Childhood Injury Prevention Interventions

Environmental Changes

Background

The difficulty with changing behavior, particularly children’s behavior, makes environmental modification particularly attractive. Changes in the traffic environment have been a traditional component of pedestrian safety programs for many years. Comprehensive programs, such as those in Europe, often go under the rubric of "traffic calming" to denote efforts to slow down traffic, thereby improving safety as well as the aesthetics of the environment.

Traffic calming refers to a group of measures designed to control traffic in an urban residential area. The definition of traffic calming adopted at the 1997 international ITE conference is, "the combination of mainly physical measu res that reduce the negative effects of motor vehicle use, alter driver behavior and improve conditions for non-motorized street users." (ITE 2001).

Measures are designed to reduce the number of commuters using residential streets and to reduce the speed of remaining traffic. Measures can include establishing a hierarchical road system, improving main roads to carry additional traffic and restricting or removing traffic from residential streets with road closures or one-way designations. Environmental changes designed to reduce speed include vertical changes in the street (speed cushions, humped pelican crossings, raised junctions), lateral changes in the street (off-set intersections), constrictions (narrowings, pinch points, pedest rian refuges), gateways at the entrances to the area, build-outs to protect on-street parking spaces, and mini-roundabouts (traffic circles).

Epidemiological studies of environmental risk factors for child pedestrian injury indicate that the likelihood of injury increases for the following conditions: increase in traffic volume (13-14 times) or speed limit (6 times), absence of play areas (5.3 times), poorly protected play area (3.5 times) and high proportion of curbside parking (3.4 times) (Roberts 1995,Wazana 1997).

In this current update (August 2001) we have included a meta-analysis of traffic calming articles published between 1971 and 1994. Additionally, we limited individual articles to those that reported changes in bicycle and pedestrian conflicts or injuries as outcomes. This update excluded individual articles that report changes in traffic volume, motor vehicle speed, or injuries to motor vehicle drivers or occupants, (Brilion 1990, Retting 2001, Kaplan 2000,Corben 1996, Mountain 1989, Engel 1992, Vis 1992, Aljanahi 1999, Cloke, TRL 1999, Holland 1996, Chorlton 1990).

Review of environmental change interventionsn (traffic calming):

Author

Elvik, 2000

Study design and target population

Meta-analysis of 33 articles.

Published from 1971-1994

Studies Include:

1-before-and after study design with matched comparison group.

2-before-and-after study design with general comparison group.

3-Simple before-and-after design with no comparison group.

Intervention

Traffic calming defined as measures to discourage non-local traffic from using residential streets and reducing speed of remaining traffic.

Outcomes

Traffic volume data by type of road before and after.

Traffic volume for whole area before and after.

Accident severity for injury and non-injury (property damage) accidents.

Effects stated only for whole area.

Effects stated for main roads.

Results

Area wide traffic calming re duces the number of accidents by 15% in area affected by traffic calming (main and local roads combined).

Validity ratings assigned to studies. Statistical analysis included random and fixed effects models, tests for heterogeneity of effects and publication bias.

Study quality and conclusions

All studies were non-experimental before and after designs.

None of studies controlled for regression to the mean or long term trends in accident occurrence.

Studies from 8 countries indicate traffic calming reduces accidents.



Author

Hyden, 2000

Study design and target population

Ecologic time-trend study design.

Vaxo, Sweden

4 months and 4 years post construction.

Intervention

Small traffic roundabouts.

Outcomes

Police accident reports, conflicts (cars giving way to other cars, bikes or pedestrians), predicted number of injury accidents, car speeds measured with radar guns.

Results

44% decrease in accident risk for pedestrians and bicyclists.

Study quality and conclusions

Size of roundabout should be as small as possible; one car lane on entrance, circulation area and exit.

Slowing MV speed reduces pedestrian injuries.



Author

Garder, 1998

Study design and target population

Meta-analysis of 33 articles.

Published from 1971-1994

Ecologic study design.

