Introduction: Pre-hospital triage for motor-vehicle crashes relies on occupant or bystander notification that a crash has occurred. Newer vehicles are being equipped with enhanced Automatic Collision Notification Systems (AACN), which automatically notify emergency systems of a crash. The vehicles transmit basic crash data using on-board sensors and an algorithm known as “URGENCY” calculates the risk of serious injuries. We sought to examine accuracy of the Urgency model and the potential impact of using this system for the triage of crashes where no contact can be made with the occupant or injuries are underestimated. Methods: Seriously injured occupants are defined as those who have sustained one or more injuries with an Abbreviated Injury Severity Score of 3 or higher (MAIS 3+). The Urgency model uses covariates such as deltaV (velocity), crash direction, belt use and number of impacts, which are all intuitively related to injury risk and predictive of injury as shown by the models. The call logs for all BMW crashes in the US from 2006 to 2008 were analyzed. Crashes were matched with police reported cases to identify the basic injury severity. Hospital transport to either a trauma center or non-trauma center was recorded. Both sources were merged using the unique Vehicle Identification Number (VIN) and the crash date as criteria for linkage. Results: Validation of the Urgency model revealed that each parameter estimate was highly significant (p value<0.05). The odds ratios for each discrete variable were significantly different than 1.0 while controlling for other factors. The overall predictive accuracy of the Urgency model suggests that 76% of the serious injured and 91% of the non-serious injured occupants would be correctly identified by Urgency. Of the 14,008 AACN crashes, 153 had no voice communication. 27% of these had injuries with 5% serious injuries. For those with injuries, 17% were treated in a non-trauma center and 3% in a trauma center. These 20% of cases in which hospital transport occurred but no voice communication was obtained represent the area where AACN can improve triage times. In contrast, for voice communication cases where the occupant indicated no injuries, police coded 23% of these actually had injuries. 20% sustained low severity injuries and 3% moderate or serious injuries. 10% were transported to a local hospital and 3% to a trauma center. This 13% who required hospitalization shows that in some cases the occupants cannot estimate their own injuries. Conclusions: The URGENCY algorithm predicted the risk of serious injuries with an ability to capture approximately 76% of MAIS 3+ injured occupants and correctly identify 91% of occupants as non MAIS 3+ injured. 20% of the “no voice” cases and 13% of those who felt they were uninjured were transported to a hospital. The AACN system and Urgency can potentially provide triage decision-support for these crashes.

The SceneScore is a simple mechanism of injury scoring system designed to facilitate the appropriate triage of crash victims. It comprises 7 variables including age, collision type, impact location, airbag deployment, steering wheel deformity, intrusion, and restraint use. A cutoff value of 7 or 8 provides the maximum balance between sensitivity and specificity, with sensitivities of 75% to 83% and specificities of 29% to 46%. For cases triaged to the trauma center based only on high suspicion of injury, the SceneScore reduces the overtriage rate by almost half. Proper application of the SceneScore may lead to improved triage and enhanced communication of mechanism of injury criteria.

This study applies NASS/CDS, GES and FARS data to examine occupant exposure plus injury and fatality rates for belted occupants in frontal crashes by seating position, age and gender. The NASS data was used to examine the distributions by crash severity. The GES data showed that when two elderly occupants (age 65+) were present, the female occupied the right front passenger position 73% of the time. A paired comparison analysis using FARS data showed that, for elderly occupants (age 65+), the fatality risk for elderly right front passengers is 42% higher than for elderly drivers. The NASS/CDS analysis found 74% of the seriously injured vulnerable passengers with MAIS 3+ injuries were in crashes less severe than 26 mph. This group of injured occupants was made up of 43% aged 50 and older and 42% younger females. The injury rates for the older (age 50+) right front passengers were 1.8 times the rates for the elderly drivers. These results suggest the need for more benign safety systems for the right front passenger that are appropriate for the lower injury tolerance of the predominant occupants of that seating position.

