Driver injury severity outcome analysis in rural interstate highway crashes: a two-level Bayesian logistic regression interpretation

There is a high potential of severe injury outcomes in traffic crashes on rural interstate highways due to the significant amount of high speed traffic on these corridors. Hierarchical Bayesian models are capable of incorporating between-crash variance and within-crash correlations into traffic crash data analysis and are increasingly utilized in traffic crash severity analysis. This paper applies a hierarchical Bayesian logistic model to examine the significant factors at crash and vehicle/driver levels and their heterogeneous impacts on driver injury severity in rural interstate highway crashes. Analysis results indicate that the majority of the total variance is induced by the between-crash variance, showing the appropriateness of the utilized hierarchical modeling approach. Three crash-level variables and six vehicle/driver-level variables are found significant in predicting driver injury severities: road curve, maximum vehicle damage in a crash, number of vehicles in a crash, wet road surface, vehicle type, driver age, driver gender, driver seatbelt use and driver alcohol or drug involvement. Among these variables, road curve, functional and disabled vehicle damage in crash, single-vehicle crashes, female drivers, senior drivers, motorcycles and driver alcohol or drug involvement tend to increase the odds of drivers being incapably injured or killed in rural interstate crashes, while wet road surface, male drivers and driver seatbelt use are more likely to decrease the probability of severe driver injuries. The developed methodology and estimation results provide insightful understanding of the internal mechanism of rural interstate crashes and beneficial references for developing effective countermeasures for rural interstate crash prevention.     

Involvement rate in two-car crashes versus driver age and car mass of each involved car

This work was performed to investigate how the likelihood of a two-car crash depends on the driver age and car mass for each of the two involved cars, and also to examine the special case of cars of similar mass crashing into each other. Data on 108 044 cars involved in police reported two-car crashes occurring in New York State in 1971 and 1972 were fitted to a function of the driver age and car mass for each of the two involved cars. Car registrations are used to estimate exposure. The special case of crashes between cars of similar mass is considered because of prior results on driver injuries in such crashes. It is found that “small-small” crashes (defined as a 900 kg car crashing into another 900 kg car) are about 0.3 times as likely as “big-big” crashes (an 1800 kg car crashing into another 1800 kg car), assuming equal numbers of cars driven by drivers of the same age. Combining the present results with earlier findings of increased injury risk in small-small crashes gives that such crashes injure about 70% as many drivers as big-big crashes when normalized for numbers of cars and driver age. That is. it is concluded that small-small crashes produce 30% fewer injuries than do big-big crashes.     

A population based estimation of the driver protection provided by passenger cars: France 1996–2005

The technology used in cars to protect occupants is constantly developing. Demonstrating the beneficial effects in the field is complex as the most recent vehicles are generally used by drivers who differ from other drivers and who drive in different traffic conditions. This paper presents an overall estimation of the consequences of changes in the secondary safety of cars, taking account of most of these factors.The data come from information collected about injury road traffic crashes by the police in France between 1996 and 2005. The risk of the driver being killed has been evaluated for the 144,034 drivers involved in two-car crashes and for the 63,621 drivers involved in single-car crashes.The study shows that when a recent car is in collision with an older car the driver of the former is better protected than the driver of the latter. These improvements in secondary safety are not observed in the case of single-car crashes, very probably because of higher impact speeds. Our findings also confirm the need for protection systems to be better adapted to the specific characteristics of users and for an improvement in the crash compatibility of vehicles, in particular to overcome the consequences of differences between the masses of vehicles.     

Evaluation of electronic stability control effectiveness in Australasia

Electronic stability control (ESC) is an in-vehicle technology aimed at improving primary safety by assisting the driver in avoiding loss of control of the vehicle. The aim of this study was to use available crash data from Australia and New Zealand to evaluate the effectiveness of ESC in reducing crash risk and to establish whether benefits estimated from overseas studies have translated to the Australian and New Zealand environments. The sample analysed included 7699 crashed vehicles fitted with ESC which comprised of 90 different models. Poisson regression was used to test whether the differences in the observed and expected crash counts for ESC fitted vehicles were significant, with exposure being induced from counts of rear end impacts. It was found that ESC reduced the risk of single vehicle crashes in which the driver was injured by 68% for 4WDs compared with 27% for passenger cars. The effect of ESC on multiple vehicle crashes in Australia and New Zealand was not clear. The long-term benefits of fitting ESC to all vehicles in Australia were also investigated based on the estimated single vehicle crash reductions.     