Gothenburg, Sweden.

Intervention

Construction of raised urban bicycle crossings at 44 road intersections.

Comparison group of conventional (unmodified) intersections.

Outcomes

Accidents obtained from police reports and hospital visits. (bicycle and pedestrian injuries recorded).

Bicycle volume and speeds,

MV speeds, conflicts between bicycles, MV, pedestrians at road intersections.

Results

Raised bicycle crossing reduced accident risk by 30% compared to conventional bicycle crossing. Some evidence that pedestrians and motorists also benefited.

Motorists speed reduced; increased bicycle volume on modified roadways; conflicts reduced from 39 per 100 hours of observation before reconstruction to 20 per 100 hours after. Biggest gain for pedestrian; conflicts involving pedestrians decreased.

Study quality and conclusions

Analysis incorporated increase in bicycle exposure (volume).

Environmental changes reduced MV speed and thus bike and pedestrian conflicts & injuries.

Increased bicycle speeds and riding against allowed direction on bike paths increased bike injuries.



Author

Webster, 1996

Study design and target population

Ecologic (before/after study) in UK.

Injuries measured in 72 traffic calming areas; speed data recorded in 32 such areas.

Intervention

20 mph zones established in 200 residential areas;

Traffic speed reduced by environmental measures.

Outcomes

Accidents, speed data.

Results

61% reduction in total injuries.

70% reduction in child pedestrian injuries.

48% reduction in child cyclist injuries.

Speeds reduced by 9 mph (from 34 mph to 25 mph).

Study quality and conclusions

Public reactions favorable.

Long period of data collection, but no control community was used. Study provides some evidence that changing urban environment reduces child pedestrian injuries.



Author

Mackie, 1998

Study design and target population

Ecologic mixed study.

5 towns in England with intervention; 5 control areas.

Intervention

Urban Safety Project.

Outcomes

 

Results

10% accident reduction overall; most benefited are pedestrians, particularly children, and cyclists.

Study quality and conclusions

Use of control areas adjusts for secular trend and changes in traffic volume.


Author

Schnull, 1992

Study design and target population

Ecologic study design.

Nordrhein-Westfallen, Germany, 1983-1990.

Intervention

Speed reduction measures on 28 roads.

Outcomes

Speed measurements.

Accidents and injuries involving MV pedestrians and bicyclists.

Results

Number of injuries involving pedestrians and bicyclists too small to evaluate.

Study quality and conclusions

Description of various interventions to reduce speed. No data presented to show that any individual measures or the total package were effective


Author

Box, 1989

Study design and target population

Ecologic time trend study design.

Naperville, IL.

Intervention

Street lighting, 1984-85 vs 1986-87.

Outcomes

Night accident reduction rates.

Results

Night accident reduction of 36%.

Night-time proportion of pedestrian bicycle accidents reduced 42% mid-block and 17% at intersections.

Lighting cost effective after 1 year.

Study quality and conclusions

Adjusted for average daily traffic volume.

Study illustrates the benefits of properly designed modern lighting.



Author

Engel and Thomsen, 1992

Study design and target population

Before-after study with control group

Denmark, countrywide

Intervention

Traffic calming, with special focus on making some streets priority for pedestrians with speeds of 15 km/hr. In addition, other streets were set at 30 km/hr.

Outcomes

Police-reported pedestrian-MV collisions and collisions involving pedestrian injury

Results

Per kilometer of road: On streets designated as 30 km/hr, 24% reduction in pedestrian-MV collisions and 45% reduction in pedestrian injuries.

Per kilometer traveled by road user: 72% reduction in pedestrian injuries (OR=0.28, 0.08-0.96).

For speed bumps, they calculated that for every 1 cm increase in the height of the bump, there is a 1 km reduction in speed.

Narrowing a road will decrease the speed by about 5 km/hr

Study quality and conclusions

The study is one of a series from Scandinavia and Europe indicating that traffic calming techniques can reduce speed and decrease the risk of injuries.