NASS/CDS 1995-2004 was used to classify rollovers according to severity. The rollovers were partitioned into two classes - rollover as the first event and rollover preceded by an impact with a fixed or non-fixed object. The populations of belted and unbelted were examined separately and combined. The average injury rate for the unbelted was five times that for the belted. Approximately 21% of the severe injuries suffered by belted occupants were in crashes with harmful events prior to the rollover that produced severe damage to the vehicle. This group carried a much higher injury risk than the average. A planar damage measure in addition to the rollover measure was required to adequately capture the crash severity of this population. For rollovers as the first event, approximately 1% of the serious injuries to belted occupants occurred during the first quarter-turn. Rollovers that were arrested during the 1 ( st ) quarter-turn carried a higher injury rate than average. The number of quarter-turns were grouped in various ways including the number of times the vehicle roof faces the ground (number of vehicle inversions). The number of vehicle inversions was found to be a statistically significant injury predictor for 78% of the belted and unbelted populations with MAIS 3+F injuries in rollovers. The remaining 22% required crash severity metrics in addition to the number of vehicle inversions.

This study examines the residual injuries reported in NASS/CDS 1997-2004 by crash mode, crash severity, body region and occupant age. It examines how serious injuries are distributed in present day crashes and identifies opportunities for further injury reduction. In planar crashes, approximately 66% of the MAIS 3+ injuries occur in crashes less severe than 25 mph delta-V. Chest injuries predominate in these crashes, particularly among elderly occupants. A reduction in chest injuries to belted elderly occupants during low severity frontal crashes offers a prime opportunity for further improvement of safety systems. Younger occupants could also benefit from improved chest protection.

The population of occupants in far-side crashes that are documented in the US National database (NASS/CDS) was studied. The annual number of front seat occupants with serious or fatal injuries in far-side planar and rollover crashes was 17,194. The crash environment that produces serious and fatal injuries to belted front seat occupants in planar far-side crashes was investigated in detail. It was found that both the change in velocity and extent of damage were important factors that relate to crash severity. The median severity for crashes with serious or fatal injuries was a lateral delta-V of 28 kph and an extent of damage of CDC 3.6. Vehicle-to-vehicle impacts were simulated by finite element models to determine the intrusion characteristics associated with the median crash condition. These simulations indicated that the side damage caused by the IIHS barrier was representative of the damage in crashes that produce serious injuries in far-side crashes. Occupant simulations of the IIHS barrier crash at 28 kph showed that existing dummies lack biofidelity in upper body motion. The analysis suggested test conditions for studying far-side countermeasures and supported earlier studies that showed the need for an improved dummy to evaluate safety performance in the far-side crash environment.

This study investigates injury occurrence for belted occupants as a function of age. An analysis of NASS/CDS 1997-2003 data was conducted to determine crash involvement rates and injury rates for front seat occupants versus mean occupant age. In frontal and near-side crashes, the average age of MAIS 3+ belted front seat occupants injured in crashes less severe than 15 mph is of the order of 50 years. The average age of the population exposed to crashes less severe than 15 mph is under 40 years old. The crash exposure and frequency if injuries to the elderly were both found to be the highest in low severity crashes. The chest is the most frequent body region injured for the elderly. These findings suggest the need for more benign safety systems to protect the elderly in low severity crashes. Design of safety systems for the elderly should give priority to reducing the chest loading in low severity frontal and near-side crashes.