Evolution of the crashworthiness and aggressivity of the Spanish car fleet

This paper investigates the relationship between a passenger car's year of registration and its crashworthiness and aggressivity in real-world crashes. Crashworthiness is defined as the ability of a car to protect its own occupants, and has been evaluated in single and two-car crashes. Aggressivity is defined as the ability to protect users travelling in other vehicles, and has been evaluated only in two-car crashes. The dependent variable is defined as the proportion of injured drivers who are killed or seriously injured; following previous research, we refer to this magnitude as injury severity. A decrease in the injury severity of a driver is interpreted as an improvement in the crashworthiness of their car; similarly, a decrease in the injury severity of the opponent driver is regarded as an improvement in aggressivity.Data have been extracted from the Spanish Road Accident Database, which contains information on every accident registered by the police in which at least one person was injured. Two types of regression models have been used: logistic regression models in single-car crashes, and generalised estimating equations (GEE) models in two-car crashes. GEE allow to take account of the correlation between the injury severities of drivers involved in the same crash. The independent variables considered have been: year of registration of the subject car (crashworthiness component), year of registration of the opponent car (aggressivity component), and several factors related to road, driver and environment.Our models confirm that crashworthiness has largely improved in two-car crashes: when crashing into the average opponent car, drivers of cars registered before 1985 have a significantly higher probability of being killed or seriously injured than drivers of cars registered in 2000–2005 (odds ratio: 1.80; 95% confidence interval: 1.61; 2.01). In single-car crashes, the improvement in crashworthiness is very slight (odds ratio: 1.04; 95% confidence interval: 0.93; 1.16). On the other hand, we have also found a significant worsening in aggressivity in two-car crashes: the driver of the average car has a significantly lower probability of being killed or seriously injured when crashing into a car registered before 1985, than when crashing into a car registered in 2000–2005 (odds ratio: 0.52; 95% confidence interval: 0.45; 0.60).Our results are consistent with a large amount of previous research that has reported significant improvements in the protection of car occupants. They also add to some recent studies that have found a worsening in the aggressivity of modern cars. This trend may be reflecting the impact of differences in masses and travel speeds, as well as the influence of consumer choices. The precise reasons have to be investigated. Also, the causes have to be found for so large a discrepancy between crashworthiness in single and two-car crashes.     

Hierarchical Bayesian random intercept model-based cross-level interaction decomposition for truck driver injury severity investigations

Traffic crashes occurring on rural roadways induce more severe injuries and fatalities than those in urban areas, especially when there are trucks involved. Truck drivers are found to suffer higher potential of crash injuries compared with other occupational labors. Besides, unobserved heterogeneity in crash data analysis is a critical issue that needs to be carefully addressed. In this study, a hierarchical Bayesian random intercept model decomposing cross-level interaction effects as unobserved heterogeneity is developed to examine the posterior probabilities of truck driver injuries in rural truck-involved crashes. The interaction effects contributing to truck driver injury outcomes are investigated based on two-year rural truck-involved crashes in New Mexico from 2010 to 2011. The analysis results indicate that the cross-level interaction effects play an important role in predicting truck driver injury severities, and the proposed model produces comparable performance with the traditional random intercept model and the mixed logit model even after penalization by high model complexity. It is revealed that factors including road grade, number of vehicles involved in a crash, maximum vehicle damage in a crash, vehicle actions, driver age, seatbelt use, and driver under alcohol or drug influence, as well as a portion of their cross-level interaction effects with other variables are significantly associated with truck driver incapacitating injuries and fatalities. These findings are helpful to understand the respective or joint impacts of these attributes on truck driver injury patterns in rural truck-involved crashes.     

