Robotic-assisted versus conventional total knee arthroplasty: a systematic review and meta-analysis of alignment accuracy and clinical outcomes

Abstract

Background:

Robotic-assisted total knee arthroplasty (RA-TKA) has emerged as an alternative to conventional TKA (C-TKA), aiming to improve surgical precision and patient outcomes. This systematic review and meta-analysis study compares the efficacy of RA-TKA versus C-TKA.

Methods:

A comprehensive search of five databases (PubMed, EMBASE, Web of Science, SCOPUS, and Cochrane Library) was conducted. We included all published randomized controlled trials (RCTs) from inception to August 2024. Meta-analysis was done using RevMan 5.4 package.

Results:

Twenty-one RCTs involving 2692 patients were involved. RA-TKA demonstrated significantly lower mechanical alignment outlier rates (risk ratio = 0.33, 95% confidence interval (CI)[0.19, 0.59], P = 0.0002) and less deviation from neutral mechanical axis (mean difference, MD = −0.93° [−1.20, −0.66], P < 0.00001) compared to C-TKA. No significant differences were found in WOMAC or Oxford Knee Scores at various follow-up points. RA-TKA was associated with longer operative times (MD = 19.94 minutes [9.2, 30.68], P = 0.0003) but showed no significant difference in intraoperative blood loss. Postoperative Knee Society Scores were slightly higher in the RA-TKA group (MD = 1.03 [0.50, 1.57], P = 0.0002).

Conclusion:

RA-TKA offers improved mechanical alignment accuracy compared to C-TKA but does not demonstrate superior short to medium-term functional outcomes. The technology is associated with longer operative times. These findings suggest that while RA-TKA may enhance surgical precision, its clinical benefits and cost-effectiveness require further evaluation, particularly in long-term studies.

HIGHLIGHTS

• The use of robotic instruments has become increasingly popular due to positive outcomes for both patients and clinicians

• Conventional arthroplasty, particularly the knee, remains the standard due to its shorter duration and positive outcomes

• Alignment accuracy is higher with robot-assisted knee arthroplasty with shorter duration but patient outcomes may not be significantly different

Introduction

Knee osteoarthritis (OA) is the most prevalent joint disorder among the elderly, affecting approximately 30% of individuals over the age of 60^[1]. Around half of these individuals experience symptoms including joint pain, stiffness, swelling, and reduced joint function^[2].

Total knee arthroplasty (TKA) is a broadly performed surgical procedure for patients with advanced osteoarthritis of the knee^[3]. As the global population ages and the frequency of knee osteoarthritis increases, the demand for TKA continues to rise^[4]. While conventional TKA (C-TKA) techniques have demonstrated good long-term outcomes, there is an ongoing pursuit of innovations to enhance surgical precision, improve patient outcomes, and optimize implant longevity^[5]. In recent years, robotic-assisted TKA (RA-TKA) has become a promising replacement to conventional methods, offering the capability for improved accuracy in component positioning and alignment^[6].

C-TKA relies on manual instrumentation and surgeon expertise to achieve proper implant positioning and limb alignment^[7]. Despite advancements in surgical techniques and implant design, challenges persist in consistently achieving optimal component placement^[8]. Factors such as variations in patient anatomy, bone quality, and soft tissue balance can complicate the procedure^[8]. Malalignment of components has been linked with suboptimal functional outcomes, increased wear, and potential early implant failure^[8].

RA-TKA aims to address these challenges by leveraging advanced imaging, computer navigation, and robotic technology to enhance surgical precision^[9]. These systems typically involve preoperative planning based on patient-specific 3D imaging, intraoperative guidance, and robotic-arm assistance for bone preparation and implant positioning^[9]. The potential advantages of robotic-assisted TKA include more precise component placement, enriched restoration of mechanical alignment, and the ability to fine-tune soft tissue balancing^[9].

While numerous studies have investigated the outcomes of RA-TKA compared to conventional techniques, the existing literature presents conflicting findings. Some studies report significant improvements in alignment accuracy and early functional outcomes with robotic assistance^[10,11], while others suggest comparable results between the two approaches^[12,13]. Additionally, concerns have been raised regarding the learning curve combined with robotic systems, increased operative time, and the substantial cost of implementing this technology^[13]. Given the rapid adoption of robotic-assisted TKA in clinical practice, there is a critical need for a broad evaluation of its efficacy and safety compared to conventional methods.