The advent of Automatic Crash Notification Systems (ACN) offers the possibility of immediately locating crashes and of determining the crash characteristics by analyzing the data transmitted from the vehicle. A challenge to EMS decision makers is to identify those crashes with serious injuries and deploy the appropriate rescue and treatment capabilities. The objective of this paper is to determine the crash characteristics that increase the risk of serious injury.Within this paper, regression models are presented which relate occupant, vehicle and impact characteristics to the probability of serious injury using the Maximum Abbreviated Injury Scale Level (MAIS). The accuracy of proposed models were evaluated using National Automotive Sampling System/Crashworthiness Data System (NASS/CDS) and Crash Injury Research and Engineering Network (CIREN) case data. Cumulatively, the positive prediction rate of models identifying the likelihood of MAIS3 and higher injuries was 74.2%.Crash mode has a significant influence of injury risk. For crashes with 30 mph deltaV, the risk of MAIS3+ injury for each mode is 38.9%, 83.8%, 47.8% and 19.9% for frontal, near side, far side and rear impact crashes, respectively. In addition to deltaV, a number of crash variables were identified that assist in the accurate prediction of the probability of MAIS 3+ injury. These variables include occupant age, partial ejection, safety belt usage, intrusion near the occupant, and crashes with a narrow object. For frontal crashes, added crash variables include air bag deployment, steering wheel deformation, and multiple impact crashes. The quantitative relationship between each of these crash variables and injury risk has been determined and validated by regression analysis based on NASS/CDS and CIREN data.

Die Entwicklung von Automatischen Unfallbenachrichtigungssystemen (ACN) bietet die Moeglichkeit, Unfaelle sofort zu lokalisieren und die Unfallmerkmale durch Analyse der vom Fahrzeug uebertragenen Daten sofort festzulegen. Dies stellt eine Herausforderung fuer EMS-Entscheidungstraeger dar, die Unfaelle mit schweren Verletzungen feststellen und die entsprechenden Rettungs- und Behandlungseinrichtungen aktivieren. Ziel dieser Untersuchung ist die Festlegung der Aufprallmerkmale, die das Risiko schwerer Verletzungen erhoehen. In dieser Studie werden Regressionsmodelle vorgestellt, die Fahrzeuginsassen, Fahrzeug und Aufprallmerkmale mit der Wahrscheinlichkeit schwerer Verletzungen abgleichen und dabei den Maximum Abbreviated Injury Scale Level (MAIS) heranziehen. Die Genauigkeit der vorgeschlagenen Modelle wurde anhand des National Automotive Sampling System/Crashworthiness Data System (NASS/CDS) und mit Hilfe der Falldaten der Crash Injury Research and Engineering Network (CIREN) bewertet. Kumulativ lag die positive Prognoserate von Modellen, die die Wahrscheinlichkeit von MAIS3 und schwereren Verletzungen feststellte, bei 74,2 Prozent. Die Art des Aufpralls hat einen deutlichen Einfluss auf das Verletzungsrisiko. Bei Aufprall mit 30 mph DeltaV liegt das Risiko fuer eine MAIS3+ Verletzung fuer jede Art bei 38,9 Prozent, 83,8 Prozent, 47,8 Prozent und 19,9 Prozent bei Unfaellen mit Frontal-, Frontseiten, Heckseiten und Heckaufprall respektive. Zusaetzlich wurden zu Delta V eine Reihe von Aufprallvariablen ermittelt, die bei der genauen Vorhersage der Wahrscheinlichkeit einer MAIS3+-Verletzung helfen. Diese Variablen beinhalteten das Alter des Fahrzeuginsassen, teilweises Herausschleudern, Verwendung des Sicherheitsgurtes, Intrusion in der Naehe des Fahrzeuginsassen und Aufprall auf ein schmales Objekt. Bei Frontalaufprall beinhalteten die zusaetzlichen Variablen das Ansprechen des Airbags, Lenkradverformung und Unfaelle mit Mehrfachaufprall. Das quantitative Verhaeltnis zwischen jeder dieser Crash-Variablen und dem Verletzungsrisiko wurde ermittelt und durch Regressionsanalyse auf der Basis von NASS/CDS- und CIREN-Daten validiert. ABSTRACT IN ENGLISH: The advent of Automatic Crash Notification Systems (ACN) offers the possibility of immediately locating crashes and of determining the crash characteristics by analyzing the data transmitted from the vehicle. A challenge to EMS decision makers is to identify those crashes with serious injuries and deploy the appropriate rescue and treatment capabilities. The objective of this paper is to determine the crash characteristics that increase the risk of serious injury. Within this paper, regression models are presented which relate occupant, vehicle and impact characteristics to the probability of serious injury using the Maximum Abbreviated Injury Scale Level (MAIS). The accuracy of proposed models were evaluated using National Automotive Sampling System/ Crashworthiness Data System (NASS/CDS) and Crash Injury Research and Engineering Network (CIREN) case data. Cumulatively, the positive prediction rate of models identifying the likelihood of MAIS3 and higher injuries was 74.2 percent. Crash mode has a significant influence of injury risk. For crashes with 30 mph deltaV, the risk of MAIS3+ injury for each mode is 38.9 percent, 83.8 percent, 47.8 percent and 19.9 percent for frontal, near side, far side and rear impact crashes, respectively. In addition to deltaV, a number of crash variables were identified that assist in the accurate prediction of the probability of MAIS 3+ injury. These variables include occupant age, partial ejection, safety belt usage, intrusion near the occupant, and crashes with a narrow object. For frontal crashes, added crash variables include air bag deployment, steering wheel deformation, and multiple impact crashes. The quantitative relationship between each of these crash variables and injury risk has been determined and validated by regression analysis based on NASS/CDS and CIREN data. (A) Beitrag zum Themenbereich Pre-/Post Crash II der Tagung Innovativer Kfz-Insassen- und Partnerschutz Fahrzeugsicherheit 2010 - der VDI-Gesellschaft Fahrzeug- und Verkehrstechnik, Berlin, 20. und 21. November 2003. Siehe auch Gesamtaufnahme der Tagung, ITRD-Nummer D352586.

