National Automotive Sampling System-Crashworthiness Data Research Articles

Sex-based differences in odds of motor vehicle crash injury outcomes

ObjectiveSeveral studies have documented the relative risk or odds of injury and fatality for females versus males in motor vehicle crashes (Parenteau et al. 2013, Forman et al. 2019, Brumbelow and Jermakian, 2022; Noh et al. 2022). Though, none combined National Automotive Sampling System–Crashworthiness Data System (NASS-CDS) and Crash Investigation Sampling System (CISS). The aim of this study was to document the relative odds of various injury outcomes for females versus males while considering a broad range of crash types, pre-crash and crash variables, and occupant characteristics. MethodsMultivariable logistic regression was carried out to study the odds of injury for females versus males. A select imputation method (Hot Deck, Approximate Bayesian Bootstrap) was applied as part of efforts to create multivariable logistic regression models for 25 different injury outcomes associated with occupants (age 13 years and older) involved in passenger vehicle crashes published in NASS-CDS (2000 to 2015) and CISS (2017–2021). Both pre-crash (n=7) and crashworthiness (n=22) predictor variables were considered, but only significant variables at p≤0.10 level were retained in final models. Six crash-type models were produced for each injury outcome; one that included all crashes, one for each of four different planar crash types (frontal, near-side, far-side, rear), and one for crashes involving rollover. These six sets of crash-type models were expanded further to include a model version that included both pre-crash/environment and crashworthiness predictor variables and one model limited to crashworthiness predictors only. Different than other recent studies, all crash types, occupant restraint conditions, and seating positions were considered. Occupant sex was retained in all models to facilitate female versus male injury outcome odds ratio (OR) assessments. ResultsFemale versus male injury OR estimates for 300 unique models are presented. Females had significantly higher odds of injury than males in 36 models (OR>1.0, p-value ≤0.05). This contrasts with 43 models where females had significantly lower odds (OR<1.0, p≤0.05). For the remaining 221 models, there was a near even split in how often the odds of injury were non-significantly higher (n=103) and non-significantly lower (n=114) for females as compared to males (p>0.05). In four cases, the OR estimate was 1.00. Amongst the results, there was a trend for females to have higher odds of AIS 2+ injuries (MAIS 2+ OR=1.75 and 1.69 for Full and Crashworthiness models, respectively for the All Crashes dataset). These increases included higher estimates for lower extremity injuries in frontal crashes, consistent with earlier studies (e.g., Forman et al. 2019). However, for certain AIS 2+ (neck, thorax) and AIS 3+ injuries (head, neck, thorax), females had significantly lower odds of injury (p≤0.05). The trends for reduced odds of injury for females were most prevalent in non-frontal crash models.

Predictive Factors of the Fatality of Motor Vehicle Passengers Involved in Far-Side Lateral Collisions: A National Crash Database Study.

Although the risks faced by passengers in near-side lateral collisions are understood, and despite the presence of side airbags for injury prevention, passengers involved in far-side lateral collisions also suffer serious and fatal injuries. The objective of this study was to determine the independent predictive factors of fatality of motor vehicle passengers involved in far-side lateral collisions. Using 2010 records from the National Automotive Sampling System/Crashworthiness Data System (NASS/CDS), we selected 86 fatal and 325 non-fatal passengers with an Abbreviated Injury Scale (AIS) score of 2 or more. The background and injury severity of the passengers and collision characteristics were compared between the two groups. In a multivariable logistic regression analysis, variables independently associated with fatalities were female sex (Ref, male) (odds ratio [OR], 0.396), age (OR, 1.029), body mass index (OR, 1.057), total delta-V (OR, 1.031), head AIS score (OR, 1.679), chest AIS score (OR, 1.330), and abdomen AIS score (OR, 1.294). This is the first report to determine factors affecting fatality in passengers involved in far-side lateral collisions. Improving the safety of the vehicle interior, such as by including additional seatbelt systems or a side airbag that deploys between seats, might help to avoid fatalities, and reduce injury severity.