This systematic review and meta-analysis study aims to address this knowledge gap by synthesizing the available high-quality evidence from randomized controlled trials (RCTs). The primary aim of this study is to compare robotic-assisted and conventional TKA in terms of alignment accuracy, functional outcomes, and perioperative parameters.

Methods

The meta-analysis was performed following the Cochrane Handbook for Systematic Reviews and Meta-Analyses^[14], while also embracing the comprehensive guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)^[15]. No ethical approval was required for this review.

Search strategy and study selection

A comprehensive literature search was conducted using five electronic databases: PubMed, EMBASE, Web of Science, SCOPUS, and Cochrane Library from their inception till August 2024. The search was aimed at identifying studies comparing RA-TKA to C-TKA. The search strategy included keywords and Medical Subject Headings^[16] terms related to “robotic-assisted total knee arthroplasty,” and “conventional total knee arthroplasty” (Supplementary Table 1, available at: http://links.lww.com/MS9/A713).

The screening process was carried out in two stages: first, titles and abstracts were reviewed, followed by a thorough full-text examination. Additionally, we performed a manual screening of previously published meta-analyses and examined the references of the included studies for further relevant RCTs.

Inclusion and exclusion criteria

Inclusion Criteria for the review were: RCTs that compared RA-TKA with C-TKA, studies that reported on key outcomes such as operative time, intra-operative blood loss, and functional outcomes including Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and Oxford Knee Score (OKS) scores. Only peer-reviewed publications were considered. Studies in languages other than English, non-randomized trials, and conference abstracts or non-peer-reviewed materials were excluded from the review.

Data extraction and quality assessment

Two reviewers separately gathered data using a standardized form, focusing on study characteristics (such as location, robotic systems used), patient demographics (e.g., age, sex), and outcome measures. Any differences between reviewers were addressed by discussion or involving another reviewer if needed.

The quality of the included RCTs was judged using the Cochrane Risk of Bias Tool version 2^[17]. This tool measures five key domains: the randomization process, deviations from the intended interventions, missing outcome data, outcome measurement, and the selection of reported results. Each of these domains was classified as having a low risk, some concerns, or a high risk of bias.

Statistical analysis

Pooled analyses were carried out using Review Manager (RevMan 5.4) software. Mean differences (MD) and 95% confidence intervals (CI) were estimated for continuous outcomes, while risk ratios (RR) with 95% CI were computed for dichotomous outcomes. When the data were homogeneous, we employed a fixed effects model to analyze the results.

Heterogeneity was evaluated using the I² statistic and the P value for heterogeneity, with values below 0.1 indicating high levels of heterogeneity. In cases where significant heterogeneity was identified (P < 0.1), we employed a random-effects model and conducted sensitivity analyses by eliminating studies that seemed to contribute substantially to the observed variability.

Results

A total of 1355 records were retrieved from five databases: PubMed (299), EMBASE (122), Web of Science (377), SCOPUS (450), and Cochrane Library (107). After removing 695 duplicate records, 660 unique records were screened based on titles and abstracts. Following this screening process, 626 records were excluded, leaving 34 reports for full-text screening. During the eligibility evaluation, 23 reports were excluded for various reasons: one was not in English, 9 were not RCTs, and 3 were conference abstracts. 21 RCTs were eligible, and 19 RCTs were included in the meta-analysis (Fig. 1).

Figure 1.

PRISMA flow diagram of study selection process.

Baseline and summary of the included studies

The studies were conducted across various sites, including the USA, UK, Belgium, Korea, China, Russia, and Thailand. Robotic systems utilized in these studies included Mako, NAVIO, ROBODOC, and others, each offering unique features and capabilities. Follow-up intervals ranged from 6 weeks to 10 years and age of patients varied, with ages ranging from 60 to 68 years. The percentage of male participants also varied, ranging from 8% to 50% throughout the different studies (Table 1).

Table 1.

