8/6/2019 Source of Errors in TKR

1/4

Sources of Error in Total Knee Arthroplasty

by LoukasKoyonos, MD; S.

David Stulberg, MD; Todd C.

Moen, MD; Gina Bart, PAC;

Michael Granieri, BS

Abstract

The purpose of this study was to identify the procedural steps in a total knee arthroplasty (TKA) in which

technical errors occur and to quantify the magnitude of these errors. Forty-nine consecutive TKAs were

performed using a traditional exposure and manual instrumentation. An image-free computer navigation

system (OrthoPilot; Aesculap AG, Tuttlingen, Germany) was used to measure and compare femoral and tibial

alignment at specific procedural points during the TKA; this data was then used to evaluate possible sources of

error in the procedure. The femoral cut tended to be made in hyperextension, the tibial cut tended to be made

in hyperextension and valgus, and the tibial component tended to be implanted in valgus. This study identified

specific points during the performance of a TKA where technical errors occur. This information suggests

technical considerations that can help a surgeon achieve more reproducible, durable, and successful

outcomes for his or her patients.

The outcome of a total knee arthroplasty (TKA) is dependent on a number of factors, one of the most important

being implant alignment. Improper alignment of the implants can lead to accelerated wear and early failure,

whereas properly positioned components have shown increased longevity.1-6

Numerous studies have shown

that errors in alignment in the coronal plane >3 lead to earlier failures and suboptimal results in TKAs.7,8

The alignment of the implants, and thus the outcome of a TKA, has been shown to be particularly sensitive to

surgical technique.5,9-13

Despite continual refinement in manual instrumentation systems to improve accuracy

of implant alignment, errors continue to occur.14

From the placement of the alignment guides and the securing

of the cutting blocks to the actual bony cuts themselves and the final cementing of the implants, errors can

occur at numerous points in a TKA.

Although studies have evaluated the accuracy of traditional TKA instrumentation, to our knowledge there has

been no study examining the precise procedural steps during a TKA in which alignment errors occur. The

purpose of this study was to identify the technical steps in which these errors occur and quantify the magnitude

of these errors with the use of an image-free computer navigation system (OrthoPilot; Aesculap AG, Tuttlingen,

Germany).

Materials and Methods

This study was approved by our Institutional Review Board. Informed consent for this study was obtained from

all participants. The senior author (S.D.S.) performed a TKA on 49 consecutive patients using a currently

available intramedullary instrumentation system. Prior to beginning the procedure, tracking diodes for the

navigation system were secured percutaneously to the distal femur and proximal tibia. The centers of rotation

of the head of the femur, knee, and ankleand thus the mechanical axis of the limbwere established using

kinematic and surface registration techniques.

The knee was exposed using a standard median parapatellar approach. The femoral intramedullary alignment

guide was set at the anatomic axis (5-9) that would result in a mechanical frontal axis of 0 based on the

8/6/2019 Source of Errors in TKR

2/4

measurements of preoperative long radiographs taken with the patient standing. The AP intercondylar

(Whitesides) line was used to establish the rotation of the femoral component. The femoral cutting block was

placed perpendicular to the AP intercondylar line. The tibial cutting block was positioned on the intramedullary

alignment guide to allow a cut to be made perpendicular to the intramedullary axis of the tibia. A 0 posteriorly

sloped cutting block was used. Both the femoral and tibial cutting guides were slotted.

The navigation system has a check plate to which is attached a diode-containing tracker. This check plate can

be applied to each of the cutting blocks that have been attached to the femur and tibia to determine their

alignment. The check plate can also be applied directly to the cut surfaces of the femur and tibia to measure

the alignment accuracy of the cuts. The check plate was used to measure the alignment of the femoral and

tibial cutting blocks before and after each actual cut was made. Alignment was recorded in the frontal, sagittal,

and axial planes. The difference between the alignment values before and after each cut was termed the

movement of the cutting block. The check plate was also used to check the final alignment of the cut.

Alignment was measured in the frontal, sagittal, and axial planes for the femoral cut, and in the frontal and

sagittal planes for the tibial cut. The final bony cuts alignment was then compared to the final alignment of the

cutting block. The difference between these 2 values was termed the accuracy of the cut. It should be noted

that after the measurements of the alignment of the initial cut were made, the senior surgeon (S.D.S.) used thenavigation system and check plate to adjust the bony resection appropriately to match the alignment of the

cutting blocks.

Finally, the check plate was used to measure frontal and sagittal alignment of the tibial component following

final implantation and cementation. This alignment was then compared to the final alignment of the bony cuts;

the difference between these 2 values was termed the accuracy of implantation. The accuracy of implantation

of the femoral component was not measured due to the femoral components lack of a flat surface on which to

apply the check plate. Given the nonparametric distribution of the data, a paired ttest and Wilcoxins signed

rank test was used to measure the differences between measurements.