Cost-Benefit Analysis to Assess the Effectiveness of an External Airbag and Autonomous Emergency Braking System

One of the recent developments in safety systems is an external airbag installed on the front bumper of a vehicle and autonomous emergency braking system. In this paper, we propose a framework for a cost-benefit analysis of the external airbag and autonomous emergency braking system in order to validate its commercialization. Road traffic crash data obtained from the National Automotive Sampling System/Crashworthiness Data System (NASS/CDS) was used, and three different crash types related to frontal damage in vehicles were extracted to estimate the safety performance of an external airbag. An ordinal logistic regression model was applied to estimate the safety performance in terms of the reduced maximum abbreviated injury scale (MAIS) based on a reduction in the total delta-v following the installation of an external airbag. Given the estimated safety performance of the external airbag, a cost-benefit analysis is conducted. According to the results, the external airbag system saves 46% of occupants with MAIS 3+ injuries and prevents 40% of fatalities. Moreover, the benefit/cost ratios of the external airbag system range from 0.496 to 0.509 depending on the scenario. Lastly, sensitivity analyses were performed with important parameters, including the initial and maximum market penetration ratio and the price of the system.

Exploring the temporal variability of the factors affecting driver injury severity by body region employing a hybrid econometric approach

The current study contributes to safety literature by incorporating the influence of temporal factors (observed and unobserved) within a multivariate model system for medical professional generated body region specific injury severity score. For this purpose, we adopt a hybrid econometric modeling approach that accommodates for the unobserved factors using two mechanisms. First, we parameterize unobserved temporal factor variation through the customization of the variance by time cohort (heteroscedasticity). Second, the common unobserved factors affecting severity across various body regions is accommodated through traditional random parameter consideration process. The proposed model system is estimated using data drawn from the National Automotive Sampling System-Crashworthiness Data System (NASS-CDS) database for the time cohorts 2003, 2006, 2009, 2012, and 2015. For the current analysis, we consider 6-point Abbreviated Injury Scale (AIS) for eight body regions (head, face, neck, abdomen, thorax, spine, lower extremity, and upper extremity). The proposed model system offers interesting insights on body region severity evolution over time. The model estimation is augmented with post-estimation exercises including hold-out sample validation analysis, illustrative policy analysis and extensive elasticity effect computation.

Identifying Pedal Misapplication Behavior Using Event Data Recorders

&lt;div class="section abstract"&gt;&lt;div class="htmlview paragraph"&gt;Pedal misapplication (PM) crashes, i.e., crashes caused by a driver pressing one pedal while intending to press another pedal, have historically been identified by searching unstructured crash narratives for keywords and verified via labor-intensive manual inspection. This study proposes an alternative method to identify PM crashes using event data recorders (EDRs). Since drivers in emergency braking situations are motivated to hit the brake hard, it follows that drivers in emergency braking situations that commit a PM would likewise hit the accelerator hard, likely harder than accelerator pedal application during normal driving. Thus, the time-series accelerator pedal position and the derived accelerator pedal application rate were used to isolate accelerator misapplications. Additional strategic filters were applied based on characteristics observed from previous PM analyses to reduce false positive PM identifications. These include a crash type filter, since PM crashes have been shown to manifest as majority road departure, end departure, rear-end, and forward impact crash types. After analyzing pre-crash EDR data from the National Automotive Sampling System Crashworthiness Data System (NASS/CDS) case years 1997 to 2015, evidence of PM was observed in 4.3% of weighted events. This result was substantially higher than the previously estimated 0.2% PM frequency [&lt;span class="xref"&gt;1&lt;/span&gt;,&lt;span class="xref"&gt;2&lt;/span&gt;]. The time-to-collision (TTC) at the point of PM was calculated for each case, and over 50% of cases had a TTC of less than 2.0 seconds. Over one-third of these drivers engaged the accelerator to 99% of pedal stroke or above and over one-eighth of drivers engaged both the brake and the accelerator pedals simultaneously during the recorded pre-crash period.&lt;/div&gt;&lt;/div&gt;

Real-Time Regression Analysis of Streaming Clustered Data With Possible Abnormal Data Batches

