Objectives. We assessed the effectiveness of speed cameras on Barcelona’s beltway in reducing the numbers of road collisions and injuries and the number of vehicles involved in collisions.

Methods. We designed a time-series study with a comparison group to assess the effects of the speed cameras. The “intervention group” was the beltway, and the comparison group consisted of arterial roads on which no fixed speed cameras had been installed. The outcome measures were number of road collisions, number of people injured, and number of vehicles involved in collisions. We fit the data to Poisson regression models that were adjusted according to trends and seasonality.

Results. The relative risk (RR) of a road collision occurring on the beltway after (vs before) installation of speed cameras was 0.73 (95% confidence interval [CI]=0.63, 0.85). This protective effect was greater during weekend periods. No differences were observed for arterial roads (RR=0.99; 95% CI=0.90, 1.10). Attributable fraction estimates for the 2 years of the study intervention showed 364 collisions prevented, 507 fewer people injured, and 789 fewer vehicles involved in collisions.

Conclusions. Speed cameras installed in an urban setting are effective in reducing the numbers of road collisions and, consequently, the numbers of injured people and vehicles involved in collisions.

Road traffic injuries are a major cause of mortality and morbidity worldwide.1 There is evidence that driving speed is a risk factor for road collisions and for increased severity of road injuries.2 The laws of physics support the view that, all else being equal, higher speeds will increase both the probability that an accident will occur and the severity of its consequences.3 Therefore, interventions aiming to reduce speeds through engineering measures and law enforcement have been recommended and widely implemented.

Speed cameras are one of the tools used in enforcing lower speeds. Although they have been widely used and many studies have been carried out to evaluate their effectiveness, evidence on their effects is still scarce. In their systematic review, Pilkington and Kinra reported that existing research consistently shows that introducing speed cameras is effective in reducing road traffic collisions and related casualties, but most studies have not included satisfactory comparison groups or suitably controlled for potential confounders.4

Variables commonly regarded as having potential confounding effects on results of observational before-and-after studies of road safety measures5 include regression to the mean, long-term trends in numbers of collisions or injuries (a systematic tendency for the number of accidents to rise or fall over a period of several years),6 general changes in the number of collisions around the time when the road safety measure is introduced (i.e., differences between the periods before and after the measure is implemented),6 changes in traffic volume,3 and other specific interventions introduced around the same time as the road safety measure.

A suitably designed study is necessary to address these factors, and that was our goal with the present investigation. We assessed the effectiveness of speed cameras installed on the beltway of Barcelona, Spain, in reducing numbers of road collisions, injuries, and vehicles involved in collisions. We also examined differences in effects between daytime and nighttime hours and between weekdays and weekends. Finally, we estimated numbers of collisions avoided and reductions in numbers of people injured and vehicles involved.

Setting and Design

Each year in Barcelona (population: 1.5 million), there are more than 10 000 motor vehicle collisions, more than 13 000 people injured, and approximately 50 road traffic deaths. In 2004, there were 1158 injurious collisions per 100 000 vehicles, 617 per 100 000 inhabitants, and 718 per 1 million km traveled; 56% of collisions involving injuries occurred at intersections.

The Barcelona city road network includes a total of 1275 km; the beltway that surrounds the city accounts for 24.1 km, and streets considered to be arterial roads account for 43.4 km. The beltway has 3 lanes in each direction separated by a median barrier, and it serves to connect the outlying metropolitan area with the city center. It absorbs 20.7% of the total number of vehicle-kilometers traveled in the city area. The speed limit over most of the beltway is 80 km per hour (49.7 miles per hour), with some stretches limited to 60 km per hour; there are no traffic lights or intersections. Arterial roads are defined as streets of 2 to 4 lanes that cross the city and interconnect with the metropolitan area. They have intersections and traffic lights and absorb 21.4% of the total vehicle-kilometers traveled in the area. The typical speed limit on these roads is 50 km per hour.

