Estimating seat belt effectiveness using matched-pair cohort methods
ABSTRACT Using US data for 1986-1998 fatal crashes, we employed matched-pair analysis methods to estimate that the relative risk of death among belted compared with unbelted occupants was 0.39 (95% confidence interval (CI) 0.37-0.41). This differs from relative risk estimates of about 0.55 in studies that used crash data collected prior to 1986. Using 1975-1998 data, we examined and rejected three theories that might explain the difference between our estimate and older estimates: (1) differences in the analysis methods; (2) changes related to car model year; (3) changes in crash characteristics over time. A fourth theory, that the introduction of seat belt laws would induce some survivors to claim belt use when they were not restrained, could explain part of the difference in our estimate and older estimates; but even in states without seat belt laws, from 1986 through 1998, the relative risk estimate was 0.45 (95% CI 0.39-0.52). All of the difference between our estimate and older estimates could be explained by some misclassification of seat belt use. Relative risk estimates would move away from 1, toward their true value, if misclassification of both the belted and unbelted decreased over time, or if the degree of misclassification remained constant, as the prevalence of belt use increased. We conclude that estimates of seat belt effects based upon data prior to 1986 may be biased toward 1 by misclassification.
- SourceAvailable from: ncbi.nlm.nih.gov
[Show abstract] [Hide abstract]
- "This method normalizes crashes by dividing the number of fatalities in a specific group of occupants to the number of fatalities in a control group. It has been used to estimate the effectiveness of restraint systems and seating positions by various researchers (Dalmotas and Krzyzewski 1987, Evans 1987, 1988, Cummings et al. 2003). "
ABSTRACT: Effectiveness of the rear seat in protecting occupants of different age groups in frontal crashes for 2000-2009 model years (MY) of vehicles was estimated and compared to 1990-1999 model years of vehicles. The objective was to determine the effectiveness of the rear seat compared to the front seat for various age groups in newer model year vehicles. The double paired comparison method was used to estimate relative effectiveness. For belted adults of the 25-49 age group, the fatality reduction effectiveness of the rear seat compared to the right front seat was 25 % (CI 11% to 36%), in the 1990-1999 model year vehicles. The relative effectiveness was -31% (CI -63% to -5%) for the same population, in the 2000-2009 model year vehicles. For restrained children 0-8 years old, the relative effectiveness was 55% (CI 48% to 61%) when the vehicles were of the 1990-1999 period. The level of effectiveness for this age group was reduced to 25% (CI -4% to 46%) in the 2000-2009 MYs of vehicles. Results for other age groups of belted occupants have followed a similar trend. All belted adult occupants of 25+ years old were significantly less protected in rear seats as compared to right front seats in the 2000-2009 model years of vehicles. For unbelted occupants however, rear seats were still a safer position than front seats, even in the 2000-2009 model years of vehicles.Annals of advances in automotive medicine 01/2010; 54:149-58.
[Show abstract] [Hide abstract]
- "Many studies have assessed the effect of individual variables, such as age, sex, safety belt use and seating position in the vehicle on the risk of death or severe injury to the driver or passengers of vehicles involved in road crashes (Bédard et al., 2002; Braver et al., 1998; Cummings et al., 2003b; Evans, 2001a,b; Huelke and Compton, 1995; Li et al., 2003; O'Donnell and Connor, 1996; Smith and Cummings, 2004; Yau, 2004; Zhang et al., 2000). To control for the confounding effect of crash severity (for example, the protective effect of the driver's safety belt may be overestimated if unbelted drivers tend to be involved in more severe crashes), some researchers have applied matched-byvehicle analyses, either with the double pair comparison method proposed by Evans (1986) or with the more efficient regression matched-pair analysis methods (Cummings et al., 2003a). "
ABSTRACT: We studied the effect of age, gender, use of restraint systems and seat position on the risk of death for rear-seated passengers of cars involved in road crashes. The data source was the Spanish register of traffic crashes with victims compiled by the Government's General Traffic Directorate. Data for crashes recorded from 1993 to 2002, inclusive, were studied. We used a matched cohort design to analyze all 5260 rear-seated passengers in vehicles occupied by two or three rear-seated passengers for accidents in which at least one of these passengers was killed. Conditional Poisson regression with death as the dependent variable was used. An increased risk of death was observed for females and children aged <3 years. For passengers aged 25 years and older, the risk increased with age. The use of restraint systems and central and right-side seats was associated with a lower risk. These results should be considered in research focused on passenger fragility and strategies to prevent injury and death.Accident Analysis & Prevention 06/2006; 38(3):563-6. DOI:10.1016/j.aap.2005.11.014 · 1.87 Impact Factor
[Show abstract] [Hide abstract]
- "A matched case-control design study in the United States (Crandall et al., 2001) reported that the use of seat belts lowered mortality (OR = 0.25, 95% CI 0.22–029). In a study using U.S. data for 1986–1988 fatal crashes (Cummings et al., 2003), the relative risk of death among belted compared with unbelted occupants was 0.39 (95% CI 0.37–0.41). Most other studies (Petridou, 1998; Cummings, 2002) reported similar findings. "
ABSTRACT: Objective. The present study was performed to clarify the relation between alcohol use and traffic fatalities in accidents involving motor vehicles in Japan. Methods. Data on traffic accidents were collected from Fukuoka Prefectural Police records of traffic accidents which occurred in that prefecture between 1987 and 1996. Multiple logistic regression models were used to assess the effect of alcohol use on the risk of traffic-accident death. Results. The data showed that 58,421 male drivers were involved in traffic accidents during the 10-year study period, and that 271 of these were killed as a result of the accident. Alcohol use was significantly associated with speed, seat belt use, time, and road form. Among male motorcar drivers, the odds ratio of alcohol use before driving, after adjusting for age, calendar year, time, and road form, was 4.08 (95% confidence interval, 3.08-5.40), which means that about 75% of fatalities (attributable risk percent among exposed) might have been prevented if drivers had not drunk before driving. Conclusions. Alcohol use before driving resulted in a 4.08-fold increase in the risk of death in a traffic accident. It is suggested that alcohol use is considered an important risk factor for fatality in traffic accidents.