Precision and Accuracy of Polyethylene Thickness in Total Knee Arthroplasty: A Comparison of Conventional Measured Resection and Robotic-Assisted Techniques

Abstract

Background

Total knee arthroplasty (TKA) aims to restore native knee kinematics through accurate bone resection and balanced flexion and extension gaps. Polyethylene thickness is a critical determinant of joint line preservation and stability. This study compared polyethylene implant accuracy and precision relative to the preoperative plan in robotic-assisted (RA) and conventional (C) TKA.

Methods

A retrospective review was performed of all patients who underwent TKA by a single surgeon at a single institution between March 1, 2018, and July 31, 2022. Patients were classified as having undergone either RA-TKA or C-TKA. A 9-mm polyethylene thickness was planned in all cases to achieve balanced flexion and extension gaps. Patient demographics and final polyethylene thickness were recorded. Mean differences from the planned thickness, deviation from the plan, and variance were compared.

Results

A total of 815 patients were included, with 404 (49.6%) undergoing C-TKA and 411 (50.4%) undergoing RA-TKA. Sex distribution did not differ between groups, though patients in the RA-TKA group were significantly older (70.46 ± 8.52 years vs 69.14 ± 8.34 years, P = .027). Mean polyethylene thickness was greater in C-TKA than RA-TKA (10.67 mm ± 1.18 mm vs 9.34 mm ± 0.61 mm, P < .001) with mean differences of 1.67 mm and 0.34 mm, respectively. Adherence to the planned thickness was higher in RA-TKA (72.26% vs 16.09%) and variance was significantly lower (0.37 vs 1.39, P < .001).

Conclusions

RA-TKA demonstrated greater accuracy and precision in achieving planned polyethylene thickness, suggesting improved execution of the surgical plan and more consistent gap balancing. Prospective studies should determine whether these differences translate to superior clinical outcomes.

Introduction

Total knee arthroplasty (TKA) is a highly successful procedure for alleviating pain and restoring joint function for patients with end-stage osteoarthritis of the knee. It is also among the most common surgical procedures performed in the United States, with over 700,000 TKAs performed each year [1]. In order to achieve an optimal postoperative outcome, TKA should seek to restore joint alignment, achieve stability, and balance soft tissues. As the only modular component of the final implant construct, the polyethylene insert plays a central role in allowing for the simultaneous balancing of both flexion and extension gaps [2,3]. Polyethylene thickness directly influences joint tension in both planes, making accurate polyethylene thickness selection critical to achieving stability [2,4,5]. Inappropriately sized polyethylene implants can lead to instability, abnormal kinematics, stiffness, and increased implant wear, all of which may decrease implant longevity and patient satisfaction [[6], [7], [8], [9]].

Robotic-assisted TKA (RA-TKA) has emerged as a potential solution for improving intraoperative implant selection and alignment. Rather than relying exclusively on traditional instrumentation to achieve accurate cuts and proper implant alignment, RA-TKA uses computer-assisted technology to help plan and execute the surgery [10]. Several studies have found that RA-TKA improves the accuracy and precision of component implantation relative to conventional TKA (C-TKA) [[11], [12], [13], [14]]. Despite the improvements in implant alignment and precision observed in RA-TKA, the impact of this technology on polyethylene insert selection remains underreported. While much of the existing literature focuses on bony alignment and component positioning, few studies have directly compared the accuracy and precision of polyethylene implantation between RA-TKA and C-TKA. As thicker polyethylene inserts have been associated with failure and higher rates of aseptic revision surgery, determining if RA-TKA confers an advantage in achieving the desired polyethylene thickness is of critical importance [7,[15], [16], [17]].

The present study aimed to compare the accuracy and precision of polyethylene insert implantation between RA-TKA and C-TKA by quantifying deviation from a standardized, preoperatively planned polyethylene insert thickness. We hypothesized that RA-TKA would result in improved accuracy and precision of polyethylene implantation with respect to implant thickness compared to C-TKA.