Examining driver injury severity outcomes in rural non-interstate roadway crashes using a hierarchical ordered logit model

Rural non-interstate crashes induce a significant amount of severe injuries and fatalities. Examination of such injury patterns and the associated contributing factors is of practical importance. Taking into account the ordinal nature of injury severity levels and the hierarchical feature of crash data, this study employs a hierarchical ordered logit model to examine the significant factors in predicting driver injury severities in rural non-interstate crashes based on two-year New Mexico crash records. Bayesian inference is utilized in model estimation procedure and 95% Bayesian Credible Interval (BCI) is applied to testing variable significance. An ordinary ordered logit model omitting the between-crash variance effect is evaluated as well for model performance comparison. Results indicate that the model employed in this study outperforms ordinary ordered logit model in model fit and parameter estimation. Variables regarding crash features, environment conditions, and driver and vehicle characteristics are found to have significant influence on the predictions of driver injury severities in rural non-interstate crashes. Factors such as road segments far from intersection, wet road surface condition, collision with animals, heavy vehicle drivers, male drivers and driver seatbelt used tend to induce less severe driver injury outcomes than the factors such as multiple-vehicle crashes, severe vehicle damage in a crash, motorcyclists, females, senior drivers, driver with alcohol or drug impairment, and other major collision types. Research limitations regarding crash data and model assumptions are also discussed. Overall, this research provides reasonable results and insight in developing effective road safety measures for crash injury severity reduction and prevention.     

Do laboratory frontal crash test programs predict driver fatality risk? Evidence from within vehicle line variation in test ratings

A number of studies have examined whether the National Highway Traffic Safety Administration's (NHTSA) frontal crash test results reliably indicate the risk of fatality or injury in serious crashes. The conclusions of these studies are mixed. Generally, studies that examine crashes in the circumstances as close as possible to those of the laboratory test find that crash test results do predict real-world risk, but studies of crashes outside those specific circumstances find either no support for the predictive validity of crash test results or limited support with important inconsistencies. We provide a new test of the predictive validity of the crash test results using information from multiple crash tests within vehicle lines, thus controlling for systematic differences in driver behavior across vehicle lines. Among drivers of passenger cars, we find large, statistically significant differences in fatality risk for vehicles with one- to four-star NHTSA ratings versus a five-star rating. We also examine the Insurance Institute for Highway Safety's frontal offset crash test, though our sample of vehicle lines tested twice or more is considerably smaller than for NHTSA ratings. Our results also support the predictive validity of the frontal offset crash test results for passenger cars, but not for trucks.     

Overall injury risk to different drivers: combining exposure,frequency, and severity models

Traffic crash risk assessments should incorporate appropriate exposure data. However, existing US nationwide crash data sets, the NASS General Estimates System (GES) and the Fatality Analysis Reporting System (FARS), do not contain information on driver or vehicle exposure. In order to obtain appropriate exposure data, this work estimates vehicle miles driven (VMD) by different drivers using the Nationwide Personal Transportation Survey (NPTS). These results are combined with annual crash rates and injury severity information from the GES for a comprehensive assessment of overall risk to different drivers across vehicle classes.Data are distinguished by driver age, gender, vehicle type, crash type (rollover versus non-rollover), and injury severity. After correcting for drivers' crash exposure, results indicate that young drivers are far more crash prone than other drivers (per VMD) and that drivers of sports utility vehicles (SUVs) and pickups (PUs) are more likely to be involved in rollover crashes than those driving passenger cars. Although, the results suggest that drivers of SUVs are generally much less crash prone than drivers of passenger cars, the rollover propensity of SUVs and the severity of that crash type offset many of the incident benefits for SUV drivers.     