Summary of included studies^[30-40]

Study ID	Site	Patients	Robotic system	Type of prosthesis	Primary outcomes	Follow-up	Conclusion	Study arms	Sample	Age, years, M ± SD	Sex, male, n (%)	Side, R/L
Airapetov et al 2023	USA	Stage 3 idiopathic osteoarthritis of the knee joint and varus deformity of the joint axis	-	Cruciate Retaining and PS Posterior Stabilized prostheses	Knee Society Score (KSS), Lysholm scale, WOMAC (Western Ontario and McMaster Universities Osteoarthritis Index), Intra-operative blood loss, Duration of intervention, Knee range of motion (Flexion, Extension) and Varus deformity correction	10 days	Despite its high cost and consumable needs, robot-assisted arthroplasty offered advantages like precise limb alignment, accurate prosthesis placement, reduced blood loss, and patient safety, though the surgeon’s role in planning, execution, and soft tissue balance remains crucial.	RA-TKA	10	60 ± 17.8	4 (40%)	5//5
								C-TKA	10	61 ± 20.7	3 (30%)	4/6
Ajekigbe et al 2024	UK	End-stage osteoarthritis	Mako	Cemented Triathlon cruciate-retaining TKA with a highly cross-linked (X3) polyethylene insert.	Spatiotemporal gait and balance parameters, including cadence, velocity, step-stride, and gait cycle parameters.	12 months	While RATKA may provide greater radiologic accuracy than COTKA, this difference was likely not clinically significant, and there appeared to be no meaningful difference in clinical outcomes, with revision and complication rates remaining inconclusive due to limited evidence.	RA-TKA	26	-	-	-
								C-TKA	23
Bollars et al 2023	Belgium	Painful and disabled knee joint resulting from osteoarthritis	NAVIO	Cemented posterior-stabilized TKA with 3° varus obliquity in the polyethylene	Implant placement accuracy, alignment in the frontal, coronal, and sagittal planes, and the occurrence of outliers (defined as deviations >3° from the perioperative planning).	6 weeks	An imageless handheld RATKA system allows accurate perioperative planning of individual implant positions, reducing alignment outliers while aiming for constitutional alignment.	RA-TKA	26	66.4 ± 7.2	9 (34.6%)	11/15
								C-TKA	26	64.4 ± 8.7	11 (42.3%)	15/11
Clement et al 2024	UK	Participants with osteoathritis of the knee	Mako	Triathlon with X-3 highly cross-linked polyethylene	Knee-specific measures (WOMAC, OKS, FJS) and HRQoL measures (EQ-5D and EQ-VAS).	12 months	Patients undergoing rTKA experienced greater knee pain relief and were more likely to meet their daytime pain relief expectations within the first 12 months compared to mTKA patients, but rTKA did not show a clinically significant difference in knee-specific function or HRQoL.	RA-TKA	43	67.0 ± 8.6	22 (51.2%)	-
								C-TKA	38	66.5 ± 8.6	16 (42.1%)
Fontalis et al 2022	UK	Symptomatic knee osteoarthritis	-	Cemented, cruciate-retaining Stryker Triathlon knee system	Local inflammatory markers (IL-6, IL-8, TNF-alpha), pain, opiate requirements, knee joint ROM, time to discharge, and number of physiotherapy sessions.	2 years	Robotic-arm-assisted total knee arthroplasty was linked to a decrease in the early postoperative local inflammatory response.	RA-TKA	15	67.9 ± 8.6	6 (40%)	8/7
								C-TKA	15	68.7 ± 9.6	7 (46.7%)	7/8
Kayani et al 2021	UK	Osteoarthritis	Mako	Triathlon cruciate-retaining knee system with patellar resurfacing usingasymmetrical components.	Postoperative levels of inflammatory markers (e.g., IL-6, TNF-α, CRP, ESR, LDH, CK), MASTI scores for soft tissue injury and bone trauma, and Accuracy of achieving the planned limb alignment and component positioning.	28 days	Robotic TKA led to a temporary decrease in the early (day 7) postoperative inflammatory response, with no difference in the immediate (<48 hours) or late (day 28) systemic inflammatory response compared to conventional TKA.	RA-TKA	15	68.7 ± 9.6	6 (40%)	8/7