Eight speed cameras went into operation on the beltway in March 2003 with the aim of reducing the number of road collisions and their consequences. The 8 cameras operate from 22 different sites (i.e., they are moved randomly from site to site). Therefore, although the cameras are visible, motorists never know whether a particular site has a working speed camera. Signs informing motorists of the presence of speed cameras are posted along the beltway. In the period preceding their installation and during their first few months of operation, they were announced to the public through a publicity campaign. The cameras take a photo of a vehicle when its speed exceeds by at least 5% the limit allowed, and a ticket is mailed to the driver within 2 to 3 weeks. The amount of the fine is €300 (approximately US $400) when the speed exceeds by 50% the limit allowed.

We conducted a natural “quasi-experiment” (a time-series study with a comparison group) designed to assess the effects of the beltway speed cameras. The “intervention group” was the beltway, and the comparison group consisted of arterial roads on which no fixed speed cameras had been installed. The pre-intervention period of the study took place from January 1, 2001, through March 31, 2003. The postintervention period spanned April 1, 2003, through March 31, 2005.

Data Source

We derived our data from the local police accident database. In Barcelona, a special division of the police department is responsible for the reporting of road collisions in the city. Trained police personnel report road collisions according to a standard protocol. All collisions that involve damage to property or injuries are reported. We considered only collisions that occurred on the beltway or on arterial roads.

Variables and Indicators

Our outcome measures were numbers of road collisions, numbers of people injured, and numbers of vehicles involved in collisions. All types of collisions were included, rather than only collisions involving injuries (the percentage of noninjurious collisions is fairly constant, at approximately 9%). We did not analyze numbers of deaths separately from numbers of injuries because the number of deaths resulting from beltway collisions is very low (approximately 1 or 2 per year).

There is often heavy daytime traffic on the beltway, particularly on weekdays; during these periods, it is impossible for motorists to exceed speed limits. Thus, speed cameras should be more effective in reducing injuries during nighttime hours and on weekends, when traffic flows are lighter. We conducted separate analyses comparing (1) daytime (7 am to 8:59 pm; this could perhaps have been more appropriately labeled “working hours” because our aim was not to assess the effects of daylight or darkness but to assess commuter traffic flows during these periods) and (2) weekdays (early Monday through late Friday) with weekends (9 pm Friday to 4 am Monday). Month was the unit of analysis.

Statistical Analysis

Initially, we calculated ratios of road collisions per 1000 vehicle-kilometers traveled per year for the beltway and the arterial roads. Subsequently, we analyzed our outcome measures using Poisson regression models.7 After adjusting for potential confounding by linear trends and seasonal patterns, we conducted an intervention analysis in which we fit a dummy variable to compare the preintervention and postintervention periods. The model for each outcome can be summarized as follows:

(1)ln(E[Yt])=B+B1×t+B2×sin(2πt/T)+B3×cos(2πt/T)+B4×Xt,

where T is the number of periods described by each sinusoidal function (e.g., T = 12 months), t is the time period (e.g., t=1 for January, t=2 for February), and Xt =1 represents the postintervention period (Xt =0 otherwise).

We used Stata statistical software to conduct the statistical analyses.8 We analyzed data for 24 months after the intervention. We derived relative risks (RRs) from the adjusted models and calculated attributable fractions from these relative risks ([RR − 1] ÷ RR) to estimate the numbers of prevented collisions and the reductions in numbers of people injured and vehicles involved in collisions.

After 2 years of speed camera operation on the Barcelona beltway, a significant decrease in the number of road traffic collisions was observed, not only on the beltway but also on the arterial roads. Before the installation of the speed cameras, probably as a result of other road safety media campaigns and actions, a decreasing general trend in the number of accidents had been evident in the city since the year 2000. This trend was taken into account in the time-series analysis.

Mean annual numbers of collisions occurring on the beltway during the preintervention and postintervention periods were 638 and 486, respectively; the corresponding means for the arterial roads were 2567 and 2394. Mean numbers of people injured yearly on the beltway in the preintervention and postintervention periods were 946 and 696, respectively; means for the arterial roads were 3254 and 3074. Finally, mean numbers of vehicles involved in beltway collisions yearly during the preintervention and postintervention periods were 1466 and 1108, respectively, and the corresponding means for the arterial roads were 5068 and 4658.

