This report addresses the effects of cell phones on driving by means of a review of the literature and an analysis of scientifically credible epidemiological and driver performance studies. A total of 84 articles were obtained covering the period 1969 – 2004. Sixty-eight articles were research papers measuring driving performance while using a cell phone and 16 articles were epidemiological studies that examined cell phone usage and their relationship to vehicular crashes. Based on an initial review of this literature, and with the agreement of the Alberta Motor Association (AMA), the analysis and report were focused on 15 epidemiological studies and 22 performance studies which were used to answer four questions: 1. Does conversation on cell phones, whether hand-held or hands-free, influence driving performance? 2. Are there differences in findings among computer-based studies, driving simulator studies and on-road studies? 3. Does performance differ between hand-held and hands-free cell phones? 4. Are some age groups more susceptible to negative influences of cell phone use on driving? The methodological approach was as follows. Where there were sufficient studies, meta-analyses were carried out to combine study results to answer the above questions. Where there were not sufficient studies, the results of individual epidemiological (i.e., crash risk) and performance (i.e., reaction time and driving variables) studies were reviewed. In addition, because of the availability of a large number of studies, a quantitative analysis of reaction time, as affected by cell phone characteristics, cell phone tasks, driving tasks and driver age was carried out. Finally other study findings of interest are reported. In answer to the first question, conversation on cell phones, both hand-held and handsfree, was found to influence driving performance. Epidemiological findings consistently showed an increase in crashes associated with use of cell phones. However, these studies did not control for exposure. Those who use cell phones more while driving may also drive more. It is well known that more driving leads to higher likelihood of having a crash. The contribution of driving exposure is a particular concern because, in most studies, phone use was not controlled at the time of the crash, and a connection is assumed. A meta-analysis of the performance studies showed moderate-to-large negative effects of the use of cell phones on driving performance. The largest negative effects were found for reaction time (an increase of 0.23 of a second on average and for older drivers, in particular, about 1/2 second). There were lesser size effects for lateral and longitudinal (headway) control, and speed control. In answer to the second question, differences in findings were evident among computer-based studies, driving simulator studies and on-road studies. A meta-analysis of the performance studies showed the strongest effects for laboratory studies, in comparison to on-road or driving simulator studies. Nonetheless, even using the most conservative analysis, on-road studies showed moderate impacts of cell phone use on performance. In answer to the third question, based on the available data, performance did not differ between hand-held and hands-free cell phones. There were insufficient studies to carry out a meta-analysis. A single epidemiological study found an unexpected effect of a slightly higher risk for hands-free use. This may be confounded by exposure to driving as well as exposure to phone use while driving which may differ between drivers using hand-held versus hands-free phones. Most driving performance studies found no difference between hands-free and handheld phones. However, the comparisons made have not focused on those situations in which hand-held phones are likely to be more of a liability with respect to physical demands of driving – for example, while merging into traffic, or while dialling or answering a call. In answer to the fourth question, the evidence is that, should cell phone risk be measured on a per kilometre driven basis, then older drivers are more likely to be at risk of a crash. This is based on an analysis combining studies of reaction time, which found older drivers showing almost a half-second delay in response, when cell phones were in use, compared to one-fifth of a second for younger drivers. However, it is likely that younger drivers use cell phones more while driving. Even though they may not have as poor a reaction time as older drivers, they are more likely to be using the phone while driving and therefore more likely to have a crash while on the phone. This provides an explanation for the results of the three epidemiological studies: two found that it was younger to middle-aged rather than older groups who were most at risk; one found no difference related to age. Our conclusions were similar to those of other reviews of the cell phone and driving literature. Additional findings of interest, not included in the meta-analyses, were that the reaction time increase was greater for lead vehicle braking as compared to other reaction time situations, and that use of a cell-phone while driving reduced the eyes-on-road time while driving, and narrowed the areas to which drivers attended. A number of gaps in research were identified. These include: - Insufficient control for exposure to driving in crash studies; - Insufficient control for exposure to cell phone use, confounding age effects; - Insufficient study of hand-held as compared to hands-free cell phones; - Lack of clarity concerning the timing of the cell phone task and a critical driving event and the performance of the cell phone