WHIPLASH

Mathematical studies Over the years a number of mathematical models have been developed to help us better understand some of the features of whiplash injury. These studies continue today. A finite element model (FEM) of the human neck (with geometry modeled from the MRI scan of a 50th percentile male spine) has been validated against cadaver tests of 15 mph rear impact crashes from Duke University. Both solid (bone) and soft elements (nucleus and anulus using linear viscoelastic material properties) were modeled based on existing literature. Collision with a pre-deployed airbag was also modeled. The model correlated well with the experimental data in the rear impact crashes and clearly demonstrated the head lag (retraction) seen in human volunteer crash tests. The model also correlated well with airbag tests. During the rear impact tests (with FEM simulations), upward motion of T1 was noted with compression of the spine. This is due either to the ramping up of the torso or straightening of the thoracic spine. Compression led to loosening of the ligaments of the neck at about 40 msec after impact. During compression, the neck becomes less stiff, diminishing its resistance to shear forces. Up to 27% capsular stretch was observed. With FEA, the more complex we make our models, the more processing time is required to solve the simulations. In one of the most sophisticated FEA head/neck models of today, an IBM supercomputer, with five processors, requires 60 hours to process only 50 msec of data.
A multibody model developed at TNO Netherlands is the Mathematical Dynamic Model (MADYMO). Research on MADYMO is ongoing, although there do not appear to be any strong human subject validations for rear impact simulations. Brain injuries are also modeled using mathematical models.
Animal studies Although such work is done less frequently these days, from the 1960s to the 1980s several researchers experimented with primates in whiplash crash simulations. Much was learned concerning the types of soft tissue lesions that could be produced, most of which were not visible using conventional x-ray techniques. Researchers have measured the subcortical EEG in rhesus monkeys exposed to simulated whiplash trauma. They found abnormal hippocampal spiking and subclinical epilepsy-an interesting finding in view of the association between memory and the hippocampus.
Researchers have more recently subjected pigs to controlled whiplash experiments, measuring pressure changes within the spinal canal which result from changes in canal volume as the neck moves in extension and flexion. The head angular accelerations and displacements were consistent with a moderate to moderate-to-severe CAD injury (peak head acceleration of ~25 g; peak displacement of ~75 deg.). None of the animals displayed any obvious neurological abnormality afterward, but minimal capsular bleeding in the cervical ganglia was discovered. Using Evans dye, they determined that many nerve cells within the spinal ganglia (mostly from C4-C7) had lost their normal blood-nerve barrier and conjectured that these changes could be sufficient to cause a similar loss and rebuilding of the afferent synaptic connections within the laminae of the posterior horn of the cord, and that this could contribute to the symptoms of whiplash in patients weeks after trauma. These experiments set the stage for the development of the Neck Injury Criterion (NIC). Note: The Spine Research Institute of San Diego is not engaged in animal research of any kind.

Cadaver studies There is a great deal of research currently available utilizing cadavers or, as they are called in this field of research, post mortem human subjects (PMHS), or even less sympathetically, post mortem test objects (PMTO). The value of using PMHS is that there is no risk to human volunteers. Moreover, unlike human volunteers, we can dissect the PMHS to identify what types of injuries might have occurred during testing. We can also attach accelerometers and other instruments, as well as photoreflective targets directly to the subjects which can provide information not attainable with live human subjects. There are, of course, a number of drawbacks and limitations as well. Nevertheless, a good deal of our current knowledge in this field was initially discovered using this kind of testing which might involve whole specimens, isolated spinal segments, or even isolated facet joints alone.