								C-TKA	15	68.7 ± 9.6	7 (46.7%)	7/8
Kim et al 2019	Korea	Osteoarthritis	ROBODOC	-	Clinical, radiographic, and CT scan evaluation.	10 years	After a minimum follow-up of 10 years, no significant differences were found between robotic-assisted and conventional TKA in functional outcomes, aseptic loosening, survivorship, or complications.	RA-TKA	674	60 ± 4	132 (19.6%)	-
								C-TKA	674	61 ± 4.75	144 (21.4%)
Li et al 2022	China	Osteoarthritis	HURWA	Legion system	Rate of malalignment for mechanical axis greater than 3°	3 months	HURWA robotic-assisted TKA was safe and effective, providing better mechanical axis alignment compared to conventional TKA.	RA-TKA	73	68 ± 7.97	13 (17.8%)	-
								C-TKA	77	69 ± 6	15 (19.5%)
Liow et al 2014	Singapore	Primary knee osteoarthritis with genu varus deformity and a fixed flexion deformity of less than 15°	ROBODOC	Zimmer NexGen LPS-Flex posterior-stabilised implants	Joint line restoration accuracy, mechanical axis alignment, and clinical outcome measures (e.g., Knee Society scores, Oxford Knee Score, SF-36 scores)	6 months	Robot-assisted TKA achieved comparable short-term clinical outcomes to conventional methods while reducing MA alignment and joint-line deviation outliers.	RA-TKA	31	67.5 ± 8.6	-	-
								C-TKA	29	68.3 ± 7.7
Liow et al 2017	Singapore	Primary knee osteoarthritis with genu varus deformity and a fixed flexion deformity of less than 15°	ROBODOC	Zimmer NexGen LPS-Flex posterior-stabilised implants	Range of motion, Oxford Knee Score (OKS), Knee Society Score (KSS), SF-36 Health Survey scores (physical functioning, social functioning, role physical, bodily pain, mental health, role emotional, vitality, general health).	2 years	Robotic-assisted TKA showed slight enhancements in patient quality of life measures compared to conventional TKA.	RA-TKA	31	67.5 ± 8.6	-	-
								C-TKA	29	68.3 ± 7.7
Liu et al 2024	China	Elderly adults (≥65 years old) undergoing their first unilateral TKA	-	-	Postoperative alignment accuracy (HKA angle deviation, CTCA, CFCA, STCA, SFCA deviation), Knee Society Score (KSS), and Visual Analog Scale (VAS) for pain	12 weeks	Given that conventional TKA outcomes can be affected by nutritional status, elderly patients with GNRI ≤ 100 might benefit more from robotic-assisted TKA, which offers greater stability and clinical benefits.	RA-TKA	11	-	-	-
								C-TKA	12
Lychagin et al 2022	Russia	Osteoarthritis	TSolution-One	-	Pain syndrome (VAS), Functional state (Oxford Knee Score – OKS, Western Ontario and McMaster Universities Arthritis Index – WOMAC), Range of Motion (ROM), and Mechanical axis deviation (teleroentgenography)	12 months	Robot-assisted TKA provided more precise mechanical axis alignment, enhancing walking biomechanics.	RA-TKA	33	67.3 ± 3.7	15 (28.3%)	-
								C-TKA	35
Park et al 2007	Korea	Osteoarthritis	ROBODOC withORTHODOC	Zimmer LPS prosthesis	Knee Society Score and Evidence of loosening, prosthetic alignment, and othercomplications	24 months	Robotic-assisted technology offered clear benefits in preoperative planning, intraoperative accuracy, and postoperative follow-up, particularly for femoral and tibial flexion angles in X-rays. However, it had a higher early complication rate.	RA-TKA	32	62.7 ± 6.51	-	-
								C-TKA	30	67.8 ± 6.44
