittency of driving.23,24 Modifications of the relative risks were assessed by comparing different subgroups, with particular attention to the prespecified contrast between hand-held cellular telephones and models that leave the hands free. All P values were two-tailed, and all relative risks were computed with 95 percent confidence intervals.
Results
We approached 5890 drivers, of whom 1064 acknowledged having a cellular telephone and 742 consented to participate in the study; the billing records of 699 of these drivers were located (Table 1). The collision times were exact for 231 subjects and inexact for 468. The group placed a total of 16,870 cellular-telephone calls and received 3643 calls during the week before the collisions (average, 3.4 calls placed and 0.7 call received per person each day). The average duration of the calls was 2.3 minutes, and 76 percent lasted 2 minutes or less (similar to cellular-telephone calling patterns elsewhere25). The monthly bill in U.S. currency for the average participant was $72, which was greater than that for the average subscriber in Toronto or the average subscriber in North America ($53 and $51, respectively).26,27,28
|
At follow-up in 1996, we located 145 subjects, of whom 72 (50 percent) were confident that they had driven during both the hazard period and the control period. Restricting our analysis to this subgroup yielded an estimated relative risk of 7.0 (95 percent confidence interval, 3.7 to 15.5) associated with cellular-telephone use. An analysis that included the entire cohort of 699 drivers and used alternative comparison days yielded similar estimates of the relative risk of a collision (Figure 1). All the alternative estimates of relative risk were adjusted for intermittent driving, and all were statistically significant (P<0.001).
|
|
|
|
Discussion
We found that using a cellular telephone was associated with a risk of having a motor vehicle collision that was about four times as high as that among the same drivers when they were not using their cellular telephones. This relative risk is similar to the hazard associated with driving with a blood alcohol level at the legal limit.29,30,31 We also found that cellular telephones have benefits, such as allowing drivers to make emergency calls quickly. A few drivers used their telephones only in the aftermath of a collision, thereby gaining some potential benefits and incurring no potential risks due to telephone use. In general, cellular-telephone calls were brief and infrequent, which explains why the rapid growth of this technology during recent years has not been accompanied by a dramatic increase in motor vehicle collisions.32
We observed no safety advantage to hands-free as compared with hand-held telephones. This finding was not explained by imbalances in the subjects' age, education, socioeconomic status, or other demographic characteristics. Nor can it be explained by suggesting that those with units that leave the hands free do more driving. One possibility is that motor vehicle collisions result from a driver's limitations with regard to attention rather than dexterity.33 Regardless of the explanation, our data do not support the policy followed in some countries of restricting hand-held cellular telephones but not those that leave the hands free.
Three weaknesses of this study should be pointed out. First, we studied only drivers who consented to participate. The fact that some persons chose not to consent might have caused us to underestimate the risks associated with telephone use if these people declined because of concern about personal liability. Second, people vary in their driving behavior from day to day — a fact that makes the selection of a control period problematic. However, it would be difficult to explain all our findings on the basis of different driving patterns, and in particular, this factor would not account for the similar results for those who remembered driving during both periods on both days. Third, case–crossover analysis does not eliminate all forms of confounding. Imbalances in some temporary conditions related to the driver, the vehicle, or the environment are possible, but we believe such factors are not likely to account for the magnitude of the association we observed.
Our study indicates an association but not necessarily a causal relation between the use of cellular telephones while driving and a subsequent motor vehicle collision. For example, emotional stress may lead to both increased use of a cellular telephone and decreased driving ability. If so, individual calls may do nothing to alter the chances of a collision. In addition, our study did not include serious injuries; hence, we do not know how — or whether — cellular-telephone use is associated with motor vehicle fatalities. Finally, the data do not indicate that the drivers were at fault in the collisions; it may be that cellular telephones merely decrease a driver's ability to avoid a collision caused by someone else.
We caution against interpreting our data as showing that cellular telephones are harmful and that their use should be restricted. Even if a causal relation with motor vehicle collisions were to be established, drivers are vulnerable to other distractions that could offset the potential reductions in risk due to restricting the use of cellular telephones. Regulations would also mean reducing benefits; in Canada, for example, half a million calls to 911 emergency services are made from cellular telephones each year.34 Yet proposals for regulation are not unreasonable, since poor driving imposes risks on others. Public debate is needed, given that cellular telephones contribute to improvements in productivity, the quality of life, and peace of mind for more than 30 million people in North America alone.
The role of regulation is controversial, but the role of individual responsibility is clear. Drivers who use a cellular telephone are at increased risk for a motor vehicle collision and should consider road-safety precautions. For them as for all other drivers, these include abstaining from alcohol, avoiding excessive speed, and minimizing other distractions. Additional strategies might include refraining from placing or receiving unnecessary calls, interrupting telephone conversations if necessary, and keeping calls brief — particularly in hazardous driving situations. Physicians should also learn to recognize patients who are at risk for a collision and who may benefit from advice regarding safety.35,36,37,38,39,40 Even limited success in reducing risk may prevent some of the death, disability, and property damage related to motor vehicle collisions.
Supported by a Career Scientist Award from the Ontario Ministry of Health (to Dr. Redelmeier), a grant from the Natural Sciences and Engineering Research Council of Canada (to Dr. Tibshirani), and a grant from the Ontario Ministry of Transportation.
We are indebted to Bell Mobility, Inc., Rogers Cantel, Inc., and the Metropolitan Toronto Police Department for their cooperation; and to the following people for help with specific aspects of the study: Mary Chipman, Ruth Croxford, Allan Detsky, Bradley Efron, Trevor Hastie, Mark Henkelman, Philip Lavori, Barry McLellan, David Naylor, Simon Min, Miriam Shuchman, and Amos Tversky.
Source Information
From the Departments of Medicine (D.A.R.), Preventive Medicine and Biostatistics (R.J.T.), and Statistics (R.J.T.) and the Program in Clinical Epidemiology and Health Care Research (D.A.R.), University of Toronto, Toronto; and the Division of Clinical Epidemiology, Sunnybrook Health Science Centre, North York, Ont., Canada (D.A.R.).
Address reprint requests to Dr. Redelmeier at Sunnybrook Health Science Centre, G-151, 2075 Bayview Ave., North York, ON M4N 3M5, Canada.
References
| |||||||||||||||||||||||||||||||||||||||||
Related Letters:
Cellular Telephones and Traffic Accidents
Cohen P. J., Quinlan K. P., Paltiel O., Ambrose A., Redelmeier D. A., Tibshirani R. J., Maclure M., Mittleman M. A.
Extract |
Full Text
N Engl J Med 1997;
337:127-129, Jul 10, 1997.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