Material and methods

Cohort identification

A retrospective review of all patients who underwent TKA between March 1, 2018, and July 31, 2022, by a single surgeon at a single institution was performed. Only patients >18 years of age undergoing elective, primary unilateral knee arthroplasties were included. Revision, conversion, and unicompartmental knee arthroplasties were excluded. The electronic medical record was retrospectively reviewed to identify patient demographics, including age and sex. Operative reports were reviewed to confirm the use of the robotic assistance and the knee implant system. The thickness of the polyethylene implant used intraoperatively was recorded. This study was submitted for Institutional Review Board review and approved with a waiver of informed consent based on its retrospective nature with minimal risk to participants.

Outcomes

The primary outcome of this study was implant accuracy, defined as the difference between the preoperatively planned polyethylene thickness (9 mm) and the final implanted polyethylene thickness. Accuracy was assessed using the mean absolute difference from the planned value. The secondary outcome was implant precision, assessed by the variance in polyethylene thickness within each cohort. Other outcomes included rate of adherence to and the frequency of deviation from the preoperatively planned polyethylene thickness of 9 mm.

Statistical analysis

Standard descriptive statistics were used to define means and standard deviations for patient demographics and polyethylene thickness. Differences were evaluated using Chi-square analyses for categorical variables and independent t-tests for continuous variables. Levene's test was used to assess variance in polyethylene insert thickness between the C-TKA and RA-TKA cohorts. P values of P < .05 were considered statistically significant. All statistical analyses were performed using STATA (version 18.0; StataCorp, College Station, TX). A post-hoc power analysis was performed using the observed means and standard deviations of the primary outcome, which confirmed that the sample size of 815 patients provided >99% power to detect the observed differences at a significance level of .05.

Surgical approach and implant selection

All procedures were performed using a midline incision followed by a medial parapatellar arthrotomy. A limited tourniquet technique was utilized in all cases. The tourniquet was inflated prior to incision and deflated shortly thereafter once the initial surgical exposure was complete. This was performed to minimize limb ischemia while maintaining a clear field during the initial approach. Mechanical alignment principles were applied to restore the coronal axis in both the C-TKA and RA-TKA cohorts. The C-TKA group utilized a posterior-stabilized Smith & Nephew Legion design (Smith & Nephew, Watford, England, UK), while the RA-TKA group used a cruciate-sacrificing Triathlon design with the Stryker Mako robotic-assisted platform (Stryker, Kalamazoo, MI). While the implant designs differ, the surgical goal of achieving balanced gaps with a target insert thickness of 9 mm was consistent. All implants were cemented. Freehand patellar resurfacing was performed in all cases, aiming to restore composite patellar thickness to within 1 mm of the original measurement. All incisions were closed using a standard layered technique.

Conventional total knee arthroplasty technique

C-TKA was performed with a measured resection technique. On the femoral side, an intramedullary guide was placed into the canal. Using a distal cut first method, the resection depth was set to 8 mm and a 5 degree valgus cut. Femoral component rotation was determined using anatomic landmarks and sizing was determined using an anterior referencing guide. An extramedullary cutting guide was used on the tibial side with a resection depth set to 7 mm off the lateral side. After tibial and femoral cuts were made, gap balancing was assessed using a spacer block. The final components were placed and cemented after confirmation of the desired balance. Soft tissue releases were performed as needed.