This article develops an incremental learning algorithm based on quadratic inference function (QIF) to analyze streaming datasets with correlated outcomes such as longitudinal data and clustered data. We propose a renewable QIF (RenewQIF) method within a paradigm of renewable estimation and incremental inference, in which parameter estimates are recursively renewed with current data and summary statistics of historical data, but with no use of any historical subject-level raw data. We compare our renewable estimation method with both offline QIF and offline generalized estimating equations (GEE) approach that process the entire cumulative subject-level data all together, and show theoretically and numerically that our renewable procedure enjoys statistical and computational efficiency. We also propose an approach to diagnose the homogeneity assumption of regression coefficients via a sequential goodness-of-fit test as a screening procedure on occurrences of abnormal data batches. We implement the proposed methodology by expanding existing Spark’s Lambda architecture for the operation of statistical inference and data quality diagnosis. We illustrate the proposed methodology by extensive simulation studies and an analysis of streaming car crash datasets from the National Automotive Sampling System-Crashworthiness Data System (NASS CDS). Supplementary materials for this article are available online.

An omni-directional model of injury risk in planar crashes with application for autonomous vehicles

Objective: Automated driving systems (ADS) are actively being deployed within the driving fleet. ADS are designed to safely navigate roadways, which entails an expectation of encountering varying degrees of potential conflict with other road users. The ADS design and evaluation process benefits from estimating injury severity probabilities for collisions that may occur. Current regression models in the literature are typically bespoke analyses involving targeted principal directions of force (PDOFs) and occupant positions. It is preferable to rely on injury severity models derived from a single source to provide a continuous function of risk for all planar collisions, while also accounting for specific vehicle and occupant characteristics. The novel feature of the proposed models is continuous, parametric injury risk surfaces that encompass the full spectrum of available United States field data. Methods: We used years 2001–2015 of the National Automotive Sampling System, Crashworthiness Data System (NASS-CDS) and years 2017–2019 of the Crash Investigation Sampling System (CISS) to estimate injury risk at the maximum abbreviated injury scale (MAIS) 3 and higher (3+) and 5 and higher (5+) levels for all adult occupants traveling in 2002 or newer passenger vehicles which were less than 10 years old at the time of the crash. The models account for occupant, vehicle, and crash characteristics. Interactions with vulnerable road users (e.g., pedestrian, bicyclist) were not considered. Results: We present statistical models suitable to predict injury in all non-rollover crashes at the maximum MAIS3+ and 5+ levels, and show that these models can be comparable to similar single scenario (e.g., frontal) crash models. We discuss challenges with imputing missing field data, and discuss handling of covariates that may not be known at the time of the crash. Conclusions: Collision severity assessment is a vital component of the ADS design process. We developed a novel injury risk function that can assess occupant injury risks across the spectrum of foreseeable planar collisions. These models can provide insight on potential outcomes of counterfactual simulations, injury risk and crashworthiness considerations for human driven vehicles, and provide an evaluation tool that can be applied in ADS safety impact evaluation.

Parallelized optimization design of bumper systems under multiple low-speed impact loads

Majority of existing design optimization of bumper systems has focused on one single impact loading case in literature, which may not offer proper performance under other loading cases. This study aims to address this issue by developing a multiobjective optimization approach, in which multiple impact loading scenarios are considered to optimize the intrusion and energy absorption of the bumper system. Modeling of real-life impact scenarios is first validated by comparing the numerical results with the experimental data. Based upon the Latin hypercube design (LHD) and Kriging surrogate approaches, energy absorption (EA) and maximum intrusion (MI) are formulated in terms of design variables to consider the structural crashworthiness and repair/maintenance cost, respectively. In order to generate Pareto solution in the multiobjective optimization effectively, the recently developed Kriging Believer based parallelized efficient global optimization (EGO) algorithm is adopted to optimize the bumper system subject to three impact loads, namely the low-speed sled crash, dynamic three-point bending and 40% offset impact, respectively. These different load cases are considered through a weighted average scheme. The optimization results indicate that the Pareto solutions largely depend on the assignment of weight factors to different load cases. In order to generate more reasonable results, the National Automotive Sampling System-Crashworthiness Data System (NASS-CDS) database is also used to determine the weighting factors for each loading case. It is found that for the intrusion objective, the design under the multiple impact loads offers an intrusion reduction of 20.27% in the sled crash, 5.84% in the 40% offset impact, in comparison with the baseline design. The intrusion of three-point bending is the same as that of baseline design to compromise other more frequently occurring load cases. For the energy absorption objective, although the design leads to increase in EA by 3.01% in the sled crash and 1.79% in 40%-offset impact, the performance in the three-point bending is actually sacrificed to compromise the other more frequently occurring load cases. The study help gain insights into the design optimization of a bumper system for real-life automotive engineering.