Ratios of collisions per 100000 vehicle-kilometers traveled on the beltway increased from 21.1 in 2001 to 21.6 in 2002 and then decreased to 17.2 in 2003 and 13.2 in 2004. On arterial roads, ratios of collisions per 100000 vehicle-kilometers traveled decreased from 100.2 to 93.2, 87.4, and 86.1 during 2001, 2002, 2003, and 2004 respectively.

Figure 1 shows numbers of road collisions according to monthly subseries. Preintervention and postintervention years are depicted for all monthly series. Numbers of beltway road collisions were consistently lower in the postintervention period than in the preintervention period. Trends were inconsistent on arterial roads.

Once we controlled for trends and seasonality by fitting Poisson distribution models, we observed a decrease of 27% in number of collisions and number of people injured and a decrease of 26% in number of vehicles involved in collisions. The RR of a road collision occurring on the beltway in the postintervention (vs preintervention) period was 0.73 (95% confidence interval [CI]=0.63, 0.85). That is, the speed cameras appeared to have a protective effect. The RR for the first year of intervention was 0.71 (95% CI=0.62, 0.82), and the effect increased in the second year (RR=0.61; 95% CI=0.5, 0.75). As can be seen in Table 1, RRs of injury were similar, as were values for vehicles involved in collisions. By contrast, on arterial roads, RRs of all 3 outcomes assessed were close to 1, indicating that there were no differences between the 2 periods.

The protective effect of speed cameras in reducing collisions was evident during both daytime and nighttime hours and on both weekdays and weekends; effects were greater in magnitude on weekends. RRs were similar for daytime (RR = 0.73; 95% CI = 0.61, 0.88) and nighttime (RR = 0.72; 95% CI = 0.52, 0.99) hours but were lower on weekends (RR = 0.66; 95% CI = 0.46, 0.94) than on weekdays (RR = 0.75; 95% CI = 0.62, 0.91).

Figure 2 shows monthly observed distributions of number of road collisions and expected distributions had speed cameras not been installed. There was a clear decrease in the number of beltway road collisions during the postintervention period, although the slope of the trend did not change (i.e., was not statistically significant). There was also a steady decrease in the number of arterial road collisions between 2001 and 2005, but again there was no change in the slope. Expected numbers of beltway collisions during the postintervention period were higher than the actual numbers observed, but no differences were apparent for arterial roads.

From the attributable fractions, it can be estimated that 364 collisions were prevented, 507 fewer people were injured, and 789 fewer vehicles were involved in collisions as a result of implementation of speed cameras (Table 2). No significant reductions were observed for arterial roads.

Our results show that speed cameras were effective in reducing the numbers of road collisions—and consequently the numbers of people injured and vehicles involved in collisions—in the context of a general overall decline in the number of collisions occurring in Barcelona. Effects were similar during daytime and nighttime hours and were greater on weekends than on weekdays. To our knowledge, ours is the first study in which time-series analyses with Poisson distribution models have been used to evaluate the effectiveness of speed enforcement cameras and the first in which the number of vehicles involved in collisions has been included as an outcome measure.

The study design we used and the statistical analyses we conducted allowed us to control for major factors that might have affected the results and that usually have an impact in evaluation studies of road safety interventions. Using time-series analyses, we controlled for regression to the mean and long-term effects, in that adjustment was made for seasonality and trends. However, the study period was not particularly long (52 months), and further research is necessary to determine trends over longer periods.

Because of the characteristics and function of the beltway as the quickest route connecting the city with the outlying metropolitan area, we do not believe that there would have been a significant number of vehicles using other routes in an attempt to avoid the speed cameras. In fact, vehicle-kilometers traveled on the beltway increased over the period under study (from 2 705 260 in 2002 to 2786 417 in 2003 and 2 788 089 in 2004).

Limitations

Comparison groups can allow for controlling of changes in traffic volume and general changes; however, in our study, the intervention group and the comparison group were not equivalent, as a result of engineering characteristics and because mobile speed cameras operated randomly on the arterial roads. The engineering features of the beltway and arterial roads are clearly different (the latter having intersections and involving lower mean speeds), but the function of the roads is the same: to connect the center city with the outlying metropolitan area. Furthermore, the belt-way and arterial roads account for similar proportions of vehicle-kilometers traveled. Because of our time-series study design, we did not need a strict comparison group, but the use of this group bolstered our findings.

