BACKGROUND: One fifth of drivers convicted of drink-driving for the first time will be convicted again for the same offence in subsequent years. Lecture-based educational programmes are believed to reduce recidivism. Little is known about the modalities of enhancing the benefits of such programs.
METHODS: This parallel randomised controlled trial measured the effects of the duration of lectures and the presence of a close relative 'in class' on rates of recidivism during the ten years following an initial drink-driving conviction. Of 1588 drivers in the Canton of Geneva convicted of a first-time offence with a blood alcohol concentration (BAC) of between 0.80 and 2.49g/kg from May 2001 to February 2004, 727 agreed to participate and were randomly assigned to either a seven-hour series of lectures, a four-hour series with a friend or close relative, or a brief two-hour lecture. Time until recidivism was retrieved from a national registry that contains details of recidivism that took place up to ten years after the first offence.
RESULTS: Significant effects of briefer lectures over the standard day-long series of lectures were observed only during the most influential time period with regards to recidivism levels-the two years following the intervention. Replacing the usual one-day series of lectures by briefer two-hour lectures would reduce, by 25% (CI95%; 3-44%), the risk of recidivism.
CONCLUSION: This study does not support policymakers' decision to rely on a seven-hour series of lectures to decrease DUI recidivism. The advantages of shorter lectures over no lecture still need to be evaluated.