Injury risk curves in far-side lateral motor vehicle crashes by AIS level, body region and injury code

Objective The objective of this study was to develop injury risk curves as a function of change in vehicle velocity for occupants in far-side lateral motor vehicle crashes (MVCs) by AIS level, body region, and specific AIS codes that commonly occur in this crash mode. Methods The National Automotive Sampling System-Crashworthiness Data System (NASS-CDS) years 2000–2015 database was queried, resulting in 4,495 non-weighted far-side crashes. For each case, occupant age, sex, and the following metadata were collected: vehicle model year, vehicle body type, lateral delta-v, normalized PDOF, multiple impacts, belt use, seat position, object contacted, striking vehicle body type, maximum crush extent and side airbag deployment. Multivariable logistic regression was used to develop risk curves for AIS 2+ through 5+ injuries, AIS 2+ injuries by body region (head, thorax, lower extremity), and for each of the 10 most frequent far-side AIS 2+ injuries. Significant covariates were determined by backwards elimination (p < 0.05). The full dataset and a subsampled dataset of only cases with side airbag deployment were used to develop risk curves. Results For AIS 2+ through 5+ injury, greater delta-V was associated with greater injury risk (OR’s: 2.48–3.66 per 11.9 kph increase) and belt use was associated with lower risk (OR’s: 0.04–0.36 compared to unbelted). Multiple impacts were significant predictors of increased AIS 3+, 4+ and 5+ injury risk (OR’s: 2.56, 2.27 and 2.83 compared to single impact). For AIS 2+ body region injuries, lateral delta-V and maximum CDC extent were positively associated with increased head, thorax and lower extremity injury risk while belt use was associated with lower risk. Increased lateral delta-v, unbelted status, and greater maximum CDC extent frequently increased injury risk for the most common far-side injuries. Side airbag deployment was not a significant covariate for the injury risk models. Conclusions The resulting risk models expand upon previous literature gaps to provide a more comprehensive view of contributors to injury risk for occupants in far-side MVCs. This study yields risk curves based on the latest available NASS-CDS data.

An examination of the performance of damaged energy-absorbing end terminals

ObjectivesGuardrail end terminals are designed to gradually decelerate vehicles during impact and protect vehicle occupants from severe injuries. It has been observed that some in-service end terminals are damaged, and it is unclear if their safety performance is still acceptable. The objectives of this study were to examine the conditions of in-service end terminals, and to evaluate the performance of damaged relative to undamaged end terminals in simulated impacts. MethodsCommon damage patterns of guardrail end terminals were investigated by using post-crash pictures collected from the National Automotive Sampling System-Crashworthiness Data System (NASS-CDS). Conditions of in-service end terminals mounted along roads in portions of six U.S. states were examined by using a sample from the second Strategic Highway Research Program-Roadway Information Database (SHRP2-RID). Finite Element (FE) models of two minorly and three severely damaged ET-Plus systems, a commonly used energy-absorbing guardrail end terminal along U.S. roads, were developed. To evaluate the performance of the damaged ET-Plus systems, we performed impact simulations with vehicle-to-damaged ET-Plus models according to the National Cooperative Highway Research Program (NCHRP) 350, test conditions 3–30. ResultsOf the 1000 in-service end terminal cases we investigated, 73% were undamaged, 18% had minor damage, and 8% had major damage. Increases in the average vehicle deceleration rates, maximum vehicle yaw angles, and vehicle local deformations were observed in simulated impacts with damaged ET-Plus end terminals relative to impacts with undamaged ET-Plus end terminals. For one damaged ET-Plus, a secondary collision was observed. Overall, we found that the damaged end terminals usually increased collision severity when compared with undamaged end terminals. ConclusionsThe findings of this study point out the need for in-service performance evaluations and proper maintenance and repair practices of end terminals. The simulation models developed in this study could be further employed to investigate device performance in crash situations that are physically impractical to test and investigate the effects of site characteristics on device performance. The simulation models could also supplement crash tests to certify new hardware designs.

