Back pain is one of the most costly and pervasive medical problems in the United States. Unfortunately, despite investing billions of dollars over the past 40 years searching for a mechanical means of replacing degenerative intervertebral discs, no form of arthroplasty has proven to significantly improve patient care. This absence stems from complexities in intervertebral joint anatomy and motion. For instance, the disc itself is not a free-standing joint, but works in conjunction with the two adjacent facet joints. Current total disc replacements (TDRs) disregard this interaction, which possibly explains their suboptimal clinical performance but little research is available to support this theory. Therefore the aim of this dissertation is to first investigate the consequences of TDR on the adjacent facet joints. Afterwards, current measures of spinal motion are improved to quantifiably compare surgical instrumentation to the spine's natural motion patterns.

In vitro biomechanical experiments demonstrated that facet forces increase after TDR implantation, most notably when the device is positioned anteriorly. During sagittal plane movement, TDRs displayed an increase in facet forces as specimens rotated into flexion, opposing the trend of the intact disc. These results indicate that over time TDRs may lead to increased facet arthritis and/or degeneration.

Two metrics capable of conveying the complex motion of the spine, the stiffness matrix and the helical axis of motion (HAM), were analyzed as possible tools for characterizing spinal motion before and after surgical alterations. By creating a method for determining the stiffness matrix which accommodates finite rotations, non-conservative forces, and energy loss, and by distilling this 36-component matrix into a single scalar, a measure of 6-D stiffness for comparative purposes was developed. This scalar proved useful in determining variations in stiffness post TDR implantation. This was followed by a thorough investigation of calculating the HAM which revealed that commonly used algorithms are too erroneous to detect changes after implanting a device. Thus, an alternate method was developed along with a set of best practices in minimizing HAM error. The combined result provides a general guideline for calculating sufficiently accurate HAM for characterizing the kinematics of motion preserving devices.