Results

The mean movement of the femoral cutting block in the coronal plane was 0.03 valgus (.34; P>.05; range,

0-1). The mean movement of the femoral cutting block in the sagittal plane was 0.04 extension (.46;

P>.05; range, 0-1). The mean movement of the femoral cutting block in the axial plane was .03 external

rotation (0.41; P>.05; Table 1).

The mean movement of the tibial cutting block in the coronal plane was 0.06 varus (0.61; P>.05; range, 0-

1). The mean movement of the tibial cutting block in the sagittal plane was 0.10 extension (0.61; P>.05;

range, 0-1; Table 1).

In the coronal plane, the femoral cut was found to be in 0.06 of varus (0.55; P>.05) with respect to the final

alignment of the cutting block. In the axial plane, the femoral cut was found to be in neutral rotation (0.36;

P>.05) with respect to the final alignment of the cutting block. In the sagittal plane, the femoral cut was found tobe in 0.96 of extension (1.19; P

8/6/2019 Source of Errors in TKR

3/4

In the coronal plane, the final alignment of the tibial component was found to be in 0.22 of valgus (0.62;

P

8/6/2019 Source of Errors in TKR

4/4

In the sagittal plane on both the tibial and femoral side, the initial bony cuts tended to be left in hyperextension

with respect to the alignment of the cutting block, with some discrepancies as high as 3. Given that the

alignment of the cutting blocks did not significantly change before and after the actual cut, these errors were

introduced during the performance of the cut itself. These errors were likely a result of deflection of the saw

blade off hardened sclerotic bone as the bony resection was performed.

In the coronal plane on the tibial side only, there was a small but statistically significant trend for the initial bony

cut to be left in valgus with respect to final alignment of the cutting block. Again, as the alignment of the cutting

blocks was found to be stable before and after the bony cut, this error was introduced during the cut itself and

again was likely due to the deflection of the saw blade off sclerotic, osteoarthritic bone. It is noteworthy that the

only significant source of error in the coronal plane occurred during the tibial resection. The initial deformity and

predominant location of osteoarthritis was not accounted for, and knees with predominantly medial

compartment diseaseand thus varus deformityand predominantly lateral compartment diseaseand thus

valgus deformitywere included. The predominance of medial compartment osteoarthritis accounts for the

small but significant trend toward a cut made in excessive valgus. If the cohort of patients in this study were

separated and studied based on location of disease, we anticipate there would be statistically stronger trends

in the anticipated direction, ie, medial compartment disease causing a cut made in excessive valgus.

The alignment of the final tibial implant following cementation also showed a small but statistically significant

trend toward implantation in valgus with respect to the alignment of the final tibial cut. This was independent of

errors introduced during the performance of the bony cuts. As mentioned, following the initial measurement of

the alignment of the bone cut, the navigation system was used to adjust the bony resection to insure an

accurate cut. Thus, the valgus error introduced was not influenced by any errors in alignment of the bone cuts.

Although the reason for this trend toward implantation in valgus is not entirely clear, in patients with medial

compartment osteoarthritis the bone of the lateral tibial plateau is softer than the sclerotic bone of the medial

plateau. This may predispose placement of the implant into valgus.

The sources of errors identified in this study suggest specific measures that can be taken to avoid these

technical pitfalls. Inherent in this discussion is the assumption that the cutting blocks were aligned correctly and

that each error compounds on itself. First, it is vital to use rigid, stiff, sharp cutting blades that can resect

sclerotic bone without deflection from the cutting block and thus avoid an inaccurate cut. Second, to further

prevent deflection from sclerotic bone, it is important to use an instrumentation system where the saw blade fits

as flush as possible within the cutting block. Third, the more sclerotic the osteoarthritic bone, the more

vigilance required by the surgeon to ensure that the bony cut that was made is flush with the alignment of the

cutting block. This point is particularly important in modern minimally invasive instrumentation systems in which

visualization of the cut surface may be suboptimal.

This study had some notable limitations. The sample size was relatively small. Although significant trends

declared themselves, there may be further conclusions to be drawn with a larger cohort of patients. Also, the

distribution of disease and deformity was not accounted for in the evaluation of errors. Inferences can be madegiven the prevalence of medial compartment osteoarthritis; however, if the specific disease patterns were

evaluated independently, a more concrete conclusion could be made as to the nature of errors.

This study identified specific points during a TKA where technical errors tend to be made. The femoral cut

tended to be made in hyperextension, the tibial cut tended to be made in hyperextension and valgus, and the

tibial component tended to be implanted in valgus. This information and the technical considerations it

suggests can assist the surgeon in achieving more reproducible, durable, and successful outcomes for their

patients following TKA.