Expansion of NASS/CDS for characterizing run-off-road crashes

Objective Run-off-road (ROR) crashes account for one-third of all annual crash fatalities in the US. The National Automotive Sampling System Crashworthiness Data System (NASS/CDS) is a dataset which may be used to understand the nature of ROR crashes. Despite the wealth of coded data available in NASS/CDS, this dataset lacks coded information about the roadside environment and the off-road trajectory of the vehicle. This information would be useful for determining lane departure warning (LDW) benefits, residual safety problems, performance of current safety hardware, lane marking inventory, LDW test procedure development, radius of curvature characterization, and effectiveness of ESC. The purpose of this paper is to demonstrate a methodology for expanding the data available in NASS/CDS to form and validate a specialized road departure database. Methods Observed, measured, and reconstructed data elements were extracted from NASS/CDS and compiled into the National Cooperative Highway Research Program (NCHRP) 17-43 database. Observed variables were primarily coded from the scene photographs and included information such as the lane markings, and geometry of the roadside cross-section. Additional variables were measured from the scaled scene diagrams including the path of the vehicle, road dimensions, and roadside object positions. The vehicle impact speed and departure speed were reconstructed using the WinSMASH delta-v, roadside object characteristics, and vehicle path. Two studies were conducted to demonstrate the usefulness of the NCHRP 17-43 database in evaluating both vehicle-based and infrastructure-based ROR countermeasures. Results The resulting NCHRP 17-43 database includes 1,581 NASS/CDS cases representing 510,154 ROR crashes. Analysis of the database found that drivers which crashed following an overcorrection were younger than drivers which did not overcorrect. This may indicate that inexperienced drivers are more likely to overcorrect when departing the roadway. The 85th percentile impact severity of ROR crashes, which occur on roads with a speed limit greater than 65 mph, is higher than the practical worst-case test conditions for roadside barriers. Conclusions The NCHRP 17-43 database contains information extracted from NASS/CDS cases to better understand the nature of ROR crashes, driver behavior in these crashes, and the potential benefits of both vehicle-based and infrastructure-based ROR countermeasures.

Estimating the number of traffic crash-related cervical spine injuries in the United States; An analysis and comparison of national crash and hospital data

BackgroundCervical spine injury is a common result of traffic crashes, and such injuries range in severity from minor (i.e. sprain/strain) to moderate (intervertebral disk derangement) to serious and greater (fractures, dislocations, and spinal cord injuries). There are currently no reliable estimates of the number of crash-related spine injuries occurring in the US annually, although several publications have used national crash injury samples as a basis for estimating the frequency of both cervical and lumbar spinal disk injuries occurring in lower speed rear impact crashes. PurposeTo develop a reliable estimate of the number of various types of cervical spine injuries occurring in the US by comparing data from national crash injury to national hospital ED and inpatient samples. Study DesignComparative cross-sectional MethodsCervical spine injury data were accessed, analyzed, and compared from 3 national databases; the National Automotive Sampling System-Crashworthiness Data System (NASS-CDS), Nationwide Emergency Department Sample (NEDS), and the Nationwide Inpatient Sample (NIS). ResultsIt is estimated that there are approximately 869,000 traffic crash-related cervical spine injuries seen in hospitals in the US annually, including around 841,000 sprain/strain (whiplash) injuries, 2800 spinal disk injuries, 23,500 fractures, 2800 spinal cord injuries, and 1500 dislocations. Because of a highly restrictive inclusion criteria for both crash and injury types, as well as a very small sample size, the NASS-CDS underestimated all types of crash-related cervical spine injuries seen in US hospital emergency departments by 84 %. The injury type with the largest degree of underestimation in the NASS-CDS was cervical disk injuries, which were estimated at an 88 % lower frequency than in the NEDS. National insurance claim data, which include cases of cervical disk injury diagnosed both in and outside of the ED, indicate that the NEDS likely undercounts cervical disk injuries by 92 %, and thus the NASS-CDS correspondingly undercounts such injuries by 99 % or more. ConclusionsBecause of a limited sample size and restrictive criteria for both crash and injury inclusion, the NASS-CDS cannot be used to estimate the number of crash-related spinal injuries of any type or severity in the US. The most inappropriate use of the database is for estimating the number of spinal injuries resulting from low speed rear impact collisions, as the NASS-CDS samples fewer than 1 in 100,000 of the cervical spine injuries of any type occurring in low speed rear impact collisions.

