Table 4

Main findings of systematic reviews on legislative interventions to improve health through transport

Modes of interventionQuality IndicesMain results
RR, risk ratio; OR, odds ratio; RCT, randomised controlled trial; NS, not significant at p<0.05 level.
Drinking and driving legislation, including administrative per se, random screening and lowering the legal blood alcohol limit449,50A systematic review on drink driving control showed that licence suspension, illegal and administrative per se laws, selective and regular enforcement patrols and sobriety checkpoints were most effective, with typical effect sizes of around 10% reduction in a variety of outcomes.
Several studies on mandatory jail sentences showed increases in crashes following implementation.
651Laws requiring a reduction to maximum 0.02% blood alcohol concentration associated with reduction in night time injuries of 17% (NS); 12% reduction in injuries in men, 24% in women; 17% reduction in fatal crashes among younger drivers (p<0.001), 1% in older drivers; 22% net reduction in fatal crashes.
Pre-post with interrupted time series (1 study): 4% reduction in serious injuries using time series, 6% reduction using pre-post – both NS.
Interrupted time series (1 study): 11% or 33% reduction in “had been drinking” crashes depending on model chosen.
552Evaluation of licence suspension or revocation through administrative determination showed no clear effect in 1/3; in 1/3, recidivism in intervention v controls OR 0.60 (0.54 to 0.68) up to but not after 36 months; in 1/3, intervention v controls in first year – drunk driving offences OR 0.78 (0.76 to 0.79), traffic crashes OR 0.65 (0.63 to 0.67) and alcohol related crashes OR 0.73 (0.70 to 0.77).
653Random breath testing reduced hospital admissions by 20%, reduced deaths and injuries by 17–35%, reduced night-time crashes by 18–19% and reduced charges for drink driving.
Checkpoints reduced night-time crash rates by 10–38% and reduced fatal crashes by 17–25%.
Graduated driver licensing among young drivers454Graduated driver licensing was associated with a reduction in hospital admissions of up to 23% and a reduction in deaths of 5.5% in 15–19 year olds; however, there was a simultaneous reduction in 15–19 year olds who drove and a 5% decrease in the teenage population in New Zealand.
A provisional licencing programme for 16–17 year olds showed a 5% decrease in daytime crashes; no effect of night time driving restriction; and a 10% decrease in traffic violation convictions.
Curfew laws in under-18s: (1/4) found no apparent effects; (3/4) found decrease in fatality by 23–28%.
Car safety belt laws (only for adults)355Prevalence of seat belt use increased by 1.08–1.3 times after laws introduced.
Primary enforcement compared with no laws found 1.5–4.1 times more prevalent seat belt use (17 studies); one outlier of 15.4 times more use of seatbelts; RR fatal injury 0.69 to 0.97 (20 studies) but 1.12 (NS) in 1 study; serious non-fatal injury RR 0.20 to 0.89 (11 studies).
Secondary enforcement compared to no laws found a prevalence of seatbelt use 2.1–2.6 times higher in the former group (6 studies); RR fatal injury 0.62 to 1.03 (7 studies), but no significant value over 1.00; serious non-fatal injury RR 0.75 to 0.85 (4 studies).
Any law compared with no law: 4 studies found that relative risk of fatal injury was 0.91 to 0.95 in the former.