Robot-assisted total knee arthroplasty technique

A preoperative computed tomography (CT) of the leg was obtained to develop a three-dimensional model for segmentation. The provisional plan was set using the 3-D model of the knee created from the preoperative CT scan. Intraoperatively, arrays were placed in the femur and tibia using the intraincisional unicortical technique as described by Yun et al. [18]. The hip center, ankle center, and femoral and tibial surfaces of the knee were registered. Starting alignment in the coronal and sagittal planes were measured. Dynamic soft tissue balancing was done to correct the preoperative deformity. The bone surface was then mapped on the CT model. The virtual implants were then adjusted to restore neutral mechanical alignment and to create symmetric flexion and extension gaps which correspond to a 9-mm tibial polyethylene insert. The preoperative plan was then modified by adjusting femoral and tibial component position and depth of femoral and tibial bone resections to create a balanced joint. After performing the bone cuts, the medial and lateral gaps in both flexion and extension were assessed, while varus and valgus stress was placed on the joint with the spacer block in place. When asymmetry in flexion or extension gaps was encountered, soft tissue release was performed and gap symmetry re-evaluated. After evaluation of the joint with trial components and placement of the final components, the selection of the final polyethylene was determined based on correction of the starting coronal and sagittal deformity.

Results

A total of 815 primary TKAs were performed, with 404 patients (49.57%) undergoing C-TKA and 411 patients (50.43%) undergoing RA-TKA. The RA-TKA cohort was slightly older (70.46 ± 8.52 years vs 69.14 ± 8.34 years, P = .027). There was no significant difference in sex distribution between the 2 groups (Table 1).

Table 1.

Demographic characteristics.

Demographics	Conventional TKA (C-TKA) N = 404		Robotic-assisted TKA (RA-TKA) N = 411		P-value
	Average	Standard deviation (SD)	Average	Standard deviation (SD)
Age (y)	69.14	8.34	70.46	8.52	0.027
	N	%	N	%	P-value
Female	228	56.44%	241	58.64%	0.572
Male	176	43.56%	170	41.36%

The average polyethylene thickness in the C-TKA cohort was significantly higher than that of the RA-TKA cohort (10.67 mm ± 1.18 mm vs 9.34 mm ± 0.61 mm, P < .001). As the planned polyethylene thickness was 9 mm in all cases, the mean difference from planned to final polyethylene thickness was 1.67 mm in the C-TKA group and 0.34 mm in the RA-TKA group (Table 2).

Table 2.

Average polyethylene thickness.

Parameter	Conventional TKA (C-TKA) N = 404		Robotic-assisted TKA (RA-TKA) N = 411		P value
	Average	Standard deviation (SD)	Average	Standard deviation (SD)
Planned (mm)	9.00	0.00	9.00	0.00	–
Actual (mm)	10.67	1.18	9.34	0.61	<0.001
Mean difference (mm)	1.67	–	0.34	–	<0.001

Among the 404 patients in the C-TKA cohort, 65 (16.09%) required a 9 mm polyethylene insert, 129 (31.93%) required a 10-mm insert, 118 (29.21%) required an 11-mm insert, 70 (17.33%) required a 12-mm insert, and 22 (5.45%) required a 13-mm insert. No patients required inserts thicker than 13 mm. Of the 411 patients in the RA-TKA cohort, 297 (72.26%) required a 9-mm polyethylene insert, 94 (22.87%) required a 10-mm insert, 15 (3.65%) required an 11-mm insert, and 5 (1.22%) required a 12-mm insert. No patients in the RA-TKA cohort required inserts thicker than 12 mm (Fig. 1). Adherence to the planned polyethylene thickness of 9 mm was markedly higher among the RA-TKA group than the C-TKA group (72.26% vs 16.09%). Furthermore, 95.13% of patients in the RA-TKA group were within 1 mm of the planned 9 mm target, compared to only 61.14% in the C-TKA group.

Figure 1.

Distribution of polyethylene implants used in conventional TKA (C-TKA) and robotic-assisted TKA (RA-TKA) by thickness (in mm).

The variance in final polyethylene thickness was 1.39 mm² in the C-TKA cohort and 0.37 mm² in the RA-TKA cohort. Levene's test demonstrated that variance was significantly greater in the C-TKA cohort (P < .001) (Table 3).

Table 3.

Distribution of polyethylene implants by thickness.