Estimating near side crash injury risk in best performing passenger vehicles in the United States

The objective of this paper was to develop an injury risk model relating real world injury outcomes in near-side crashes with U.S. New Car Assessment Program (NCAP) test performance, crash, and occupant properties. The study was motivated by the longer-term goal of predicting injury outcomes in a future fleet in which all vehicles are expected to have passive safety performance equivalent to a 5-star NCAP rating level (the highest star rating and lowest risk of injury).The dataset used to evaluate injury risk was the National Automotive Sampling System / Crashworthiness Data System (NASS/CDS). Case years 2010–2015 were used. An injured occupant was defined as a vehicle occupant who experienced an injury of maximum Abbreviated Injury Scale (AIS) of 2 or greater, or who were fatally injured. Injury severity was scored using AIS-2005 (2008 update). Cases were selected in which front-row occupants of late-model vehicles were exposed to a near-side crash. Logistic regression was used to develop an injury model with delta-v, belt status, age, and gender as predictor variables. The side crash performance of each vehicle was identified and added to the model by matching each case with the associated performance in the NCAP moving deformable barrier side impact crash test.NCAP MDB test performance, delta-v, and occupant age, sex, and BMI were found to be significant predictors of injury risk. The effect of a 5 % higher risk in the MDB test (approximately one star rating worse) was comparable to a 2.84 km/h increase in delta-v. This model informs the development of active safety systems in a future fleet where vehicle passive safety performance is higher, quantified by the NCAP MDB test.

Evaluating the influence of knee airbags on lower limb and whole-body injury

Objective: Knee airbags (KABs) have become increasingly common in the vehicle fleet. Previous studies (Weaver et al., 2013, Patel et al. 2013) showed indications that KABs may be protective for some lower extremity injuries and associated with increased risk for others. Since KABs have become significantly more common in recent model year vehicles, we revisited these findings using the most recent available data.Methods: We compared injury rates below the knee, from the knee to the hip, and above the hip in years 2000-2015 of the National Automotive Sampling System, Crashworthiness Data System (NASS-CDS) and the Crash Injury Research and Engineering Network (CIREN). Injury rates were compared with matched analyses and with Bayesian multiple logistic regression.Results: Both analyses showed that KAB to have an Odds Ratio of approximately 0.6 for knee to hip injuries, with the Bayesian model strongly significant and the matched model borderline insignificant. In the Bayesian model, KAB was borderline significant for a decrease in above the waist injuries, while the matched model pointed toward a protective effect but was not significant. Both models pointed toward an increased risk of below knee injuries, but neither was statistically significant.Conclusions: Knee airbags may be protective for knee to hip injuries and above waist injuries. If KABs continue to be widely implemented in the vehicle fleet, the field should continue to monitor and evaluate below knee injuries.

Accuracy of algorithms to predict injury severity in older adults for trauma triage