task; - Lack of clarity regarding the meaning of reported driving performance variables with respect to changes in risk. * Exposure, Cell Phone Use and Crash Risk: In studies of crash risk associated with cell phone use there are issues of both cell phone exposure (frequency of use while driving) and driving exposure (kilometres driven per year), which can confound results. Frequent users of cell phones may be more able to carry out the division of attention required. On the other hand, they are more likely to be exposed to attention-dividing circumstances, and whether or not they can handle themselves better than less frequent cell phone users, they may be more likely to have a crash involving the use of a cell phone. Epidemiological studies are required that consider both cell phone and driving exposure. Performance studies are required to examine the impact on driving of cell phone use, for both experienced and inexperienced users of cell phones. * Cell Phone Use, Age and Driving Experience: While this is expected to change somewhat over the next decade, currently young drivers are more likely than older drivers to be frequent users of cell phones while driving. Young drivers are also at higher risk of a crash. Higher frequency of cell phone use while driving, and greater experience as a driver may both reduce the impact of cell phone use on driving performance. The naïve assumption that either negates the impact of cell phone use on driving needs further investigation. * Hand-Held vs. Hands-Free Cell Phone Use: Studies examining cell phone use while driving have predominantly focused on hands-free versions of cell phones even though the majority of drivers are using a hand-held phone. More studies are needed comparing the differences between these two types of units, particularly looking at newer phone interactions, such as text-messaging, that require more visual processing time. Various interfaces should also be researched in more depth to determine those that may mitigate or exacerbate existing problems with hand-held phones. * Secondary Task Performance: Studies of driver response to a critical event while performing a secondary distraction task, such as using a cell phone, frequently do not report the timing of the critical event relative to the distraction task. Furthermore studies do not generally report whether the drivers ignore the distraction task in order to maintain performance on driving tasks. This information would be helpful in determining safety impacts. * Meaning of Driver Performance Variables: Interpretation of the meta-analysis of cell phone effects on driver performance measures (e.g., lane position) is problematic because of ambiguity in the interpretation of these measures vis-à-vis driver risk. For example, a change in mean lateral lane position may indicate either increased or decreased risk depending on roadway and traffic characteristics. Future research should indicate a priori the pattern in these measures that is associated with greater risk. With respect to policy implications, four issues arise: - Legislation concerning cell phone use while driving and the trade-off of costs and benefits; - The need to consider restrictions on cell phone use for inexperienced drivers; - Private sector response to cell phone crash risk; - Crash reporting and database improvement. Each of these is discussed below. * Legislation: In the debate to prohibit the use of cell phones while driving, policy makers must weigh the social and economic benefits of using cell phones while driving against the growing literature that suggests that using cell phones while driving has negative impacts. Currently, 45 countries have implemented bans on using cell phones while driving. In Canada, only Newfoundland and Labrador has introduced a ban on cell phone use while driving. In the U.S., although the majority of states have looked at the issue only three states now have full cell phone bans. No legislation has dealt specifically with novice drivers and cell phone use. Since younger drivers are heavier cell phone users, and since the least experienced drivers have high crash rates, a specific ban for these drivers may be warranted. * Inexperienced Drivers and Cell Phone Use Graduated licensing programs are becoming the norm to licensing new drivers in North America, and represent an opportunity to address the issue of inexperienced drivers interacting with cell phones. During the probationary period, drivers are only allowed to operate vehicles under restricted conditions. In many jurisdictions, for example, new drivers cannot have any alcohol in their blood while operating an automobile. No restrictions are made for cell phone use. This despite the fact that young people are among the largest users of cell phones, and according to one study, most likely to be involved in distraction related crashes. * Private Sector Response A number of companies have instituted bans on cell phone use while driving by their employees. Phone companies and insurance companies have begun to warn users of the increased crash risk associated with cell phone use. * Better Crash Reporting Currently, our knowledge of cell phone impacts is limited because Canadian accident report forms do not require use of a cell phone at the time of the crash to be reported. This requirement has been added to the forms recommended for use in the U.S. Once reporting improves, it seems likely that the number of crashes identified as being associated with cell phones will grow. Use is high – one study showed that about 60% of drivers had a cell phone and 30% made or received calls on a regular basis while driving. (Author/publisher)
Abstract