The process of relating a specific event to a specific injury is often complicated when the facts surrounding the event are in dispute. Expert testimony is often used to evaluate the recollections in many litigation matters. Specifically, some experts will assert that their analysis does not support the claimed injury mechanism, which can be interpreted to imply that an injury is unlikely. The converse is also true – experts will sometimes state that the biomechanical factors are consistent with the injury mechanism, based on biomechanical data.
There are many steps that occur in developing an expert opinion on the mechanism of a reported injury, such as accident reconstruction, analysis of the interaction of a human body with a structure, and an estimate of loads on a specific body part. These steps are based on known approximations for the calculated loads on the body. From these analyses and the estimated circumstances of an event, there are generally three overall approaches that lead to opinions on injury, based on the biomechanical data available in professional and scientific literature.
Activities of daily living
One approach is to compare the loads or accelerations in the event at issue with the loads that occur in the activities of daily living, such as getting into or out of a chair, climbing stairs, coughing, or sneezing. The estimates for those activities are available from tests where volunteers perform the activities and measurements are made of accelerations, reaction forces, or movements during the activity. In fact, there is some data, from many years ago, where volunteers sat in automobiles that were struck from behind at low speeds, and their reaction to the impact, such as transient neck pain or headaches, were recorded along with the mechanical data.
The key issue with these tests is the state of the volunteers. They are typically healthy, young, and “normal” for two primary reasons. First, for obvious legal reasons, there was a very low expectation of injury. These tests are typically conducted in universities, with an Institutional review board overseeing the tests. Second, the reaction of a person with a pre-existing condition is going to be very different from the reaction of a different person with a different pre-existing condition. To produce a reproducible result and to be relevant to a large population, a scientific study would narrow its focus to normal people. Follow-on studies could be done that address the effects of specific abnormalities such as arthritis or disc degeneration, but they are rare, especially since those conditions have a wide degree of severity. So young, healthy volunteers typically generate this data.
The second common way that biomechanical data is used to develop an injury likelihood is using a general indicator to predict an injury, without a specific mechanism of injury involved. That is, the overall mechanism (an impact or a high force) is the cause of the injury, but the specific failure process (cracking at a specific part of a bone, for example) is not specified.
The best example of that approach is the use of a “HIC” (head injury criterion) score to indicate concussion or TBI. Despite decades of research, the specific mechanisms for traumatic brain injury are unclear, even as knowledge concerning the symptoms, clinical course, and treatments continues to improve. The general mechanisms are well-known – head impacts, particularly those including rotational motion, are a cause of a wide array of symptoms, which have been well-documented. However, at this time there are significant uncertainties and incomplete knowledge concerning the connection between a documented physical injury (on a medical scan, for instance) and a functional deficit. This leaves biomechanical analyses without a clear quantity or estimate that can be taken from a mechanical model and used in making a prediction of injury. This area of research is moving ahead, however, and medical imaging techniques are developing that show promise.
As of today, the quantities currently used as indices for TBI prediction are based primarily on intuition and on very general physiological data. For example, the HIC score was initially developed to quantify the impacts that would lead to skull fracture. Smaller values were thought to be connected to the degree of brain injury, and in general decreasing the HIC score for a given impact leads to less severe brain injury, but the correlation is not based on any specific injury to any specific part of the brain. Similarly, mechanical models of the head in impact have been developed that calculate the overall volume of brain tissue which has a peak strain larger than a set value, but the specific part of the brain that has a high strain is not accounted for, because of a lack of knowledge for which part of the brain would cause functional deficits if it was significantly stretched.
There are many similar head injury indicators and many detailed models of head injury, but the key question to ask is whether these indicators can classify or predict whether a person will develop mild, moderate, or severe TBI. In most cases, these indicators are not able to offer predictive data. This could change with improved medical knowledge of the injury process in TBI.
Other injury predictions are similarly based on overall indicators without developing a specific mechanism of injury. The criteria for ankle fracture or spine fracture, in many cases, uses overall load as the indicator, without specifying the location or mechanism of fracture. In the ankle, for instance, an array of fractures to the Calcaneus, Talus, or the lower Tibia are observed, but the indicator for fracture is often the overall load on the tibial shaft. To be fair, if the objective is to avoid any injury at all, then this approach will cover the significant injuries, but the approach does not address the cause of a specific injury.
