Engineering efforts directed at better occupant safety require a thorough understanding of available epidemiologic data. Epidemiologic studies using clinical as well as accident information facilitates the prioritization of biomechanics research so that controlled laboratory experimentation and/or analytical models can be advanced. This information has also value in dictating levels and types of injury that are critical to the development of anthropomorphic test devices used in crash environments.

In this paper, motor vehicle accident related (excluding pedestrians, bicyclists, and motorcyclists) epidemiologic data were obtained from clinical and computerized accident (National Accident Sampling System-NASS) files. Clinical data were gathered from patients admitted to the Medical College of Wisconsin Affiliated Hospitals, and fatalities occurring in Milwaukee County, State of Wisconsin. NASS database with specific focus on spinal injuries of motor vehicle occupants was also used.

The purpose of the clinical study was to determine the most commonly injured anatomic levels of the cervical spine, to classify these injuries based on an impairment scale, to determine the mechanism of injury at each spinal level, to evaluate the differences, if any, between the patient (survivors) and fatality data with respect to the location and mechanism of injury, and to compare this data obtained from a localized population with literature results and national samples. The purpose of the NASS study was to estimate the annual occurrence of cervical and thoracolumbar spinal injuries, to elicit the gross anatomical distribution of these injuries (eg., cord vs. skeletal);​ to tabulate the distribution of these injuries by crash type (i.e., frontal, side, rear, and rollover);​ to estimate the risk of these injuries as a function of crash type, and restraint use, and finally, to briefly explore the association of craniofacial trauma with spinal injury. Results indicated that while injuries to the cervical column are complex and may occur at any spinal level, they concentrate statistically in two primary zones, at the craniocervical junction for fatal victims, and in the lower cervical spine for survivors. The majority of paralyzing injuries (survivors), both complete and incomplete quadriplegia, were produced by flexion-compression loading, with disruption of the posterior elements and compressive fractures of the vertebral bodies. However, rotation appeared to be an important component in non-paralyzing cervical fractures. A strong association was found to exist between craniofacial and cervical spine trauma. While only cervical spine injuries were evaluated from the clinical database, analysis of NASS files included both cervical and thoracolumbar spine. Results indicated that the thoracolumbar injuries at AIS 3+ level were primarily bone related (96%). In contrast, neck trauma at this severity was 20% cord and 65% bone related emphasizing the relative importance of protecting the spinal cord in this region. For the cervical spine, within the limitations in the ensemble size, introduction of restraints appeared to significantly reduce the incidence of injury at the AIS 3+ level, while increasing at the AIS 1 level. However, little change was observed at the AIS 3+ level for the thoracolumbar spine.