A Comparison of Outcomes with CAS versus Conventional TKA

    A recent study from New Zealand evaluated outcomes of total knee arthroplasty from the standpoint of whether surgeons routinely used computer-assisted surgery or conventional instrumentation for more than 90% of cases.


    Amy Kaplan, MD, and Ajit Deshmukh, MD


    Roberts TD, Frampton CM, Young SW. Outcomes of computer-assisted surgery compared with conventional instrumentation in 19,221 total knee arthroplasties: results after a mean of 4.5 years of follow-up. J Bone Joint Surg Am. 2020 Jan 22. doi: 10.2106/JBJS.19.00852. [Epub ahead of print]


    Roberts et al retrospectively collected data from the New Zealand Joint Registry (NZJR) to evaluate differences in clinical outcomes between surgeons who routinely performed total knee arthroplasty (TKA) with computer-assisted surgery (CAS) and those who did not routinely perform TKA with CAS. Data was collected for a 12-year period, 2006 through 2018.

    RELATED: Primary TKA Using Computer Navigation – an ICJR Featured Surgery

    Previously reported evidence from the NZJR showed no difference in functional scores and implant survival rates between patients who had CAS and those who did not. In this study, Roberts et al, attempted to eliminate potential bias from surgeons who only used CAS for complex cases and only used it occasionally by analyzing surgeon groups instead of TKA cases. Surgeons who performed more than 50 TKAs recorded by the registry were divided into 2 groups:

    • Those who performed more than 90% of their cases with CAS
    • Those who performed more than 90% of their cases without CAS

    A third category was surgeons who were “mixed,” meaning those that performed from 10% to 90% of their cases with CAS. Data from this mixed group were compared with data from the first 2 groups as a secondary analysis.

    Outcomes evaluated were:

    • Primary outcome: Cumulative revision rates, defined as the total number of revisions divided by the number of component years and multiplied by 100
    • Secondary outcome: Oxford Knee Scores recorded at 6 months, 5 years, and 10 years

    A total of 19,221 TKAs were identified and included in this study:

    • CAS group: 9501 cases performed by 25 surgeons
    • Conventional instrumentation group: 7672 TKAs performed by 22 surgeons
    • Mixed group: 2048 TKAs performed by 8 surgeons

    No difference was observed in:

    • Revision rate per 100 component years when comparing routine CAS surgeons and routine conventional instrumentation surgeons
    • Revision rate per 100 component years in a subgroup analysis of patients less than age 65
    • Oxford Knee Scores between routine CAS surgeons and routine conventional instrumentation surgeons

    Significant decreases were observed in:

    • Oxford Knee Scores at 5 years between the mixed group and the routine conventional instrumentation group (40.19 vs 41.77, respectively; P=.0006)
    • Oxford Knee Scores at 5 years between the mixed group and the routine CAS group (40.19 vs 42.26, respectively; P <0.001)

    Although statistically significant, the difference in Oxford Knee Scores was hypothesized to be clinically inconsequential.

    RELATED: Does Computer-Assisted Surgery Improve Outcomes of TKA?

    In addition, CAS cases had longer operative times. Subgroup analysis revealed that it was surgeons who routinely used CAS who had longer operative times. It is unclear whether longer operative times are secondary to CAS use or due to surgeon factors.

    Clinical Relevance

    Computer-assisted surgery for TKA has become increasingly popular; however, the clinical significance of CAS is unclear. Studies show that accuracy increases with CAS, especially in the coronal plane, but they do not indicate whether this improved accuracy will translate to improved clinical outcomes for patients.

    Supporters of CAS argue that clinical benefits will not be apparent until longer-term outcome studies become available. This study included a mean follow up of 4.5 years, with a range of 0 to 12 years, which is the longest follow-up known to evaluate the use of CAS as it applies to clinical outcomes. It corroborates previous intermediate studies showing no difference in revision rates or clinical functional scores between surgeons who used CAS more than 90% of the time and those who used conventional techniques more than 90% of the time.

    The primary outcome of revision rates was not different between the groups, indicating that computer-assisted technology for TKA may assist in more accurate cuts, but may not provide any additional benefit to the patient. There was a statistically significant decrease in Oxford Knee Scores at the 5-year interval in the mixed group when compared with both the CAS group and the conventional instrumentation group. This difference was not seen at the 6-month or the 10-year interval, making it unclear how clinically relevant this difference actually is.

    In addition, there was a time difference noted in cases in which CAS was used: Surgeons who routinely used CAS took about 10 minutes longer. All these results may contribute to a recent topic of interest, which is operative cost savings. Because no differences in primary or secondary outcomes were seen between the 2 groups, the longer operative time for CAS may contribute to higher operative costs, including the use of the technology and billing for the actual operative time.

    Another area of concern in past studies was the selection bias introduction when using CAS. Roberts et al were able to control for case complexity and the learning curve of using CAS, both of which were previously noted biases, by analyzing groups based on surgeons rather than technology. In this sense, the study eliminated potential bias where CAS was used for complex cases, which may have resulted in inferior patient outcomes due to other factors. 

    This study contributes to a growing body of literature demonstrating that although computer-assisted technology provides a more-accurate surgery, it does not necessarily translate to improved clinical outcomes for patients. Importantly, surgeons who primarily used CAS did not show inferior outcomes, indicating that good outcomes are attainable with either CAS or traditional techniques. In addition, this study supports the fact that functional outcomes, including the rates of component revisions, are more complex than just a technically sound surgery.

    Author Information

    Amy Kaplan, MD, is an orthopedic surgery resident at NYU Langone Orthopedic Hospital, New York, New York. Ajit Deshmukh, MD, is a Clinical Associate Professor of Orthopedic Surgery in the Division of Adult Reconstruction at NYU Langone Orthopedic Hospital, New York, New York.

    Disclosures: The authors have no disclosures relevant to this article.