A multinomial logit model-Bayesian network hybrid approach for driver injury severity analyses in rear-end crashes

Rear-end crash is one of the most common types of traffic crashes in the U.S. A good understanding of its characteristics and contributing factors is of practical importance. Previously, both multinomial Logit models and Bayesian network methods have been used in crash modeling and analysis, respectively, although each of them has its own application restrictions and limitations. In this study, a hybrid approach is developed to combine multinomial logit models and Bayesian network methods for comprehensively analyzing driver injury severities in rear-end crashes based on state-wide crash data collected in New Mexico from 2010 to 2011. A multinomial logit model is developed to investigate and identify significant contributing factors for rear-end crash driver injury severities classified into three categories: no injury, injury, and fatality. Then, the identified significant factors are utilized to establish a Bayesian network to explicitly formulate statistical associations between injury severity outcomes and explanatory attributes, including driver behavior, demographic features, vehicle factors, geometric and environmental characteristics, etc. The test results demonstrate that the proposed hybrid approach performs reasonably well. The Bayesian network reference analyses indicate that the factors including truck-involvement, inferior lighting conditions, windy weather conditions, the number of vehicles involved, etc. could significantly increase driver injury severities in rear-end crashes. The developed methodology and estimation results provide insights for developing effective countermeasures to reduce rear-end crash injury severities and improve traffic system safety performance.     

Modeling occupant-level injury severity: An application to large-truck crashes

Most of the injury-severity analyses to date have focused primarily on modeling the most-severe injury of any crash, although a substantial fraction of crashes involve multiple vehicles and multiple persons. In this study, we present an extensive exploratory analysis that highlights that the highest injury severity is not necessarily the comprehensive indicator of the overall severity of any crash. Subsequently, we present a panel, hetroskedastic ordered-probit model to simultaneously analyze the injury severities of all persons involved in a crash. The models are estimated in the context of large-truck crashes. The results indicate strong effects of person-, driver-, vehicle-, and crash-characteristics on the injury severities of persons involved in large-truck crashes. For example, several driver behavior characteristics (such as use of illegal drugs, DUI, and inattention) were found to be statistically significant predictors of injury severity. The availability of airbags and the use of seat-belts are also found to be associated with less-severe injuries to car-drivers and car-passengers in the event of crashes with large trucks. Car drivers’ familiarity with the vehicle and the roadway are also important for both the car drivers and passengers. Finally, the models also indicate the strong presence of intra-vehicle correlations (effect of common vehicle-specific unobserved factors) among the injury propensities of all persons within a vehicle.     

An explanatory analysis of driver injury severity in rear-end crashes using a decision table/Naïve Bayes (DTNB) hybrid classifier

Rear-end crashes are a major type of traffic crashes in the U.S. Of practical necessity is a comprehensive examination of its mechanism that results in injuries and fatalities. Decision table (DT) and Naïve Bayes (NB) methods have both been used widely but separately for solving classification problems in multiple areas except for traffic safety research. Based on a two-year rear-end crash dataset, this paper applies a decision table/Naïve Bayes (DTNB) hybrid classifier to select the deterministic attributes and predict driver injury outcomes in rear-end crashes. The test results show that the hybrid classifier performs reasonably well, which was indicated by several performance evaluation measurements, such as accuracy, F-measure, ROC, and AUC. Fifteen significant attributes were found to be significant in predicting driver injury severities, including weather, lighting conditions, road geometry characteristics, driver behavior information, etc. The extracted decision rules demonstrate that heavy vehicle involvement, a comfortable traffic environment, inferior lighting conditions, two-lane rural roadways, vehicle disabled damage, and two-vehicle crashes would increase the likelihood of drivers sustaining fatal injuries. The research limitations on data size, data structure, and result presentation are also summarized. The applied methodology and estimation results provide insights for developing effective countermeasures to alleviate rear-end crash injury severities and improve traffic system safety performance.     

