Intended for healthcare professionals

Rapid response to:

Papers

Role of mobile phones in motor vehicle crashes resulting in hospital attendance: a case-crossover study

BMJ 2005; 331 doi: https://doi.org/10.1136/bmj.38537.397512.55 (Published 18 August 2005) Cite this as: BMJ 2005;331:428

Rapid Response:

Authors' reply

We have read the responses to our study with interest and have a
number of comments to make.

Absolute versus relative risk:
In this study we examined the effect of mobile phone use on the odds of
having a crash requiring hospital attendance and demonstrated a fourfold
increase in the likelihood of crashing. This is akin to calculating a
relative risk, as distinct from an absolute risk. The study did not
determine the absolute risk of crashing while using a mobile phone. Of
course, the risk of having a crash during a given trip is usually very low
and may vary by some degree depending on individual factors. Our findings
suggest that using a mobile phone will increase that risk fourfold. Thus
the absolute risk of having a crash while using a mobile phone remains low
and this explains why many drivers have used a phone while driving but not
had a crash while doing so. Despite this, given the large number of road
injuries and deaths each year, coupled with the fact that the prevalence
of mobile phone use varies and may be high in countries without
legislation, we believe that a factor that increases the likelihood of a
serious crash by fourfold is important.

Other driver distractions:
We recognise that mobile phone use is one of many potential distractions
that may occur while driving. However, the availability of phone activity
records during the hazard and control intervals makes this distraction
particularly amenable to study using a case-crossover design.
Furthermore, as the use of mobile phones continues to increase worldwide,
the potential impact on road safety is a concern.

Mobile phone use while driving has a number of potential effects:
cognitive (conversation), visual (looking for the phone or at the display
or keypad), auditory (listening to conversation) and manual (holding the
phone or dialling a number). Under certain circumstances, any of these
could lead to an adverse event.

The role of additional occupants was raised. However, measuring
conversation as an exposure using a case-crossover design is difficult as
validating conversation during hazard and control intervals would be
limited to self report. The presence and number of passengers has been
associated with crashes resulting in driver deaths among teenagers.1 That
said the effect of passengers varies by the age and gender of the driver
and passenger, and among adult drivers, passengers can have a protective
effect. Generally-speaking, as discussed by Dr Grove, a passenger is aware
of the road conditions and can moderate his or her conversation as needed.
The same cannot be said of a caller on the other end of a phone line. A
recent on-road experimental study reported that in the most demanding
urban roads conversations with passengers were suppressed, whereas mobile
phone conversations were not.2

The study by Stutts and colleagues (2001)3 found that distraction was
cited in 8.3% of crashes from the US National Accident Sampling System
Crashworthiness Data System between 1995 and 1999 (presence of distraction
unknown or no driver present in 35.9% of crashes) and Dr Grove has made
some valuable observations that we will not repeat here. This study was
not designed to provide risk estimates. For this, exposure data from an
appropriate group of controls (or an appropriate control window) would
have been needed for comparison. Moreover, the ownership and use of
mobile phones has increased since that study was conducted, and police-
reported data on driver distractions at the time of a crash may be
incomplete, inconsistent and unreliable.

Methodological issues:
The only drivers we could include in the case-crossover analysis were
those who had (1) driven during at least one control interval and (2) had
phone activity records available. This is why the final number of drivers
in the case-crossover analysis was 456. Of those, 40 (9%) had used a
mobile phone in the period up to 10 minutes before the crash. We stress
‘up to’ because for those drivers who had driven less than 10 minutes
prior to the crash (63%) we considered only phone activity during the
stated duration of that trip (for both the hazard and corresponding
control intervals). Of the 941 drivers who were interviewed, 72 (8%)
reported using the phone during the trip in which they crashed (Table 1).4

Dr Lombardi raised the possibility of uncontrolled within-subject
confounding in our study. This is an issue we discussed in the
limitations section ‘circumstances of the control driving intervals may
have differed from the crash driving interval’. While this problem is to
an extent unavoidable, we did plan our study to explore some of these
transient factors. Firstly, we collected information on the presence of
other driver factors that may have contributed to the crash outcome
including speed, alcohol, drugs and fatigue (Table 2).4 Secondly, we
collected self reported data on weather conditions at the time of crash
and during the control intervals and validated the information using
rainfall reports from the Australian Bureau of Meteorology. While not
presented in the paper, weather conditions (defined as wet or dry) did not
alter the association between mobile phone use and crash. Finally, we
asked drivers about the trip route during the hazard and control
intervals. About half (47%) were travelling along the same route during
both intervals.

Culpability:
We did consider driver culpability in planning the study and collected
information on the circumstances of the crashes from drivers. While
culpability was clear for some crashes, often it was not. Among those
drivers involved in paired crashes (i.e. crashes in which two or more
parties involved in the same crash were interviewed) accounts sometimes
varied. As a result of the difficulty in assigning fault, we did not
explore this further. Importantly, it is possible that the use of a
mobile phone may impair a driver’s ability to avoid a crash, even if the
driver is not at fault.

We thank all respondents for their contributions.

1. Chen L-H, Baker SP, Braver ER, Li G. Carrying passengers as a risk
factor for crashes fatal to 16- and 17-year old drivers. JAMA
2000;283:1578-82

2. Crundall D, Bains M, Chapman P, Underwood G. Regulating
conversation during driving: a problem for mobile telephones?
Transportation Research Part F 2005;8:197-211

3. Stutts JC, Reinfurt DW, Staplin L, Rodgman EA. The role of driver
distraction in traffic crashes. Prepared for AAA Foundation for Traffic
Safety, May 2001

4. McEvoy SP, Stevenson MR, McCartt AT, et al. Role of mobile phones
in motor vehicle crashes resulting in hospital attendance: a case-
crossover study. BMJ doi:10.1136/bmj.38537.397512.55 (published 12 July
2005)

Competing interests:
Authors of the study

Competing interests: No competing interests

15 September 2005
Suzanne P McEvoy
Senior Research Fellow
Stevenson MR, McCartt AT, Woodward M
The George Institute for International Health PO Box M201 Missenden Rd Sydney NSW 2050 Australia