PE thickness	Conventional TKA (C-TKA) N = 404		Robotic-assisted TKA (RA-TKA) N = 411		P value
	N	%	N	%
9 mm	65	16.09%	297	72.26%	<0.001
10 mm	129	31.93%	94	22.87%	<0.001
11 mm	118	29.21%	15	3.65%	<0.001
12 mm	70	17.33%	5	1.22%	<0.001
13 mm	22	5.45%	0	0.00%	<0.001

Discussion

The present study found that RA-TKA resulted in improved accuracy and precision of polyethylene implantation compared to C-TKA. Average deviation from preoperatively planned polyethylene thickness was significantly lower in the RA-TKA group relative to C-TKA group (0.34 mm vs 1.67 mm, P < .001), with 72.3% of RA-TKAs achieving the exact target of 9 mm vs only 16.1% of C-TKA cases. Additionally, 95% of RA-TKAs were within 1 mm of the preoperative plan, compared to 48% of C-TKAs. Variance in final polyethylene thickness was also significantly reduced in RA-TKA relative to C-TKA (0.37 vs 1.39, P < .001), indicating more reliable intraoperative execution. These findings suggest that robotic assistance enables greater precision and accuracy of polyethylene thickness implantation than conventional techniques in the setting of primary TKA.

The present study's findings align with the existing body of literature investigating the role of RA-TKA in implant precision and accuracy. Prior research has found that RA-TKA improves alignment, component positioning, and restoration of the mechanical axis relative to C-TKA [13,19,20]. However, the impact of surgical technique on polyethylene implant thickness remains unexplored. Deckey et al. [11] reported that the use of RA-TKA resulted in a lower rate of distal femoral recuts than C-TKA (10% vs 22%), suggesting improved cut precision with RA-TKA systems and indirectly supporting the findings of the present study. Despite this, a direct comparison of the ability to adhere to the preoperative plan for the polyethylene implant between RA-TKA and C-TKA has not yet been performed. This study uniquely focuses on the consistency of polyethylene thickness as a marker for balancing strategy and intraoperative fidelity, adding a novel dimension to the understanding of robotic assistance in TKA beyond coronal and sagittal plane alignment.

In C-TKA, postresection polyethylene thickness adjustments are often a compromise between the flexion and extension gaps within the limits of postresection soft tissue releases, especially in cases of sagittal plane deformity. For example, knees that extend into hyperextension will often require a thicker polyethylene insert, while knees with residual flexion contracture receive a thinner insert. While the extension gap is frequently the primary driver of thickness selection, the iterative process of balancing the flexion gap with conventional techniques (increasing posterior slope or progressive posterior cruciate ligament release) can be less precise compared to the real-time, millimeter-level data provided by robotic systems. It would be an overgeneralization to suggest the flexion gap is ignored; rather, the conventional iterative process lacks the high-fidelity predictive capabilities inherent to robotic platforms. In contrast, RA-TKA allows surgeons to make intraoperative adjustments to bone cuts and perform real-time evaluation of flexion and extension gaps allowing for increased bone preservation [11]. Two hypotheses may explain the reduced polyethylene thickness and variance seen in RA-TKA. First, CT-based systems may reduce the risk of over-resection by identifying virtual anatomic landmarks with higher precision than surface-based techniques. Second, the availability of intraoperative quantitative data may allow for smaller, more targeted soft tissue releases, avoiding the need to compensate with thicker polyethylene. This is clinically significant as prior studies have associated polyethylene thickness ˃13 mm with an increased risk of revision, possibly due to altered joint line mechanics or increased bone resection [16,17]. Though none of the patients included in either cohort required a polyethylene insert larger than 13 mm, 13 mm inserts were used in 22 C-TKA cases (5.45%) while none of the RA-TKA cases necessitated a 13-mm insert. The improved precision observed in the RA-TKA group may reduce the need for thicker polyethylene components and allow for longer-term implant durability.