No separate data for children. Data appear to be reliable since they are based on police reports.



Author

Van Houten and Malenfant, 1992

Study design and target population

Before-after design

Motorists and pedestrians (all ages) in Dartmouth, Nova Scotia.

Intervention

Signs prompting drivers to stop farther back from the crosswalk to increase visibility of pedestrian to other drivers with or without line on pavement indicating where to stop

Outcomes

(1) distance from crosswalk at which car stopped;

(2) driver yielding to pedestrian;

(3) pedestrian-MV conflicts

Results

(1) OR for stopping less than 30 feet from crosswalk was 0.63 for sign; addition of line made little difference;

(2) OR for pedestrian-MV conflicts was 0.40-0.47 following introduction of sign, and 0.19-0.24 following sign plus line. Intervention was equally effective whether or not pedestrian-activated flashing light was used.

(3) The odds of drivers not yielding to pedestrian was 0.86-0.89 after the sign; line had no added benefit.

Study quality and conclusions

The results indicate that the sign did decrease all 3 hazards to pedestrians; with the exception of pedestrian-MV conflicts, line added little. Effects were maintained one month and one year later.

Study had before-after data only; no control groups.



Author

Vis et al., 1992

Study design and target population

Non-equivalent control group design

15 experimental areas in the Netherlands.

Intervention

Use of traffic calming measures to lower speeds to 30 km/hr: signs, speed bumps, narrowing the road, entrance constructions, partial barricades, pedestrian refuges.

Outcomes

Reduction in speed

Reduction in traffic volume.

Police reported MV collisions and collisions with injuries.

Results

Reduction in speed by: humps-40% reduction; narrowed entrance-28%; speed bump-25%; mini roundabout-25%; diagonal closure of street-23%. Signs only did not reduce speed. However, effect of measure varied enormously from one area to another.

Traffic volume decreased by 5-30%.

Injuries from MV crashes decreased by 20% (OR=0.80).

Study quality and conclusions

The measures appeared to slow the speed of traffic and reduce the number of injuries. However, the effect appears to be highly variable. This may be due to the small sample size. In addition, pedestrian injuries and child injuries are not separated out.


Author

Garder, 1989

Study design and target population

Signalization at 115 intersections in Stockholm and Malmoe, Sweden

Use of traffic conflicts technique on 1728 observations

Intervention

Traffic signals at intersections

Outcomes

Number of pedestrian motor vehicle conflicts as a proxy for pedestrian-MV collisions.

Results

In general, at intersections with mean speed >30km/hr, signalization appears to reduce the risk of pedestrian injury by 50%. Pedestrian refuges decrease risk of injury by about 30%.

Study quality and conclusions

No data specifically on children. No actual injury data. Number of observations for particular kinds of intersections was small. No multi-variate analysis.

Summary of environmental modification programs:

There are no randomized controlled trials of traffic calming, and some of the studies use only before-after comparisons, thus not accounting for secular trends. The majority of the studies are from 7 European countries, the remainder from Australia, Canada and the United States. Outcome measures vary among the studies and data specifically on children are largely absent.

Despite these concerns, traffic calming appears to be a promising intervention. It not only reduces the risk of injuries, but also makes the cities more esthetically pleasing. The success of traffic calming is best illustrated by the Elvik 2000 meta-analysis which found area-wide urban traffic calming schemes reduced the number of injury accidents by 15% (25% on residential stree ts and 10 % on main roads). This figure is currently the best estimate of the effect of traffic calming. Studies reporting higher success rates are in general, weak designs without comparison groups.

Changes in transport policy to encourage walking and cycling for short trips and bus, train, air trips for longer distances should be considered. (Roberts 1995) This would include environmental changes to make walking safer and a philosophical shift to value pedestrian and cyclists' convenience over that o f motor vehicle drivers.

Recommendations on environmental modifications:

These are recommended as being a probably effective pedestrian injury prevention strategy. They will fit nicely with the Haddon matrix and the general principles of passive vs. active strategies.