NASS 1998-2000 was queried to determine the frequency of serious injuries in multiple impact crashes and the distribution of injuries by crash sequence. The data set included all passenger cars and light trucks in NASS/CDS. The results showed that 42% of the MAIS 3+ injuries were in crashes that involved more than one harmful event. Approximately 24% of the MAIS 3+ injuries involved two harmful events, and 18% involved 3 or more harmful events. For multiple crashes with serious injuries, the most frequent initial impact direction was frontal (50%) followed by side (44.9%). The most frequent second impact was side (48.4%) followed by frontal (27.6%). The most harmful sequences were side-side (27.7%), front-side (15.8%) and front-front (14.9). The data suggests the need for further investigation and classification complex multiple impact crashes to aid in the in the design of safety systems.

A finite element model of the Hybrid III crash test dummy is developed for computer crash simulations. A description of the major components of the Hybrid III dummy and their finite element representations are given. The results of testing procedures required by the Code of Federal Regulations on the physical dummy are also presented and compared with results obtained from the computer model. The reasonable accuracy obtained from the model makes it useful for crashworthiness simulations when combined with other vehicle and restraint system models.

In this paper, the comparative stiffness of vehicle side and frontal structures is determined by available static test and crash test data. NHTSA has conducted a series of staged crash tests where a Honda Accord is impacted by different bullet vehicles at a closing velocity of 32.5 mph. These staged front-to-side crash tests are examined in order to assess the extent of damage to both the bullet and struck vehicle. The load cell barrier data for the bullet vehicles used in NHTSA's vehicle-to-vehicle front-to-side crash tests are examined to determine the geometric and stiffness properties of the frontal structures as measured in the NCAP tests. The geometric and stiffness measurements during the early stages of frontal crush are most influential in front-to-side crashes. The barrier data provides useful stiffness information. However, the number of rows of load cells may be insufficient to provide geometric information. The bumper height and width appears to be a more accurate measurement of load path that initially contacts the side of the struck vehicle. The bumper height for 20 representative vehicles was measured and summarized in the paper.