Investigating driver injury severity patterns in rollover crashes using support vector machine models

Rollover crash is one of the major types of traffic crashes that induce fatal injuries. It is important to investigate the factors that affect rollover crashes and their influence on driver injury severity outcomes. This study employs support vector machine (SVM) models to investigate driver injury severity patterns in rollover crashes based on two-year crash data gathered in New Mexico. The impacts of various explanatory variables are examined in terms of crash and environmental information, vehicle features, and driver demographics and behavior characteristics. A classification and regression tree (CART) model is utilized to identify significant variables and SVM models with polynomial and Gaussian radius basis function (RBF) kernels are used for model performance evaluation. It is shown that the SVM models produce reasonable prediction performance and the polynomial kernel outperforms the Gaussian RBF kernel. Variable impact analysis reveals that factors including comfortable driving environment conditions, driver alcohol or drug involvement, seatbelt use, number of travel lanes, driver demographic features, maximum vehicle damages in crashes, crash time, and crash location are significantly associated with driver incapacitating injuries and fatalities. These findings provide insights for better understanding rollover crash causes and the impacts of various explanatory factors on driver injury severity patterns.     

Car occupant safety in frontal crashes: a parameter study of vehicle mass, impact speed, and inherent vehicle protection

A new mathematical model was developed to estimate average injury and fatality rates in frontal car-to-car crashes for changes in vehicle fleet mass, impact speed distribution, and inherent vehicle protection. The estimates were calculated from injury/fatality risk data, delta-V distribution and collision probability of two vehicles, where delta-V depends on impact speed and mass of the colliding vehicles. The impact speed distribution was assumed to be unaffected by a change in fleet mass distribution.

The results showed that safety in frontal crashes would improve 27–35% by a 10% increase in fatality risk parameters, which reflected substantial improvement in inherent vehicle protection. A 40% safety improvement was attained by a 10% impact speed reduction. Consequences of vehicle fleet mass were not as strong, but depended on the average mass ratio of the fleet. A reduction in mass range would be the most beneficial, while a uniform mass reduction of 20% would increase the fatality rate by 5.4%. The model estimates trends in traffic safety and may help to identify priorities in active and passive safety.     

Analysis of large truck crash severity using heteroskedastic ordered probit models

Long-combination vehicles (LCVs) have significant potential to increase economic productivity for shippers and carriers by decreasing the number of truck trips, thus reducing costs. However, size and weight regulations, triggered by safety concerns and, in some cases, infrastructure investment concerns, have prevented large-scale adoption of such vehicles. Information on actual crash performance is needed. To this end, this work uses standard and heteroskedastic ordered probit models, along with the United States’ Large Truck Crash Causation Study, General Estimates System, and Vehicle Inventory and Use Survey data sets, to study the impact of vehicle, occupant, driver, and environmental characteristics on injury outcomes for those involved in crashes with heavy-duty trucks. Results suggest that the likelihood of fatalities and severe injury is estimated to rise with the number of trailers, but fall with the truck length and gross vehicle weight rating (GVWR). While findings suggest that fatality likelihood for two-trailer LCVs is higher than that of single-trailer non-LCVs and other trucks, controlling for exposure risk suggest that total crash costs of LCVs are lower (per vehicle-mile traveled) than those of other trucks.     

A logistic model of the effects of roadway, environmental, vehicle, crash and driver characteristics on hit-and-run crashes

Leaving the scene of a crash without reporting it is an offence in most countries and many studies have been devoted to improving ways to identify hit-and-run vehicles and the drivers involved. However, relatively few studies have been conducted on identifying factors that contribute to the decision to run after the crash. This study identifies the factors that are associated with the likelihood of hit-and-run crashes including driver characteristics, vehicle types, crash characteristics, roadway features and environmental characteristics. Using a logistic regression model to delineate hit-and-run crashes from nonhit-and-run crashes, this study found that drivers were more likely to run when crashes occurred at night, on a bridge and flyover, bend, straight road and near shop houses; involved two vehicles, two-wheel vehicles and vehicles from neighboring countries; and when the driver was a male, minority, and aged between 45 and 69. On the other hand, collisions involving right turn and U-turn maneuvers, and occurring on undivided roads were less likely to be hit-and-run crashes.     