Song et al 2011	Korea	Bilateral osteoarthritis	ROBODOC	NexGen prosthesis	Range of motion, HSS scores, WOMAC scores (for pain and function), Mechanical axis alignment, Operative time, Incision length, and Drainage amount	12 months	The improved alignment accuracy of robotic TKA and its positive clinical results may enhance both clinical and radiological outcomes.	RA-TKA	30	67 ± 6.3	0 (0%)	Both
								C-TKA	30
Song et al 2013	Korea	Primary osteoarthritis of the knee and a mechanical axis between 20 angle and 5 angle valgus	ROBODOC1 System	Posterior cruciate-retaining prosthesis	Percentage of cases within ± 3° of neutral in coronal alignment	41 months	Robotic-assisted TKA reduced mechanical axis alignment outliers and improved flexion-extension gap balance, with similar clinical scores and complication rates compared to conventional methods.	RA-TKA	50	66.1 ± 7.1	4 (8%)	-
								C-TKA	50	64.8 ± 5.3	5 (10%)
Thiengwittayaporn et al 2021	Thailand	Primary knee osteoarthritis	NAVIO	Fixed-bearing posterior stabilized implant	Accuracy of implant positioning, alignment of mechanical axis, femoral and tibial inclinations, changes in joint line, and posterior femoral offset.	6 weeks	Imageless robot-assisted TKA offerd improved alignment accuracy and a short learning curve, making it an appealing choice for TKA.	RA-TKA	75	69.0 ± 8.3	6 (8%)	45/32
								C-TKA	77	69.1 ± 7.3	15 (19.5%)	40/35
Tian et al 2023	China	Osteoarthritis	Jianjia	Zimmer Biomet	Postoperative HKA angle deviation, coronal and sagittal tibial and femoral component angles, clinical outcomes (Knee Society score, VAS, ROM)	12 weeks	The new robotic-assisted TKA system proved to be safe and effective. Preoperative HKA angle deviation influences postoperative HKA angle deviation.	RA-TKA	62	68.17 ± 7.59	13 (21%)	29/33
								C-TKA	61	68.84 ± 7.12	15 (24.6%)	31/30
Vaidya et al 2022	India	Primary osteoarthritis of the knee joint with varus deformity	NAVIO	Posterior-stabilized prosthesis	Accuracy of mechanical axis deviation (MAD), femoral and tibial coronal alignment (FCA and TCA), joint line elevation	Postoperative	The new imageless, handheld semi-autonomous robotic system for TKA offered superior accuracy in component positioning and mechanical alignment in the coronal plane compared to conventional TKA.	RA-TKA	32	62.2 ± 10	8 (25%)	19/13
								C-TKA	28	59.9 ± 8	4 (14.3%)	14/14
Xu et al 2022-1	China	End-stage knee osteoarthritis	YUANHUA-TKA	Unified fixed-platform and posterior cruciate ligament sacrificing prosthesis	Knee function (ROM, KSS, WOMAC), postoperative complications, duration of surgery, length of hospital stay, theoretical blood loss	90 days	RA-TKA took more time than CM-TKA, likely due to the learning curve and intraoperative registration.	RA-TKA	37	64.5 ± 5.3	11 (29.7%)	16/21
								C-TKA	35	63.4 ± 7.2	7 (20%)	19/16
Xu et al 2022-2	China	End-stage osteoarthritis, Kellgren–Lawrence (KL) staging III–IV, no response to conservative treatment for >6 months	YUANHUA robot	Standard posterior substitution prosthesis	Clinical safety, and efficacy	Postoperative	Compared to CM-TKA, RA-TKA reduced trauma, shortened bone cutting and gap balancing time, minimized mechanical errors in osteotomy and prosthesis placement, and improved alignment accuracy. It also enhanced postoperative comfort and reduced inflammation and medication use.	RA-TKA	16	67.3 ± 3.5	3 (18.75%)	-
								C-TKA	16	66.6 ± 3.7	3 (18.75%)
Yuan et al 2024	China	End-stage knee osteoarthritis	YUANHUA-TKA	Cemented posterior stability total knee system	WOMAC and opioids consumption	12 months	Compared to cmTKA, raTKA with the YUANHUA robot reduced patient pain and trauma while enhancing functional recovery and improving the accuracy of prosthesis positioning and axial alignment reconstruction.	RA-TKA	28	65.2 (6.4	9 (28.13%)	-
								C-TKA	32	65.4 (8.0	4 (12%)

