Rotary Task Kinematic and Kinetic Analysis of the Lower Extremity After Total Knee Arthroplasty: Stability and Strategies

Abstract

Knee instability has been identified as a contributor to the need for total knee replacement (TKR) revision surgery. As TKR revisions increase, the importance of understanding the mechanisms of knee instability becomes a priority to the surgeons, rehabilitation specialists and designers in industry. Objective measurements of knee instability have been recorded by instruments such as the knee arthrometer or with the leg encased in a boot driven by a dial test tool for rotation. Both of these tests are performed in the open-chain position and the knee arthrometer only measures laxity in the sagittal plane. Despite capturing total range of motion (ROM) excursion or laxity measurement, these tests do not reproduce what is described as the working definition for knee instability. Functional knee instability, defined as the subjective report of the knee `buckling' or `giving way', correlates to a dynamic event rather than an open-chain occurrence. Therefore, a weight-bearing, dynamic test would be required in order to identify knee instability. In order to test dynamic knee instability, rotation should be included to add a dimension of out-of-plane movement that correlates to pivoting, change of direction or rotating with bending and extending or reaching while loading and unloading the knee. Twenty-eight individuals (10 TKR, 12 Healthy, 6 Unstable) performed two tasks, Stair Task and Target Touch Task (TTT), and kinetic and kinematic data were captured by motion analysis and force platforms. The Stair Task included a pivot turn after stair descent and the TTT required a series of button pushes while squatting and extending, with rotation or crossing mid-line. Variables from both tasks were imputed to a Principle Components Analysis (PCA) in order to identify differences of performance across the groups. In the Stair Task, the TKR group had less Ground Reaction Force (GRF) on initial impact after stair descent compared to the unaffected leg (p=0.021) and during mid-stance compared to the healthy group (p=0.049). The affected stance leg had less knee flexion during mid-stance in both the straight trial (p=0.002) and turn (p=0.010). Similarly, the TKR individuals maintained a more extended knee position for both affected (14.7°) and unaffected (17.8&#730;) during the TTT when compared to the healthy (25.5&#730;) when approaching mid-line to transfer weight in order to push the low button when squatting (p<.05). Further, a difference was noted between the TKR legs during the motion to push the high button. At 90% of this cycle of movement, the unaffected knee was more flexed when compared to the affected knee (p<.05). A large variation of loading during mid-cycle was noted with the TKR group suggesting difficulty with stabilization during the transition of side to side motion with rotation. The PCA model was utilized to compile 29 variables selected from the two tasks in order to identify related performance variables for all groups. Interestingly, no dominant concepts were derived from the analysis; rather, the percentage of variables loaded up to 12 PCs to achieve 80% of the explained variation. Temporal variables dominated PC1 when comparing Unstable to Healthy as well as TKR to Healthy. Both the Unstable and TKR groups performed the tasks slower than the Healthy group. While delay or slowed response may be related to a protective response to avoid fall, the hesitation may be associated with information response from the Central Nervous System (CNS). In addition, compensatory movements from adjacent joints were noted in the TKR group such as utilizing torso movements to achieve the TTT rather than bending the knee. Altered mechanics that may be a result of pre-operative habits or a result of surgery, could lead to injuries to other joints. Rehabilitation interventions should address symmetrical movements with weight transfer and rotation in order to avoid future injuries.