Alcohol ignition interlocks may stop repeat drink driving offences, but only as long as they are still fitted

Convicted drink drivers are sometimes offered the choice of a standard punishment, or for an alcohol ignition interlock to be fitted to their car for a fixed period. To operate a vehicle equipped with an interlock, the driver must first give a breath specimen. If the breath alcohol concentration of the specimen is too high, the vehicle will not start. A number of studies have been conducted to see whether the interlock stops drink drivers from offending again. Most of these studies have not been of high quality. The interlock seems to reduce re-offending as long as it is still fitted to the vehicle, but there is no long-term benefit after it has been removed. However, more studies of good quality are needed to confirm these findings. The low percentage of offenders who choose to have an interlock fitted also makes it difficult to reach firm conclusions about their effectiveness.

Authors' conclusions: 

In order to eliminate potential selection bias, more RCTs need to be conducted in this area so that effectiveness, as well as efficacy, can be ascertained. The interlock programme appears to be effective while the device is installed in the vehicle of the offender. Studies need to address ways of improving recidivism rates in the long term, as the major challenges are participation rates, compliance and durability.

Read the full abstract...
Background: 

An ignition interlock device is part of a multi-dimensional programme aimed at reducing recidivism in convicted drink drivers. To operate a vehicle equipped with an ignition interlock device, the driver must first provide a breath specimen. If the breath alcohol concentration of the specimen exceeds the predetermined level, the vehicle will not start. As a measure to reduce circumvention of the device (i.e. someone else blows into the mouthpiece), random retests are required while the vehicle is running. Other components of the drink driving programme include information seminars for the driver and downloading data from the device's data logger, which logs all test attempts and records all passes, warnings and failures.

Objectives: 

To systematically assess the effectiveness of ignition interlock programmes on recidivism rates of drink drivers, by examining rates of recidivism while the ignition interlock device was installed in the vehicle and after removal of the device.

Search strategy: 

We searched The Cochrane Injuries Group's Specialised register (Sept 2002), MEDLINE (1966 to August 2002), PubMed (to Aug 2002), EMBASE (1980 to Sept 2002), TRANSPORT (1988 to 2002 issue 06), CENTRAL (The Cochrane Library 2002, Issue 3), The Science Citation Index (1980 to Sept 2002)
National Research Register (2002, issue 3). We also searched the Internet using various search engines.

Selection criteria: 

Controlled trials in which offenders have been charged with drink driving and have either been sentenced to participate in an ignition interlock programme or the usual punishment (either licence suspension or some form of treatment programme). This study was not restricted by language or status of publication.

Data collection and analysis: 

One randomised controlled trial (RCT) and ten controlled trials were identified, and also three ongoing trials. Data regarding recidivism while the interlock is installed in the vehicle; after the interlock has been removed from the vehicle and total recidivism during the study were extracted and entered into analyses using RevMan.

Main results: 

The RCT showed that the interlock programme was effective while the device was installed in the vehicle; relative risk 0.36 (95% confidence interval 0.21 to 0.63). Controlled trials support this conclusion, with a general trend − in both first-time and repeat offenders − towards lower recidivism rates when the interlock device is installed. Neither the RCT nor the controlled trials provide evidence for any effectiveness of the programmes continuing once the device has been removed.

Share/Save