Objective: Older adults make up a rapidly increasing proportion of motor vehicle occupants and previous studies have demonstrated that this population is more susceptible to traumatic injuries. The CDC recommends that patients anticipated to have severe injuries (Injury Severity Score [ISS] ≥ 16) be transported to a trauma center. The recommended target rate for undertriage is ≤ 5% and for overtriage is ≤ 50%. Several regression-based algorithms for injury prediction have been developed in order to predict severe injury in occupants involved in a motor vehicle collision (MVC). The objective of this study to was to determine if the accuracy of regression-based injury severity prediction algorithms decreases for older adults.Methods: Data were obtained from the National Automotive Sampling System – Crashworthiness Data System (NASS-CDS) from the years 2000–2015. Adult occupants involved in non-rollover MVCs were included. Regression-based injury risk models to predict severe injury (ISS ≥ 16) were developed using random split-samples with the following variables: age, delta-V, direction of impact, belt status, and number of impacts. Separate models were trained using data from the following age groups: (1) all adults, (2) 15–54 years, (3) ≥45 years, (4) ≥55 years, and (5) ≥65 years. The models were compared using the mean receiver operating characteristic area under curve (ROC-AUC) after 1,000 iterations of training and testing. The predicted rates of overtriage were then determined for each group in order to achieve an undertriage rate of 5%.Results: There were 24,577 occupants (6,863,306 weighted) included in this analysis. The injury prediction model trained using data from all adults did not perform as well when tested on older adults (ROC-AUC: 15–54 years: 0.874 [95% CI: [0.851–0.895]; 45+ years: 0.837 [95% CI: 0.802–869]; 55+ years: 0.821 [95% CI: 0.775–0.864]; and 65+ years: 0.813 [95% CI: 0.754–0.866]). The accuracy of this model decreased in each decade of life. The performance did not change significantly when age-specific data were used to train the prediction models (ROC-AUC: 18–54 years: 0.874 [95% CI: 0.851–0.896]; 45+ years: 0.836 [95% CI: 0.798–0.871]; 55+ years: 0.822 [95% CI: 0.779–0.864]; and 65+ years: 0.808 [95% CI: 0.748–0.868]). In order to achieve an undertriage rate of 5%, the predicted overtriage rate by these models were 50% for occupants 15–54 years, 61% for occupants ≥ 55 years, 70% for occupants ≥ 55 years, and 71% for occupants ≥ 65 years.Conclusion: The results of this study indicate that it is more difficult to accurately predict severe injury in older adults involved in MVCs, which has the potential to result in significant overtriage. This decreased accuracy is likely due to variations in fragility in older adults. These findings indicate that special care should be taken when using regression-based prediction models to determine the appropriate hospital destination for older occupants.

Crash Telemetry-Based Injury Severity Prediction is Equivalent to or Out-Performs Field Protocols in Triage of Planar Vehicle Collisions.

With the increasing availability of vehicle telemetry technology, there is great potential for Advanced Automatic Collision Notification (AACN) systems to improve trauma outcomes by detecting patients at-risk for severe injury and facilitating early transport to trauma centers. National Automotive Sampling System Crashworthiness Data System (NASS-CDS) data from 1999-2013 were used to construct a logistic regression model (injury severity prediction [ISP] model) predicting the probability that one or more occupants in planar, non-rollover motor vehicle collisions (MVCs) would have Injury Severity Score (ISS) 15+ injuries. Variables included principal direction of force (PDOF), change in velocity (Delta-V), multiple impacts, presence of any older occupant (≥55 years old), presence of any female occupant, presence of right-sided passenger, belt use, and vehicle type. The model was validated using medical records and 2008-2011 crash data from AACN-enabled Michigan (USA) vehicles identified from OnStar (OnStar Corporation; General Motors; Detroit, Michigan USA) records. To compare the ISP to previously established protocols, a literature search was performed to determine the sensitivity and specificity of first responder identification of ISS 15+ for MVC occupants. The study population included 924 occupants in 836 crash events. The ISP model had a sensitivity of 72.7% (95% Confidence Interval [CI] 41%-91%) and specificity of 93% (95% CI 92%-95%) for identifying ISS 15+ occupants injured in planar MVCs. The current standard 2006 Field Triage Decision Scheme (FTDS) was 56%-66% sensitive and 75%-88% specific in identifying ISS 15+ patients. The ISP algorithm comparably is more sensitive and more specific than current field triage in identifying MVC patients at-risk for ISS 15+ injuries. This real-world field study shows telemetry data transmitted before dispatch of emergency medical systems can be helpful to quickly identify patients who require urgent transfer to trauma centers.

Factors contributing to serious and fatal injuries in belted rear seat occupants in frontal crashes

