Mobile phones and driving

BMJ 2014; 348 doi: http://dx.doi.org/10.1136/bmj.g1193 (Published 04 February 2014) Cite this as: BMJ 2014;348:g1193
  1. Charles Pless, family physician 1,
  2. Barry Pless, professor emeritus, McGill University2
  1. 1Clinique de Médecine Familiale, Notre-Dame-CSSS Jeanne-Mance, Montreal, QC, Canada
  2. 234 Lansdowne, Westmount, QC, Canada, H3Y 2V2
  1. barry.pless{at}mcgill.ca

Time to act: they’re responsible for a quarter of crashes in the US

Although a review of the recent literature found that the evidence for a causal association between mobile phone use while driving and crash related injuries was not clear cut,1 with a quarter of crashes in the United States now attributed to mobile phone use, we can’t wait for perfect evidence before acting.2

In 1997, Redelmeier and Tibshirani found that mobile (cell) phone use was associated with a quadrupled risk of crashes,3 although last year a study cast doubt on some components of the association.4 While most early studies unequivocally supported the view that mobile phones made driving more dangerous, some later reports arrived at contradictory conclusions, especially with regard to hands-free phones.5 Part of the confusion may be due to the mix of laboratory and observational epidemiological studies that characterise this field. Nevertheless, given the proliferation of mobile phones, the prevalence of distracted driving is undoubtedly increasing.

Texting is obviously the riskiest activity because the distraction is cognitive and visual; handheld phone use (particularly when making or receiving calls) comes next; and hands-free use is probably the least dangerous. Convincing causal associations are notoriously difficult to determine, particularly in this area. Causality can be inferred only from randomized trials, and it is doubtful if a real world trial could be designed that would receive ethical approval. Laboratory studies using such designs are possible (and plentiful) but not persuasive. However, physicians and policy makers must often make decisions and act before they have solid proof. We therefore decided to proceed by taking as “given” that the risk was causal, substantial, and likely to grow unless more successful preventive measures are introduced. Possible interventions include the “usual suspects”: education, legislation, and technology.

Education is the preferred choice of many, especially governments, because it is inexpensive, inoffensive, and politically “easy.” Generally, health education is of two kinds: counseling of individual patients by health professionals or more broadly targeted media campaigns. Evidence supporting the effectiveness of patient education is limited.6 In some domains, such as smoking cessation, physician counseling can be surprisingly effective.7 8 Counseling may be more effective when provided in an appropriate context, such as in the use of seat restraints with parents of newborns.9 Experience suggests that patients in the emergency room after an injury are also more open to receiving messages about preventive strategies and equipment. But, even if we had solid evidence that counseling about the dangers of distracted driving was effective, it is unlikely that it would reach those at highest risk or be the best use of limited physician time.

Another, perhaps more powerful, form of education comes through the media. Public service announcements may prove to be persuasive, whether on TV or the web. So far, no studies have determined how often mobile phone users watch such announcements or what effect they have on driving habits. We can, however, assume that over time they would reinforce other measures, much as they did with seat restraints.

Legislation also seems promising, but here too evidence is inconsistent. Some studies show that laws prohibiting mobile phones while driving result in reduced phone use but not necessarily in fewer crashes or injuries; some show both and some show neither.10 These disparities may reflect variable enforcement. Clearly, a well enforced law with stiff penalties and highly visible enforcement will be more effective than a weaker one. Studies concluding that legislation is ineffective often neglect these crucial aspects. Enforcement requires that distracted driving be classed as a primary violation, which would enable police to stop drivers who are using phones regardless of whether another offence has been witnessed. As well as higher fines and more license demerit points, increases in insurance premiums or deductibles for crashes related to distracted driving may also be needed to change drivers’ behavior and initiate a new culture.

The paradox seems inevitable: the most promising solution to distracted driving caused by the mobile phone may well be more technology. Johnston has described some of these measures and cast them in the context of behavioral economics.11 The idea is to find gadgetry that nudges people towards desirable behaviors. This includes software that prevents texting while driving set as a factory default; convenient mobile phone pull-out areas with free wifi access; and automatic messages informing callers that the recipient is driving. Other fixes include a sensor such as a signal jamming key that prevents mobile phone reception when the ignition is engaged (http://key2safedriving.com). Another gadget uses software to detect that a mobile phone is in a moving car; non-emergent outbound calls are then blocked, inbound calls are routed to voicemail, and inbound text messages are stored (www.aegismobility.com). There are now many apps that can be installed to achieve these goals. Ultimately, a technical solution is needed that blocks texting and conversations by drivers while permitting passengers to use their phones as they wish. Until nudging works fully, regulatory bodies must be instructed to incorporate the best available technological preventive measures into all new mobile phones and cars.12

In most countries, impaired driving is a criminal offence and there is a strong social taboo against drinking and driving. Unfortunately, there was a long delay between the first scientific evidence and the public’s recognition that drunk driving is unacceptable. We cannot accept such a long process in the case of distracted driving. The stakes are too high. Studies must continue, but like the emergency room doctor, society must accept that we cannot always wait for perfect evidence to act. Doing nothing, or avoiding the tough options, can have disastrous consequences.


Cite this as: BMJ 2014;348:g1193


  • Competing interests: We have read and understood the BMJ Group policy on declaration of interests and declare the following interests: None.

  • Provenance and peer review: Commissioned; not externally peer reviewed.