Image-based computer-assisted total knee arthroplasty leads to lower variability in coronal alignment

In a prospective randomized controlled trial, one group of 50 patients had total knee arthroplasty using conventional instruments; another group of 50 patients had total knee arthroplasty using a fluoroscopy-based computer navigation system. The variability of postoperative alignment of the lower limb in the coronal and saggital plane, early clinical outcome scores, and morbidity of the procedure were compared. In addition, the image-based computer navigation system was used to assess the following presumptions: the accuracy of the calculation of the kinematic center of rotation of the hip and the reliability of full-leg standing xrays in determining overall coronal alignment of the lower limb. Variability in the coronal plane was significantly reduced in the computer-assisted surgery group compared with the conventional group (p < 0.0001). Early clinical outcome and complication rates were similar for both groups. The correlation between full-leg standing xrays and the computer navigation system for determining the mechanical alignment of the lower limb in the coronal plane was high (r = 0.987). The maximum deviation between the calculated kinematic center of rotation of the hip and the fluoroscopically determined anatomic center of the hip was 5 mm (mean, 1.6 mm). An image based computer-assisted system can be safely used and leads to a consistent and perfect postoperative alignment in the coronal plane.