Objectives: Earlier research has shown that the rear row is safer for occupants in crashes than the front row, but there is evidence that improvements in front seat occupant protection in more recent vehicle model years have reduced the safety advantage of the rear seat versus the front seat. The study objective was to identify factors that contribute to serious and fatal injuries in belted rear seat occupants in frontal crashes in newer model year vehicles.Methods: A case series review of belted rear seat occupants who were seriously injured or killed in frontal crashes was conducted. Occupants in frontal crashes were eligible for inclusion if they were 6 years old or older and belted in the rear of a 2000 or newer model year passenger vehicle within 10 model years of the crash year. Crashes were identified using the 2004–2015 National Automotive Sampling System Crashworthiness Data System (NASS-CDS) and included all eligible occupants with at least one Abbreviated Injury Scale (AIS) 3 or greater injury. Using these same inclusion criteria but split into younger (6 to 12 years) and older (55+ years) cohorts, fatal crashes were identified in the 2014–2015 Fatality Analysis Reporting System (FARS) and then local police jurisdictions were contacted for complete crash records.Results: Detailed case series review was completed for 117 rear seat occupants: 36 with Maximum Abbreviated Injury Scale (MAIS) 3+ injuries in NASS-CDS and 81 fatalities identified in FARS. More than half of the injured and killed rear occupants were more severely injured than front seat occupants in the same crash. Serious chest injury, primarily caused by seat belt loading, was present in 22 of the injured occupants and 17 of the 37 fatalities with documented injuries. Nine injured occupants and 18 fatalities sustained serious head injury, primarily from contact with the vehicle interior or severe intrusion. For fatal cases, 12 crashes were considered unsurvivable due to a complete loss of occupant space. For cases considered survivable, intrusion was not a large contributor to fatality.Discussion: Rear seat occupants sustained serious and fatal injuries due to belt loading in crashes in which front seat occupants survived, suggesting a discrepancy in restraint performance between the front and rear rows. Restraint strategies that reduce loading to the chest should be considered, but there may be potential tradeoffs with increased head excursion, particularly in the absence of rear seat airbags. Any new restraint designs should consider the unique needs of the rear seat environment.

Residual road departure crashes after full deployment of LDW and LDP systems

Objective: Road departures are one of the most severe crash modes in the United States. To help reduce this risk, vehicles are being introduced in the United States with lane departure warning (LDW) systems, which warn the driver of a departure, and lane departure prevention (LDP) systems, which assist the driver in steering back to the roadway. Previous studies have estimated that LDW/LDP systems may prevent one third of drift-out-of-lane road departure crashes. This study investigates the crashes that were not prevented, to potentially set research priorities for next-generation road departure prevention systems.Methods: The event data recorder (EDR) data from 128 road departure crashes in the National Automotive Sampling System Crashworthiness Data System (NASS-CDS) from 2011 to 2015 were mapped onto the vehicle trajectory and simulated with LDW/LDP to assess the potential for crash avoidance. The model predicted that 63–83% of single-vehicle road departure crashes may not be prevented by an LDW system and 49% may not be prevented by an LDP system.Results and Conclusions: For LDP systems, which were assumed to have zero latency, no crashes were avoided if the time-to-collision (TTC) from lane crossing to impact was less than 0.55 s. Obstacles such as guardrails and traffic barriers, which tend to be very close to the road, were more common among the remaining crashes. The study shows that LDW/LDP systems are limited by two factors, driver reaction time and TTC to the roadside object. Thus, earlier driver response and longer TTC may help in these situations.

Vehicle deformation depth based injury risk function for safety benefit evaluation of crash avoidance and mitigation systems

As the delta-v-based injury models used to evaluate intelligent driving systems are always fitted with European or American crash database, they cannot achieve wide application in those countries where limited in-depth analysed crash data are recorded. An injury model which is based on easily accessible information, is urgently needed. In this study, a deformation depth based injury risk model is proposed to overcome the limitation of delta-v. First, a correlation between the vehicle deformation depth and occupant injury risk is verified from the aspects of retrospective safety assessment and stiffness cluster analysis using German in-depth accident study and national automotive sampling system-crashworthiness data system. Furthermore, injury risk-deformation functions are regressed for different stiffness clusters using the crash data. The fitting accuracy reaches 97%, higher than the existing literature. A novel safety benefit assessment simulation platform is built with the regressed injury risk model. Based on this platform, an autonomous emergency braking system is evaluated. Only 1% error of the safety benefit exists between the proposed model and the delta-v based one.

Rear-facing versus forward-facing child restraints: an updated assessment

ObjectivesThe National Highway Traffic Safety Administration and the American Academy of Pediatrics recommend children be placed in rear-facing child restraint systems (RFCRS) until at least age 2. These recommendations are...