Other road safety interventions that we did not control for, such as screening for drinking and driving or general safety campaigns, could also have influenced the trends observed here, although they would have affected both the beltway and the arterial roads. There also could have been effects associated with motorists driving at reduced speeds on other city streets. Such factors would have tended to produce reductions in collisions, leading to underestimates of the effects produced by the introduction of speed cameras.

Speed Camera Effectiveness

It is difficult to compare our results with those of other published studies, because of the different methodologies used. It must be borne in mind that our study focused on the beltway of a large city of which the speed limit is 80 km per hour. Few published studies have reported maximum speeds, and many have focused on highways. After we controlled for time-dependent confounders, we estimated a reduction of 27% in the number of collisions and the number of people injured and a reduction of 26% in the number of vehicles involved in collisions. Reported reductions in outcomes across other studies have ranged from 5% to 69% for collisions, 12% to 65% for injuries, and 17% to 71% for deaths in the immediate vicinity of camera sites.4

A more recent systematic review reported preintervention–postintervention reductions in the vicinity of camera sites from 14% to 72% for all collisions, 8% to 46% for collisions involving injuries, and 40% to 45% for collisions resulting in fatalities or serious injuries.9 Hess, in a study conducted in the United Kingdom, estimated a 46% global reduction in the number of accidents occurring in the vicinity of cameras and a 29% reduction in the number of accidents occurring on urban roads.10 Christie et al. reported a 51% reduction of injurious collisions in a study focusing on 101 mobile speed cameras implemented in South Wales.11

Chen et al. reported a reduction of 16% in the number of accidents on a British Columbia highway on which the maximum speed was 80 to 90 km per hour.12 Mountain et al.,3 in a study focusing on speed cameras installed in 48 km per hour areas of the United Kingdom, reported a 24% reduction in personal injury accidents, of which 19% was attributable to the effects of the cameras. Overall, these comparisons show that our results fall into the range of those from previously reported studies.

We explored differences in the effectiveness of speed cameras during daytime versus nighttime hours and on weekdays versus weekends because on some stretches of the beltway during certain hours, especially on weekdays, traffic jams make it impossible to drive at speeds higher than those permitted. To take into account times during which there are typically higher traffic flows, we defined “daytime” as 7 am to 8:59 pm. Our aim was not to assess effects of daylight or darkness but, rather, to examine commuter traffic flows on workdays. In Spain, most people start work at 8 am, and shops primarily close around 8:30 pm.

As expected, speed cameras had a significant effect in terms of reducing road injuries on weekends, although confidence intervals overlapped with weekday results. Contrary to expectations, however, the protective effects observed were similar for daytime and nighttime hours. To our knowledge, only 1 study has provided RRs of injurious accidents according to time of day,10 and that study showed daytime and nighttime RRs of 0.46 and 0.55, respectively. Although in that investigation daytime and nighttime referred to daylight and lack of daylight, the results were similar to those revealed in our study. We were unable to find any studies reporting weekday–weekend comparisons of injury outcome measures.

Although our study was set in a very localized area, the characteristics of the roads studied (i.e., an urban beltway and arterial roads) are probably similar to those of many other cities, not only in Spain but elsewhere. Thus, we believe that the results described here would be generalizable to other cities.

Our findings show that speed cameras are effective in reducing the numbers of road collisions and consequently the numbers of people injured and numbers of vehicles involved in collisions in an urban setting. Installation of speed cameras in urban areas should be considered in efforts to increase road safety.