The third and most detailed method that is used to predict injury from a biomechanical analysis is based on a calculated stress due to an incident, with the mechanism of injury clearly stated. The calculated stresses are compared to material tests that determine the failure stress. The typical example for this type of case is rib fracture, where the mechanism of failure is relatively clear (tensile cracks perpendicular to the axis of the rib).
At first glance this seems like the clearest and most accurate way to determine failure, similar to the stress analyses used in most other types of engineering. Large-scale finite element structural modeling generates an estimate for the stress level at a specific location, and the material will yield according to criteria from independent material tests. However, the variability in human anatomy and the sensitivity of stress levels to small shape features (“stress risers”) makes this approach very difficult to use. Simply put, if a stress-based failure criterion is being used, then the prediction is only as accurate as the calculated stresses.
Calculating a stress level in a human bone requires high fidelity information on the anatomy of the bone and the material state of the bone. As is well-known, most bones have an outer cortex that varies in thickness from up to 5 mm in the mid-shaft of the femur to less than 1/10th of a millimeter around the joints. Underneath the cortex, near the joints, cancellous bone (also called trabecular bone or spongy bone) supports the joint surfaces. This cancellous bone varies in stiffness and strength widely, and it varies from person to person as well.
The fundamental problem with a stress-based criterion for tissue injury lies in the nature of the stress distributions within the bone. The stresses will vary substantially due to small variations in geometry and material properties, and the local strength of the bone will vary as well. Further, the small variations in geometry (due to circulatory access to the interior of the bone, for example) are not well-established or controlled.
One approach to dealing with the wide variability of human bone shape has been to run many analyses that incorporate the variability of the bony structure into the calculations, and to develop probabilistic estimates for failure. A statistically variable estimate of peak local stress replaces the overall indicator (such as force or moment on the bone) to provide an injury prediction.
Where does the Data come from?
Each of the methods summarized above requires experimental data. As discussed above, tests and analyses of experimental specimens require the specimens to be relevant to the population in which the injury estimate is being developed. Specimens with “normal” anatomy are favored for consistency in the data being produced, and so that the injury criteria are relevant to a large population.
To evaluate whether an injury criterion is relevant to a case involving an injury, the relevance of the group of specimens within the study that produced the injury criterion is a key question. If the study population is very different from the plaintiff or if the plaintiff in the case has significant pre-existing conditions, then concerns should be raised. In addition, medical imaging results from the plaintiff at the time of the injury can reveal some prior conditions. Some conditions are the result of long-term changes, while some others are clearly short-term reactions. Abnormal intervertebral discs, evidence of osteoarthritis, or other conditions are often the result of long-term changes, and they would have existed prior to a recent event. By contrast, inflammation, hematomas, and similar findings are typically short-term reactions. The time scales for these conditions are not the province of a biomechanical expert, however – any opinion on whether an imaging finding was short or long-term would need to be made by the appropriate medical professional. A Biomechanical expert could evaluate whether the changes due to pre-existing conditions are significant enough to invalidate an opinion based on data from the injury studies.
In conclusion, injury data relied upon by biomechanical experts should get a close look to evaluate relevance if a plaintiff has evidence of pre-existing conditions. In addition, the strengths and weaknesses of the typical methods for developing estimates for injury likelihood should be kept in mind to assess whether the predictions and the data that they are based on are reliable.
Timothy P. Harrigan, ScD, MBA, PE, is a Mechanics / Biomechanics and Modelling and Simulation Engineer with over 35 years of experience. He has spent 15 years in Orthopedic departments and the rest of his career in Biomedical Engineering. Dr. Harrigan has done accident reconstruction through analytical and computational models, and extensively used FEA, CFD and, Fluid-structure models. He works on both Design and Manufacturing Defects in Medical Devices, including total joint replacements, IV tubing, insulin pumps, IV fertilization, and heart valves. With 3 US Patents, and having passed the patent bar, Dr. Harrigan works effectively with IP professionals on expert reports Injury Biomechanics and Intellectual Property.
©Copyright - All Rights Reserved
DO NOT REPRODUCE WITHOUT WRITTEN PERMISSION BY AUTHOR.