Objective: The purpose of this retrospective study is to compare the efficiency of computer-assisted surgery (CAS) and conventional method (CONV) in TKA using mechanical axis (MA) and component alignment measured on the post-operative radiograph in the same patient by different technique for TKA on both sides of the knee.
Material and Method: Fifty-two TKA in twenty-six patients with primary osteoarthritis of both knees that underwent stage bilateral TKA by computer-assisted surgery one side and conventional method on the other side were inclusion criteria. Digital long-leg weight-bearing radiographs were taken. The mechanical axis (MA), femoral component in coronal plane (FFC), tibial component in coronal plane (FTC), femoral component in sagittal plane (SFC) and tibial component in sagittal plane (STC) were measured and compared.
Results: The MA indicated that computer-assisted surgery (CAS) is significantly improved accuracy compared with conventional method (178.12° + 1.56° and 176.15° + 1.85° respectively, p = 0.00). For FFC alignment, the results showed that CAS group is significantly more accurate than CONV group (88.58° + 1.30° and 87.38° + 2.02° respectively, p = 0.07). CAS group showed less distribution and fewer outliers of data than CONV group. For FTC, SFC and STC alignment, the means of both groups were no difference (p > 0.05). Otherwise, the numbers of outlier CONV group trend toward greater than CAS group (FTC 3.8% and 0%, SFC 30.8% and 0%, respectively). There was no report of change in the navigator group procedure to conventional method during surgery and no perioperative or postoperative complications were noted.
Conclusion: Computer-assisted surgery (CAS) is a safe and useful intraoperative tool for total knee arthroplasty to improve accuracy of mechanical axis, good implant position and reduce number of postoperative implant outlier. Clinical studies will be required for clinical outcome assessment.

Keywords: Computer-assisted surgery, Arthroplasty, Total knee replacement