Antilock brakes and the risk of driver injury in a crash: a case-control study

While antilock brakes can improve steering and reduce stopping distance in some test situations, there is little evidence that they reduce the risk of crash-related injury. We sought to estimate the association between presence of antilock brakes and the risk of driver injury. We conducted a case-control study using claims data from the Insurance Corporation of British Columbia, Canada, for passenger vehicles insured during July 1, 2003, to June 30, 2004. Cases were 5000 vehicles with a driver crash injury during the study period. Controls were 49,994 vehicles insured at the mid-point of the study interval. The adjusted risk ratio for a crash with driver injury in a vehicle with antilock brakes was 1.06 (95% confidence interval, 0.95-1.17), compared with a vehicle without antilock brakes. If this estimated association is causal, antilock brakes do not prevent crash-related driver injuries.     

Quasi-induced exposure method: Evaluation of not-at-fault assumption

Crash rates are used to establish the relative safety of various variables of concern such as driver classes, vehicle types and roadway components. Appropriate exposure data for estimating crash rates is critical but crash databases do not contain information on driver or vehicle exposure. The quasi-induced exposure method, which uses not-at-fault driver/vehicle data as an exposure metric, is a technique used in order to overcome this problem. The basic assumption made here is that not-at-fault drivers represent the total population in question. This paper examines the validity of this assumption using the Kentucky crash database to define two samples of not-at-fault drivers. One sample included only not-at-fault drivers selected from the first two vehicles in a multi-vehicle crash (two or more vehicles involved) while the other included the not-at-fault drivers from multi-vehicle crashes with more than two vehicles involved and excluding the first two drivers. The assumption is that the randomness of the involvement of drivers in the second sample is more reasonable than the drivers in the first two vehicles involved in crashes. The results indicate that these two samples are similar; there is no statistical evidence demonstrating that both samples represent two different populations in the maneuvers and other variables/factors examined here; and they are representative simple random samples of the driver population with respect to the distribution of the driver age when there is no reasonable doubt about investigating officers’ judgments. Thus, estimating relative crash propensities for any given driver type by using the quasi-induced exposure approach will yield reasonable estimates of exposure.     

A mixed logit analysis of two-vehicle crash severities involving a motorcycle

Using motorcycle crash data for Iowa from 2001 to 2008, this paper estimates a mixed logit model to investigate the factors that affect crash severity outcomes in a collision between a motorcycle and another vehicle. These include crash-specific factors (such as manner of collision, motorcycle rider and non-motorcycle driver and vehicle actions), roadway and environmental conditions, location and time, motorcycle rider and non-motorcycle driver and vehicle attributes. The methodological approach allows the parameters to vary across observations as opposed to a single parameter representing all observations. Our results showed non-uniform effects of rear-end collisions on minor injury crashes, as well as of the roadway speed limit greater or equal to 55 mph, the type of area (urban), the riding season (summer) and motorcyclist's gender on low severity crashes. We also found significant effects of the roadway surface condition, clear vision (not obscured by moving vehicles, trees, buildings, or other), light conditions, speed limit, and helmet use on severe injury outcomes.     

Incidence and characteristics of school bus crashes and injuries

Objective

Studies of school bus crashes have focused on the biomechanics of catastrophic collisions, with very few examining crash incidence.

Methods

Crashes in the state of Iowa were examined from January 2002 through December 2005. School bus crashes were identified through the Iowa Crash Data, a comprehensive database of all reported crashes in the State of Iowa. School bus mileage data were provided by the Iowa Department of Education. School bus crash, fatality, and injury rates were calculated and differences in crash and injury characteristics between school buses and other vehicles were examined.

Results

The school bus crash, fatality and non-fatal injury rates were 320.7, 0.4 and 13.6 per 100 million bus miles travelled, respectively. School bus crash fatality and injury rates were 3.5 and 5.4 times lower than overall all vehicle crash fatality and injury rates, respectively. Drivers of other vehicles were more likely to have caused the crash than the bus driver (P < 0.001).

Conclusions

School buses experience low crash rates, and the majority of crashes do not lead to injury. Buses are among the safest forms of road transportation, and efforts to educate drivers of other vehicles may help reduce crashes with buses.