Table
TABLE 1— Relative Risks (RRs) for Outcome Variables Relative Risks During the Postintervention Period Relative to the Preintervention Period: Barcelona, Spain, January 2001–March 2005
TABLE 1— Relative Risks (RRs) for Outcome Variables Relative Risks During the Postintervention Period Relative to the Preintervention Period: Barcelona, Spain, January 2001–March 2005
 CollisionsInjuriesVehicles Involved in Crashes
Time and Postintervention SubperiodBeltway, RR (95% CI)Arterial Roads, RR (95% CI)Beltway, RR (95% CI)Arterial Roads, RR (95% CI)Beltway, RR (95% CI)Arterial Roads, RR (95% CI)
Overall
    April 2003–March 20040.71* (0.62, 0.82)1.00 (0.90, 1.11)0.72* (0.60, 0.87)1.00 (0.89, 1.13)0.73* (0.63, 0.84)0.99 (0.89, 1.10)
    April 2004–March 20050.61* (0.50, 0.75)1.02 (0.88, 1.19)0.62* (0.48, 0.81)0.99 (0.83, 1.17)0.66* (0.54, 0.81)1.04 (0.89, 1.20)
Daytimea
    April 2003–March 20040.72* (0.60, 0.87)1.00 (0.89, 1.12)0.70* (0.56, 0.88)1.01 (0.89, 1.15)0.73* (0.62, 0.87)1.00 (0.90, 1.12)
    April 2004–March 20050.66* (0.51, 0.86)1.06 (0.90, 1.24)0.62* (0.45, 0.87)1.01 (0.84, 1.21)0.69* (0.54, 0.89)1.08 (0.93, 1.27)
Nighttimeb
    April 2003–March 20040.71* (0.52, 0.96)0.99 (0.85, 1.16)0.79 (0.54, 1.15)1.02 (0.85, 1.23)0.70* (0.50, 0.99)0.96 (0.82, 1.14)
    April 2004–March 20050.50* (0.33, 0.78)0.93 (0.74, 1.17)0.57* (0.33, 0.98)0.96 (0.74, 1.24)0.53* (0.33, 0.87)0.88 (0.70, 1.12)
Weekdays
    April 2003–March 20040.73* (0.61, 0.89)0.99 (0.88, 1.11)0.79* (0.66, 0.96)0.98 (0.86, 1.13)0.76* (0.64, 0.91)0.99 (0.87, 1.11)
    April 2004–March 20050.65* (0.49, 0.85)1.05 (0.89, 1.23)0.70* (0.53, 0.92)1.01 (0.83, 1.23)0.69* (0.54, 0.89)1.06 (0.90, 1.26)
Weekends
    April 2003–March 20040.65* (0.46, 0.93)1.00 (0.83, 1.21)0.57* (0.35, 0.93)1.05 (0.86, 1.29)0.63* (0.43, 0.93)1.00 (0.82, 1.21)
    April 2004–March 20050.50* (0.31, 0.83)0.92 (0.70, 1.20)0.46* (0.23, 0.90)0.91 (0.68, 1.23)0.56* (0.33, 0.97)0.93 (0.70, 1.24)

Note. CI = confidence interval. Values were derived from time-series analyses using Poisson regression models in which we controlled for trends and seasonality.

aDaytime was defined as 7 am to 8:59 pm.

bNighttime was defined as 9 pm to 6:59 am.

*P < .05.

Table
TABLE 2— Relative Risks, (RR) Attributable Fractions, and Observed, Expected, and Prevented Numbers of Collisions, Injuries, and Vehicles Involved in Collisions After 24 Months of Speed Camera Operation: Barcelona, Spain, January 2001–March 2005
TABLE 2— Relative Risks, (RR) Attributable Fractions, and Observed, Expected, and Prevented Numbers of Collisions, Injuries, and Vehicles Involved in Collisions After 24 Months of Speed Camera Operation: Barcelona, Spain, January 2001–March 2005
 RR (95% CI)Attributable Fraction (95% CI)No. ObservedNo. ExpectedNo. Preventeda
Beltway
    Collisions0.73* (0.63, 0.85)−0.37 (−0.59, −0.18)9711335−364
    Injuries0.73* (0.61, 0.89)−0.37 (−0.64, −0.12)13911898−507
    Vehicles involved in collisions0.74* (0.64, 0.85)−0.35 (−0.56, −0.18)22153004−789
Arterial roads
    Collisions0.99 (0.90, 1.10)−0.01 (−0.11, 0.09)47884821−33
    Injuries1.10 (0.90, 1.14)0.09 (−0.11, 0.12)6147609453
    Vehicles involved in collisions0.99 (0.89, 1.09)−0.01 (−0.12, 0.08)93159451−136

Note. CI = confidence interval. Values were derived from time-series analyses using Poisson regression models in which we controlled for trends and seasonality.

aObserved minus expected.

