To the Editor: Taylor et al found that: “Mobile phone use is common among Melbourne metropolitan drivers despite restrictive legislation” and suggest that this constitutes “a preventable exposure to injury risk”.1 This raises two questions:

• Does mobile phone use while driving affect road safety?

• If so, do hands-free devices reduce the risk?

In their introduction, Taylor et al cited six publications to provide evidence that the use of a handheld mobile phone while driving increases the risk of a road crash. The studies by Lamble et al2 and McKnight and McKnight3 involved a hands-free device and examined driver impairment, not crash risk. The three papers by Violanti4-6 had significant limitations, including no phone billing information to demonstrate that drivers were using their phones at the time of the crash,4-6 reliance on police accident reports that may have involved more thorough investigations into fatal crashes than non-fatal ones,5 and small sample size with only 14 mobile phone users in one study.6 These limitations reduce the validity of the research.

The best of the epidemiological studies was a case-crossover study of 699 drivers in collisions involving property damage only.7 However, the oft-quoted four-fold increase in risk comes from the analysis of mobile phone use in a 10-minute hazard interval before the collision. This does not provide conclusive evidence that these drivers were on the phone at the time of their crash and indicates a statistical association only. Although shorter hazard intervals were also examined, one needs to be wary of the potential for misclassifying post-crash calls as pre-crash calls because the time of collision may be imprecise, mobile phone use is common following a crash and a call to the emergency services may not be the first call made after the event. If we conclude that the data are valid despite these limitations, then the fact that hands-free models did not reduce the risk must be noted.

Returning to our questions, although there is good evidence demonstrating driver impairment in laboratory-based studies, the epidemiological research has limitations that need to be dealt with to determine the real-world effect of mobile phone use while driving. We are currently undertaking two large epidemiological studies in Perth, involving about 2000 drivers over an 18-month period. The limitations have been addressed in the design of our studies. Furthermore, the evidence to date suggests that hands-free devices do not confer a safety advantage and this issue should not be ignored in driver education.

To the Editor: In “In this issue” of the Journal of 4 August 2003,1 you write regarding “Dialling drivers” that “the jury is still out on the health effects of mobile phones”.

Well, the jury is back with a guilty verdict. In June 2003, Injury Prevention2 quoted an evaluation by the Harvard Center for Risk Analysis that “the use of cell phones by drivers may result in about 2600 deaths, 330 000 moderate to critical injuries, 240 000 minor injuries and 1.5 million instances of property damage in America per year”.3

Taylor et al suggest, “Further interventions aimed at decreasing mobile phone use among drivers should be considered.”4

Occupational safety professionals consider that a worker not complying with the safe practices for using a tool should be offered remedial education. If education fails, they stop the worker using that tool.

Wise parents also consider taking away a child’s toy until the child can learn to use it safely.

And so with mobile phones used while driving. Driver safety education is not very effective. Police have powers to impound items related to other offences, and so should have powers to impound mobile phones used when driving. The driver could then claim it, say, four weeks later, from the police station on payment of a fee-for-service to the police that covers, at least, the relative value of the expenses of the police. The driver would also incur demerit points. Repeated offences would mean they forfeit the phone or their licence.

To the Editor: The claim made in the recent article that mobile phone use while driving is more dangerous than drink driving is misleading.1

The 1997 study by Redelmeier is often misinterpreted and cited for the proposition that driving while using a mobile is the same as driving drunk. However, Redelmeier wrote to the New England Journal of Medicine to correct this inaccuracy, saying, “. . . alcohol circulates in the blood for hours, whereas a telephone call lasts only minutes. The cumulative risks associated with intoxication are greater than those associated with cellular telephones”.2

This is supported by a recent Australian study, which compared the blood alcohol levels of drivers involved in real car crashes, rather than driving simulators, and found the risk of an accident was increased by 25 times at a blood alcohol concentration of 0.08.3 Mobile phones have not been shown to present this level of risk in any research.

In 2002 in the United States, alcohol was a factor in about 41% of all fatal traffic crashes and in 6% of all crashes.4 In comparison, data collected by about 20 state highway authorities show that mobile phones were a factor in an estimated one half of one percent of all accidents in the US last year.

Furthermore, mobile phone subscribers provide the extra eyes and voice for police in reporting aggressive, reckless or drunk drivers, accidents and other road hazards. Almost a third of all genuine calls to 000 are made from mobile phones.5

A recent US survey found that at any given time only 3% of drivers are actively using their mobile phones,6 although it is legal to use a handheld phone in almost all states. Therefore, Taylor’s overall result that less than 2% of Melbourne drivers use a handheld mobile phone, while illegal in Australia and undesirable, is not unexpected.

However, the unrelated and misleading comparison made with drink driving is not supported by the facts.

No one is questioning that mobile phone use imposes physical, visual, and cognitive demands on the driver. Although technology can help to address physical and visual factors, education is required to address cognitive factors. The Australian Mobile Telecommunications Association has developed 10 safety tips for mobile phones and driving (see www.amta.org.au) and, by adhering to these simple common-sense practices, drivers can make full, productive and safe use of mobile phones.

In reply: McEvoy and Stevenson raise some important issues. The first relates to the confusion between driver impairment and crash risk. Intuitively, this association seems valid, as any level of driver impairment could be expected to affect driving skill. However, they are correct to suggest that the two should not be used interchangeably without supporting evidence.

Secondly, I agree that the quality of evidence directly linking mobile phone use with crash risk is poor.1-4 This largely relates to the difficulty in confirming mobile phone use at the exact time of the crash. Reported direct observation is uncommon, billing records are inexact, and self-report may be subject to prevarication bias.

The use of hands-free devices was not examined in our study, mainly because of difficulties in detecting their use. There is anecdotal evidence of a trend towards the use of these devices while driving. However, while their use might avoid the need to physically hold the phone, they may not significantly diminish driver impairment resulting from distraction.

Many questions remain, and I encourage McEvoy and Stevenson in their endeavour to more clearly evaluate the real-world risk of mobile phone use, both handheld and hands-free, by drivers.

Chalker et al draw attention to the comparison of crash risk for mobile phone use while driving and drink driving. I acknowledge that interpretation of published studies is confusing. Redelmeier’s statement that alcohol circulates for hours and that a telephone call may last for only minutes relates to individuals. From the highway perspective, when one driver completes a call, another is likely to be starting one and effectively assuming the increased collision risk. This concept is consistent with our findings. Almost 2% of drivers were using mobile phones when they passed our observation points, and were therefore at risk at that time. The exact extent of this risk awaits clarification. Chalker et al provide US alcohol and mobile phone related crash statistics. Unfortunately, the latter were not referenced and their value is therefore questionable.

Finally, Chalker et al are to be commended for publishing safety tips for mobile phone use while driving. However, their claim that common-sense practices can make mobile phone use safe is extraordinary and disregards emerging evidence. Indeed, this statement appears to contradict their first safety tip, which states “a hands free device can reduce the physical effort to make and receive calls; however, it alone doesn’t make using a mobile phone while driving safer”.5 At best, therefore, common-sense practices will not make mobile phone use while driving safe, only possibly safer.