The National Highway Traffic Safety Administration (NHTSA) routinely measures the force exerted on the barrier in crash tests. Thirty-six load cells on the face of the rigid barrier measure the force. This study examines the load cell barrier data collected during recent years of NCAP testing to determine how it can be used to assess vehicle compatibility in vehicle-to-vehicle front-to-side crashes. The height of the center-of-force measured by the columns of load cells is proposed as a metric for quantitatively describing the geometric properties of the crash forces. For front-to-side crashes, the geometric and stiffness properties of frontal structures during the early stages of crush are applicable. Consequently, geometric and stiffness measurements at a crush of 125 mm are presented in this paper. This paper shows the range of the compatibility and stiffness parameters measured on cars, pickups, vans, and multi-purpose vehicles.

The URGENCY algorithm uses vehicle crash sensor data in Automatic Crash Notification (ACN) systems to assist in instantly identifying crashes that are most likely to have time critical injuries. The algorithm also provides the capability of improving injury identification, using data obtained from the scene. The prime purpose of the algorithm is to automatically provide emergency medical responders with objective information on crash severity to assist in detecting the approximately 1% of crashes with serious injuries needing the most urgent medical care. The algorithm calculates the risk of a MAIS 3+ injury being present in the crashed vehicle, instantly at the time of the crash. The prediction can be subsequently updated as more information becomes available. The algorithm was based on a multiple regression analysis using data from the National Accident Sampling System/Crashworthiness Data System, (NASS/CDS) years 1988-95. In this paper, the accuracy of the algorithm was evaluated for near side crashes by applying it retrospectively to the population of injured occupants in NASS 1997-2000. URGENCY was applied to the population of injured occupants in near side crashes. Using an injury risk criterion of 50%, URGENCY identified 69% of the crashes with MAIS 3+ injuries. By lowering injury risk criterion to 40%, URGENCY identified 78% of the crashes with MAIS 3+ injuries. Vehicle side intrusion was found to be a highly influential variable. By changing side intrusion from a binary to a continuous variable, the correctly identified crashes increased from 69% to 81%. Examination of the consequence of missing variables found that unknown values of occupant height and weight had a negligible effect on the ability to capture the MAIS 3+ injured. However, lack of knowledge of these variables did increase the magnitude of the false positives.

The URGENCY algorithm uses data from on board crash recorders to assist in identifying crashes that are most likely to have time critical injuries. The algorithm calculates the risk of an AIS 3+ injury being present in the crashed vehicle. The basis for the algorithm is a multiple regression analysis using data from the National Accident Sampling System/Crashworthiness Data System, years 1988-95. The algorithm was deployed and evaluated by a NHTSA during a Automatic Crash Notification Field Operational Test in Erie County, New York. Additional validation was conducted using case data from the William Lehman Injury Research Center. The URGENCY algorithm offers promise of augmenting the Automatic Crash Notification System by identifying those crashes with seriously injured occupants.

The National Highway Traffic Safety Administration (NHTSA) routinely measures the force exerted on the barrier in NCAP crash tests. Thirty-six load cells on the face of the rigid barrier measure the distribution forces. This study examines the load cell barrier data collected during recent years of NCAP testing to determine how it can be used to assess vehicle compatibility in vehicle-to- vehicle front-to-side crashes. The height of the center-of-force measured by the columns of load cells is proposed as a metric for quantitatively describing the geometric properties of the crash forces in the vertical direction. Another proposed metrics is the load distribution when the total force reaches 50kN. Load distribution factors are proposed, based on the number and location of load cells that carry the majority of the barrier crash forces. For front-to-side crashes, the geometric and stiffness properties of frontal structures during the early stages of crush are applicable. Consequently, compatibility measurements at a frontal crush of 125 mm or less are presented in this paper. This paper shows the values for the proposed geometric compatibility parameters measured on twenty-three late model passenger cars.