* P < .05.

The work described in this article was partially supported by the Red Española de Centros de Epidemiologia y Salud Pública (grant C03/09).

We thank Mercé Navarro, Diego Navarro, Angel López, Joan Mañosa, and Alex Culubret, as well as Manuel Haro, and Josep Royuela (Guàrdia Urbana de Barcelona) for providing the data.

Human Participant Protection No protocol approval was needed for this study.

References

1. Peden M, Scurfield R, Sleet D, et al., eds. World Report on Road Traffic Injury Prevention. Geneva, Switzerland: World Health Organization; 2004. Google Scholar
2. Hirst WM, Mountain LJ, Maher MJ. Are speed enforcement cameras more effective than other speed management measures? An evaluation of the relationship between speed and accident reductions. Accid Anal Prev. 2005;37:731–741. Crossref, MedlineGoogle Scholar
3. Mountain LJ, Hirst WM, Maher MJ. Are speed enforcement cameras more effective than other speed management measures? The impact of speed management schemes on 30 mph roads. Accid Anal Prev. 2005;37:742–754. Crossref, MedlineGoogle Scholar
4. Pilkington P, Kinra S. Effectiveness of speed cameras in preventing traffic collisions and related casualties: systematic review. BMJ. 2005;330:331–334. Crossref, MedlineGoogle Scholar
5. Hauer E. Observational Before-After Studies in Road Safety: Estimating the Effect of Highway and Traffic Engineering Measures on Road Safety. Oxford, England: Pergamon Press; 2005. Google Scholar
6. Elvik R. The importance of confounding in observational before-and-after studies of road safety measures. In: Proceedings of the 13th International Cooperation on Theories and Concepts in Traffic Safety Workshop. Corfu, Greece: International Cooperation on Theories and Concepts in Traffic Safety; 2000:105–117. Google Scholar
7. Stolwijk AM, Straatman H, Zielhuis GA. Studying seasonality by using sine and cosine functions in regression analysis. J Epidemiol Community Health. 1999; 53:235–238. Crossref, MedlineGoogle Scholar
8. Stata, Version 9 [computer program]. College Station, Tex: StataCorp; 2005. Google Scholar
9. Wilson C, Willis C, Hendrikz JK, Bellamy N. Speed enforcement detection devices for preventing road traffic injuries (Cochrane Review). In: The Cochrane Library, Issue 2. Oxford, England: Update Software; 2006. Google Scholar
10. Hess S. An analysis of the effects of speed limit enforcement cameras with differentiation by road type and catchment area. Available at: http://www.cts.cv.ic.ac.uk/StaffPages/StephaneHess/papers/Hess_slec_2004.pdf. Accessed May 18, 2007. Google Scholar
11. Christie SM, Lyons RA, Dunstan FD, Jones SJ. Are mobile speed cameras effective? A controlled before and after study. Inj Prev. 2003;9:302–306. Crossref, MedlineGoogle Scholar
12. Chen G, Meckle W, Wilson J. Speed and safety effect of photo radar enforcement on a highway corridor in British Columbia. Accid Anal Prev. 2002;34: 129–138. Crossref, MedlineGoogle Scholar

Related

No related items

TOOLS

SHARE

ARTICLE CITATION

Katherine Pérez, PhD, Marc Marí-Dell’Olmo, MPH, Aurelio Tobias, PhD, and Carme Borrell, PhDKatherine Pérez and Marc Marí-Dell’Olmo are with the Agència de Salut Pública de Barcelona, CIBER Epidemiología y Salud Pública, Barcelona, Spain. Aurelio Tobias is with the Escuela Nacional de Sanidad, Instituto de Salud Carlos III, Madrid, Spain. Carme Borrell is with the Agència de Salut Pública de Barcelona, CIBER Epidemiología y Salud Pública, Barcelona, and the Departament de Ciències Experimentals i de la Salut, Universitat Pompeu Fabra, Barcelona. “Reducing Road Traffic Injuries: Effectiveness of Speed Cameras in an Urban Setting”, American Journal of Public Health 97, no. 9 (September 1, 2007): pp. 1632-1637.

https://doi.org/10.2105/AJPH.2006.093195

PMID: 17666698