Accurate polyethylene implant selection has meaningful clinical implications for the longevity of the TKA construct. Polyethylene thickness directly influences joint line restoration and overall limb alignment, both of which contribute to function and component survivorship. Thicker polyethylene inserts may lead to joint line elevation while increasing the risk of anterior knee pain. In contrast, thinner polyethylene components may demonstrate unfavorable wear characteristics and decreased durability necessitating revision surgery for liner exchange [21,22]. By improving the consistency with which surgeons are able to achieve the intended polyethylene thickness, RA-TKA may promote more reliable joint line restoration, reduce the need for compensatory soft tissue releases, and preserve subchondral bone stock. The technical advantages conferred by RA-TKA allowing for increased accuracy and precision in polyethylene insert thickness may translate into improved long-term kinematics and more predictable clinical outcomes after surgery.

This study has limitations that must be acknowledged. First, as a retrospective analysis of cases performed at a single institution by a single surgeon, the generalizability of these findings may be limited. Second, the C-TKA and RA-TKA cohorts utilized different implant systems (posterior-stabilized Smith & Nephew Legion vs cruciate-sacrificing Stryker Triathlon). These designs possess different kinematic profiles and constraint levels, which inherently influence gap-balancing requirements and subsequent polyethylene selection. While both designs involve sacrificing the posterior cruciate ligament, the variation in design-specific constraint represents a confounding variable that limits our ability to attribute all observed differences solely to the surgical platform. However, the markedly reduced variance in the RA-TKA group suggests that robotic assistance enables higher fidelity to a planned target regardless of these design differences. Future studies using matched implant designs or subgroup analyses are needed to isolate the independent effect of robotic assistance on polyethylene accuracy. Third, this study did not assess postoperative alignment, clinical outcomes, or patient-reported satisfaction scores. Although improved polyethylene accuracy is theoretically beneficial, the long-term clinical relevance of this metric remains to be validated.

There are also several notable strengths. All procedures were performed by a single high-volume arthroplasty surgeon, thus reducing variability in surgical technique and decision-making. This uniformity strengthens internal validity and allows for clearer attribution of observed differences to the surgical platform rather than the surgeon (at the expense of decreased generalizability). Additionally, polyethylene thickness was treated as a continuous, objective intraoperative variable that directly reflects component positioning and soft tissue balancing, thereby offering a novel, quantifiable metric of precision and accuracy. By evaluating adherence to a consistent preoperative plan across 2 techniques, this study avoids subjective interpretation and focuses on a reproducible outcome that is easily compared.

Conclusions

RA-TKA significantly improved polyethylene implant accuracy and precision in comparison to conventional measured resection techniques. Future larger, multisurgeon studies incorporating variations in technique, alignment philosophy, and matched implant designs are needed to further validate these findings. Additionally, prospective research should examine whether improved precision and the associated bone preservation lead to a reduced risk of revision TKA and superior patient-reported outcome scores.

CRediT authorship contribution statement

Kory B. Dylan Pasko: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Conceptualization. McKenzie W. Culler: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis. Torie Oishi: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis. Avinash Iyer: Writing – review & editing, Writing – original draft, Methodology, Investigation. Nathanael D. Heckmann: Writing – review & editing, Supervision, Formal analysis, Conceptualization. Andrew G. Yun: Writing – review & editing, Supervision, Methodology, Formal analysis, Conceptualization.

Conflicts of interest

Andrew Yun receives royalties from Smith & Nephew, is a speaker for paid presentations, and is a paid consultant for Avanos and Mizuho. Nathanael Heckmann is a paid consultant for Intellijoint Surgical, MicroPort Orthopedics, Corin USA, and Zimmer; holds stock in Intellijoint Surgical; and serves as a board member of AAOS, AJRR, AAHKS, the Knee Society, and the Hip Society. The other authors declare there are no conflicts of interest.

For full disclosure statements, refer to https://doi.org/10.1016/j.gie.2026.05.037.

Appendix A. Supplementary data

Conflict of Interest Statement for Iyer