Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Campaigns about alcohol and driving

Systematic review

Many years ago Bandolier saw a drunk driver take a corner at an insane speed and crash head-on into another car. The injuries were horrible, more to the crashed-into than the crasher. So campaigns in the media to tell people not to drink and drive have seemed like a good idea. Whether they work is another matter. A systematic review of studies [1] is both a compelling read, because it makes one think about what is involved in media campaigns, and provides compelling evidence that mass media campaigns work.


This is not familiar territory for most of us, so a brief introduction to some aspects of media campaigns might help. It isn't as simple as one might think.

First is the content of the message. Is it going to be fear (pictures of bodies, drunk drivers in handcuffs and irons, courts, jail, whatever)? Is it going to be sweet reason (nominated drivers, responsible citizenship)? Whatever the choice, content is crucial, and may have to alternate between fear and reason to maintain awareness over the longer term.

Second is message delivery. The message has to be of good quality for a start. Think of the old-fashioned public service messages with cheap graphics and voice-overs compared with modern campaigns that are indiscernible from advertisements for cars, soft drinks, or beer.

Then there is the issue of testing the message and delivery before the campaign starts. You might know what it means, but the people on the receiving end might have a different perception. The paper gives the example of a campaign for responsible drinking where a third of viewers thought it was an advertisement for beer. The message and delivery are likely to change depending on the target audience, and with time.

Systematic review

For inclusion, studies had to be primary research published in English up to end-2001, related to mass-media interventions, have objective data on one or more outcomes related to alcohol-impaired driving, and meet minimum quality criteria for study design and execution. This last was a set of criteria laid out for evidence-based community preventive services.

Studies were only accepted if levels of other alcohol-impaired driving activities did not change substantially, or where statistical models accounted for any changes that did occur.

Studies were classified by whether they concentrated on legal consequences, or social and health consequences of drinking and driving. Outcomes used were fatalities or injuries related to alcohol-impaired driving, or people driving over the legal limit.


There were six studies between 1975 and 1998, from Australia, New Zealand, and USA, with follow up from two to 37 months. A variety of interventions were used, and most were before-after studies, as well as time series with concurrent comparisons. All were evaluated in areas where actual enforcement levels during the campaign were at approximately the same level as before the campaign.

Two studies reported the proportion of drivers who were over the legal limit. Both had substantial reductions, of 30% and 37% in the number of drivers over the limit, and one had a very large sample base.

Four studies reported all crashes, or fatal, injury, or alcohol-related crashes where the driver had been drinking. All showed reductions, of between 6 and 18%.

Two studies also reported on cost-benefit analyses. One Australian study estimated costs of the campaign were (in US$ adjusted to 1997) $400,000 for advertising development, supporting media, media placement and concept research. Savings from medical costs were $3.2 million per month, with total cost savings (medical, productivity, pain and suffering, property damage) of £8.3 million per month. A US study estimated total costs of a campaign to be $800,000, with savings of $7 million.


Preventative campaigns tend to be sporadic. For driving, in the UK we have a Christmas and New Year campaign, but little for the rest of the year. Perhaps we should do this more often, because the evidence is that it works. This may be a different sort of evidence from that we are used to, but one way of reducing medical costs is to advertise effectively against drinking and driving.

One additional thought is that this paper, while in an area unfamiliar to most of us, talks about evidence in ways we have come to expect. Quality, validity, and size are themes that run throughout. That is what makes the result believable.

A second additional thought is why limit this to drink driving? Why not ask, and try and answer, questions about getting over messages about healthy living, vegetables, fish, obesity, exercise, or sexually transmitted infections?

This paper opens a door on delivery of messages on health. It directly addresses some fundamental themes of how to make such media campaigns work effectively, and about how to assess their effects. One might even contemplate a future in which evidence-based advertising on good health vied with salt and fat rich snacks and prepared meals. Now there's a thought.


  1. RW Elder et al. Effectiveness of mass media campaigns for reducing drinking and driving and alcohol-involved crashes. A systematic review. American Journal of Preventive Medicine 2004 27: 57-65.

previous or next story