Total knee arthroplasty (TKA) is a common surgery to relieve knee pain and increase range of motion due to osteoarthritis (OA) in older patients. Minimally invasive, computer navigated techniques are gaining popularity for knee replacement surgery. These techniques may have potential to provide better functional outcomes over a shorter period of time. Little data exists comparing the early functional recovery of patients following total knee replacement surgery performed using various common approaches. This study compares the functional gait of patients two months after surgery performed using one of four common approaches to determine if differences exist in the immediate recovery. This knowledge will aid surgeons determine the best approach to use when performing surgery.

This study was approved by the appropriate Institutional Review Board. Subjects volunteered to participate in the study and signed informed consent prior to testing. Subjects were excluded if the had significant diseases of the other joints of the lower extremity or a diagnosed disorder with gait disturbance. Patients were randomly assigned to receive unilateral primary TKA using standard parapatellar, mini-parapatellar, mini-midvastus, or mini-subvastus approaches. All patients received the same preoperative, perioperative, and postoperative critical pathways and standard orders. All incisions were five inches and all patients and examiners blinded to type of approach. Surgery was performed by one of two fellowship trained orthopedic surgeons. Patients visited the gait laboratory two months after receiving TKA. Motion data was captured using a ten-camera motion capture system (Motion Analysis Corp., Santa Rosa, CA). Three-dimensional force data was recorded using four floor embedded force platforms (AMTI Inc., Watertown, MA). Patients were asked to walk at a self selected speed along a 6.5 metre walkway. A minimum of five good foot strikes for each limb were recorded. Data were collected using EVaRT 5 software (Motion Analysis Corp., Santa Rosa, CA) and analyzed using OrthoTrak 6.2.8 (Motion Analysis Corp., Santa Rosa, CA) and MatLab software (The Mathworks Inc., Natick, MA). Statistical analysis was performed using SPSS 14.0 software (SPSS Inc., Chicago, Il) (α = 0.05).

Fifty-two patients (72 ± 6 years) volunteered to participate in the study. The approaches used were: standard parapatellar – 12; mini-parapatellar – 12; mini-midvastus – 14; mini-subvastus – 14. Statistical analysis found no significant differences in any of the variables measured except minimum knee flexion angle during stance (p=.046). The variables measured included the maximum and minimum injured lower limb joint angles in all planes during both stance and swing phase of gait. Also measured were the maximum joint moments in all planes during stance and hip, knee, and ankle powers.

Patients who received TKA using the mini-subvastus approach had greater knee extension through much of the single stance phase of the gait cycle which contributed to a lower (but not significant) peak knee flexion moment. These patients also had the highest ground reaction shear forces with higher ankle power absorption at foot strike and generation at push off. Mini-subvastus patients used a higher cadence to walk with a greater velocity then patients who received surgery using the other approaches.

The results of this study show only minor differences in gait between patients who have received surgery using the different approaches. The limited numbers of patients in the study and the large variation in outcomes so soon after surgery mean that in most cases the differences that were measured do not reach significant level. This study shows that the surgical approach used to implant the device has no apparent effect on the ability of the person to ambulate following surgery, however further study with increased numbers of patients and observation over a longer period of time will allow a stronger conclusion. The knowledge gained from this and future studies will enable surgeons to make decisions on type of approach based on factors other than expectations of functional outcome.

Previous studies (Chen et al., 2003; Kaufmann et al., 2001) have shown that persons with osteoarthritis (OA) walk more slowly with lower cadence, have lower peak ground reaction forces and load their injured limb at a lower rate than healthy age matched subjects. However, another study (Mündermann et al., 2005) found that patients with severe bilateral OA loaded their knee joint at a higher rate. They also found these patients had higher knee adduction moments and lower hip adduction moments. It has been reported (McGibbon and Krebs 2002) that when subjects with knee OA are required to walk at the same speed as healthy subjects they generate more power at the hip joint to help overcome reduced knee power and aid in the advancement of the leg prior to the swing phase of the gait cycle. Myles et al. (2002) reported that patients with knee OA have reduced knee range of motion during walking. This paper presents detailed kinematic and kinetic data collected on a large group of patients with advanced knee osteoarthritis to show the differences in the gait of these patients just prior to surgery compared with age-matched control group.

This study was approved by the Sun Health Institutional Review Board. Subjects volunteered to participate in the study and signed informed consent prior to testing. Subjects were excluded if the had significant diseases of the other joints of the lower extremity or a diagnosed disorder with gait disturbance. Motion data was captured using a ten-camera motion capture system (Motion Analysis Corp., Santa Rosa, CA). Three-dimensional force data was recorded using four floor embedded force platforms (AMTI Inc., Watertown, MA). Patients were asked to walk at a self selected speed along a 6.5 meter walkway. A minimum of five good foot strikes for each limb were recorded. Data were collected using EVaRT 5 software (Motion Analysis Corp., Santa Rosa, CA) and analyzed using OrthoTrak 6.2.8 (Motion Analysis Corp., Santa Rosa, CA) and MatLab software (The Mathworks Inc., Natick, MA). Statistical analysis was performed using SPSS 14.0 software (SPSS Inc., Chicago, Il) (α = 0.05).

Eighty-six patients (71 ± 7 years) along with sixty-four control subjects (65± 10 years) volunteered to participate in the study. All measured temporal and spatial parameters showed significant differences between the OA patients and the control group. The OA patients were found to walk at a significantly lower velocity (p< .01) and cadence (p< .01) using a wider step width (p< .01) than the control subjects. Patients had their injured knee significantly more flexed at foot strike (p< .01) but flexed the knee significantly less during swing (p< .01) when compared to the control group. Patients had significantly higher knee flexion angles as well as hip flexion and abduction angles during stance. Knee varus angles were significantly higher for the OA patients during stance (p< .01) but not during swing when compared to the control group.

Significant increases in pelvic tilt and pelvic obliquity were measured during the stance phase. Hip abduction angles during stance were significantly lower for the OA group. Patients generated significantly lower vertical ground reaction forces during stance (p< .01) while sagittal plane kinetic analysis showed significantly lower external knee flexion moments (p< .01) and knee power generation (p< .01) during this phase of the gait cycle. Analysis of frontal plane angles showed OA patients had a significantly higher maximum knee varus angle during stance as well as generating a higher external knee varus moment (p=.03) during this phase of the gait cycle.

Changes in gait measured in this study support and enhance findings from previous studies. OA patients appeared to walk with a more crouched posture with higher knee and hip flexion angles through mid stance. This along with lower velocity and cadence and a larger step width would indicate a desire for more stability while walking. Patients also flexed their knees more at foot strike in an attempt to absorb the forces generated during weight acceptance. While knee flexion angles measured for the OA group were similar to the control subjects during the initial period of stance, the OA patients did not extend their knees as much during mid stance indicating a desire to reduce the angular rotation of the knee while in single support. Changes measured in frontal plane angles of the hip and pelvis may be an attempt to compensate for the different angles generated by the knee during stance. The differences in hip and knee angles measured during stance for patients and controls allowed patients to have reduced peak external knee flexion moments during initial stance but a higher knee flexion moment at mid stance. The reduction in knee angular change during stance and the reduced cadence meant power absorption during early and late stance and generation during mid stance was much lower for the OA patients than the control group. All the changes noted appear to be designed to limit the movement of the knee joint while loaded and reduce the peak loads in an effort to reduce pain at the affected joint while at the same time increase stability during gait. These data show the differences that exist between the gait patterns of patients with advanced osteoarthritis and healthy age-matched persons and highlight the changes that are necessary following knee replacement surgery and rehabilitation to return the gait of these patients to normal.