ATDs Anthropometric test devices (ATD), a.k.a. crash test dummies, have been used for many years as surrogates or stand-ins for humans in tests that are deemed too dangerous for human test subjects. The most familiar of these ATDs to most Americans is the Hybrid III dummy which is currently the designated model used in FMVSS crash tests. While it does serve as a useful surrogate in these higher speed (30-35 mph) frontal crash tests, it does not have sufficient neck flexibility or compliance for use in low speed rear impact crash tests-it is said to lack biofidelity. For example, because the Hybrid III does not have an articulated thoracic spine, it cannot experience the flattening of the kyphotic thoracic curve that results in spinal compression and upward motion seen in human volunteers. Its cervical spine is also too stiff to simulate a relaxed human cervical spine. Thus, there has been a need for a biofidelic rear impact dummy (RID) ATD to use in the development of more effective automotive safety systems. In recent years, two such ATDs have been developed. The RID (currently the RID2), was developed at TNO Netherlands and is manufactured by First Technology Safety Systems, of Plymouth, MI. It uses a modification of the Hybrid III torso which is designed for testing the chest loads imparted by safety restraints. It has rib units and a single joint in the thoracic spine and is called the test device for human occupant restraint (THOR). It has been used to evaluate restraint systems. A completely modified neck, which moves in all cardinal planes (flexion-extension, lateral flexion, and rotation), was added to this dummy to make the RID2.
The second RID, the biofidelic rear impact dummy (BioRID II in its current stage of development) was developed at Chalmers University. Unlike the RID2, it has a fully articulated spine from top to bottom, but moves only in the anterior to posterior (flexion-extension) plane. Currently it is manufactured by Robert A. Denton, Inc./Denton ATD, Inc., in Rochester Hills, MI. Both dummies have been extensively tested by institutional members of the European Whiplash Consortium, the International Insurance Whiplash Prevention Group (IIWPG) formed by Allianz Zentrum fur Technik (AZT), the German Insurance Institute for Traffic Engineering (GDV), IIHS, and the Motor Insurance Repair Research Center (MIRRC), Thatcham. Finally, the Spine Research Institute of San Diego conducted full scale, human subject validation tests of both the RID2 (2002) and the BioRID II (2003). Both ATDs have been shown in SRISD tests to have good biofidelity.
Human subject crash test studies Severy et al. conducted the original full scale rear impact crash tests in the 1950s and 1960s and deserves tribute for their pioneering efforts. Despite the fact that the cars they used in their first series of tests were WWII era Plymouths and Hudsons, and the fact that the equipment today is much more sophisticated than what was used back then, and despite the fact that those old cars had relatively rigid bumpers, no head restraints, and no shoulder harnesses, the results they obtained back then are surprisingly similar to those we obtain today in our modern fleet of cars sporting microchip technology. Most notably, Severy's group were the first to show that the acceleration of a volunteer's head in LOSRIC could be up to 2-3 times (or more) higher than that of his vehicle because of the unique and complex occupant-vehicle coupling of this type of crash.
Subsequently, a number of researchers have conducted human subject crash tests-some using seats mounted on hydraulically accelerated sleds, others in full scale, car-to-car configurations. These include the work of West et al., Szabo et al., McConnell et al., Castro et al., Ono et al., Siegmund et al., van den Kroonenberg et al., Davidsson et al., and Croft et al. (see Croft AC, Haneline MT, Freeman MD: Differential occupant kinematics and head linear acceleration between frontal and rear automobile impacts at low speed: evidence for a differential injury risk. International Congress on Whiplash-Associated Disorders, Berne, Switzerland, March 9-10, 28, 2001; and Croft AC, Haneline MT, Freeman MD: Differential Occupant Kinematics and Forces Between Frontal and Rear Automobile Impacts at Low Speed: Evidence for a Differential Injury Risk, International Research Council on the Biomechanics of Impact (IRCOBI), International Conference, September 18-20, 2002, Munich, Germany, 365-366). This research has taught us, collectively, a great deal about how the human subject interacts with the vehicle; knowledge that simply cannot be gained using mathematical models, animal models, or human cadavers.
Some authors have reported that crash test subjects begin to complain of neck pain or headaches in rear impact crashes when crash velocities reached about 5 mph delta V and these comments have gradually been transmogrified into a threshold for human tolerance, albeit through no fault of these authors. There are, unfortunately, several reasons why such extrapolations cannot be made from these tests. In many cases, the crash test subjects were exposed to multiple impacts. It is likely that tolerance to these crashes is diminished with successive tests. More importantly, none of the studies has been designed specifically to determine human tolerances to these forces. Such a study would require relatively large numbers of subjects who would need to be representative of the general population and who would also need to be tested under representative crash conditions. The results of the tests would have to be subjected to statistical analysis in order to determine that the results were not likely to be simply the result of chance. No published tests to date satisfy those scientific requirements. So, while they can tell us much about human kinematics and other important factors, they cannot be used to determine injury thresholds or to develop tolerance corridors.
Researchers recently conducted low speed crash tests and reported that 29% of their subjects developed symptoms in tests of only 2.5 mph delta V, providing compelling evidence against the popular 5 mph delta V threshold theory. Moreover, in a large German study in which real world crashes were reconstructed, the authors reported that of the rear impact crashes investigated, in 42% the crash speed was below 6.2 mph delta V. (For a more in-depth explanation of the limitations of attempting to establishing injury thresholds using this literature, see Freeman MD, Croft AC, Rossignol AM, Weaver DS, Reiser M: A review and methodologic critique of the literature refuting whiplash syndrome. Spine 24(1):86-96, 1999.)

Epidemiological and clinical literature In 1995 the Quebec Task Force on Whiplash-Associated Disorders set out to synthesize the existing whiplash literature. After searching the world literature on this topic, they found over 10,000 citations; most of this can be found in the clinical literature (see Spitzer WO, Skovron ML, Salmi LR, Cassidy JD, Duranceau J, Suissa S, Zeiss E: Scientific monograph of the Quebec task force on whiplash-associated disorders: redefining "whiplash" and its management. Spine (Supplement) 20(8S):1S-73S, 1995). Using rigid criteria that excluded more than 99% of that literature from further review severely limited the breadth and, consequently, the validity of that document.