C-TKA: Conventional Total Knee Arthroplasty; CK: Creatine Kinase; CFCA: Coronal Femoral Component Angle; CRP: C-Reactive Protein; CTCA: Coronal Tibial Component Angle; EQ-5D: EuroQol 5-Dimension; EQ-VAS: EuroQol Visual Analogue Scale; ESR: Erythrocyte Sedimentation Rate; GNRI: Geriatric Nutritional Risk Index; HKA: Hip-Knee-Ankle Angle; HRQoL: Health-Related Quality of Life; IL: Interleukin; KSS: Knee Society Score; LDH: Lactate Dehydrogenase; Lysholm: Lysholm Knee Score; MASTI: Modified Anterior Soft Tissue Index; RA-TKA: Robotic-Assisted Total Knee Arthroplasty; SFCA: Sagittal Femoral Component Angle; STCA: Sagittal Tibial Component Angle; TNF-alpha: Tumor Necrosis Factor-alpha; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index.

Quality assessment

Two studies had a high risk of bias due to the selection of reported results^[11,18]. Five RCTs raised concerns due to insufficient details regarding randomization^[13,19], missing outcome data^[20], or the methods used for outcome measurement^[21,22]. The remaining 14 RCTs exhibited a low risk of bias, consistently demonstrating minimal bias across all five domains (Supplementary Figure 1, available at: http://links.lww.com/MS9/A712).

Operative time (minutes)

The pooled analysis of 10 RCTs indicated a significant longer operative time in the RA-TKA compared with C-TKA (MD = 19.94 [9.2, 30.68], P = 0.0003), but the data revealed unresolvable heterogeneity (P < 0.00001, I² = 98%; Fig. 2).

Figure 2.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: operative time (minutes).

Intra-operative blood loss (ml)

There was insignificant difference between RA-TKA in the intra-operative blood loss (MD = −41.72 [−100.07, 16.64], P = 0.16), but there was substantial heterogeneity that could not be resolved (P < 0.0001, I² = 80%; Fig. 3).

Figure 3.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: intra-operative blood loss (ml).

Change in WOMAC (10 days to 2 months)

There was insignificant difference between RA-TKA in the change in after 10 days to two months (MD = 0.47 [−4.97, 5.9], P = 0.87), and the data showed homogeneity (P = 0.36, I² = 6%; Fig. 4A).

Figure 4.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: change in WOMAC score. (A) Before resolving heterogeneity. (B) After resolving heterogeneity.

Also, there was insignificant difference between RA-TKA in the change in WOMAC after two to six months) (MD = 2.26 [−3.93, 8.45], P = 0.47), and the data showed homogeneity (P = 0.76, I² = 0%; Fig. 4A).

Moreover, there was insignificant difference between RA-TKA in the change in WOMAC after 12 to 36 months (MD = −0.32 [−1.54, 0.90], P = 0.60), but the data showed heterogeneity (P = 0.002, I² = 73%). Even after resolving heterogeneity by excluding Song et al 2013 (P = 0.28, I² = 21%), the results remained insignificant (MD = −1.54 [−4.47, 1.39], P = 0.30; Fig. 4A and 4B).

Change in OKS

There was insignificant difference between RA-TKA in the change in OKS after 6 months (MD = 0.62 [−0.26, 1.51], P = 0.17), and the data showed homogeneity (P = 0.78, I² = 0%; Fig. 5)

Figure 5.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: change in Oxford Knee Score (OKS).

Also, there was insignificant difference between RA-TKA in the change in OKS after (12 months to 24 months) (MD = 0.56, 95%CI [−0.25, 1.37], P = 0.18), and the data showed homogeneity (P = 0.36, I² = 6%; Fig. 5).

Mechanical alignment outlier rate

The pooled analysis showed a lower mechanical alignment outlier rate in the RA-TKA group compared with C-TKA (RR = 0.33 [0.19, 0.59], P = 0.0002), but the data were heterogenous (P = 0.03, I² = 52%). Also, after resolving heterogeneity by excluding Tian et al 2023 (P = 0.12, I² = 38%), the results stayed significantly lower in the RA-TKA group (RR = 0.41 [0.25, 0.66], P = 0.0003; Fig. 6).

Figure 6.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: mechanical alignment outlier rate. (A) Before resolving heterogeneity. (B) After resolving heterogeneity.

Post-operative KSS

The pooled analysis showed a higher post operative KSS score in the RA-TKA group compared with C-TKA (MD = 1.03 [0.50, 1.57], P = 0.0002), and the data showed homogeneity (P = 0.60, I² = 0%; Fig. 7).

Figure 7.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: post-operative Knee Society Score (KSS).

Post-operative ROM

There was insignificant difference between RA-TKA and C-TKA regarding postoperative ROM (MD = −0.34, 95%CI [−6.19, 5.51], P = 0.91), but the data showed unresolvable heterogeneity (P < 0.00001, I² = 99%; Fig. 8).

Figure 8.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: post-operative Range of Motion (ROM).

Deviation from neutral mechanical axis

The pooled analysis showed a less deviation from neutral mechanical axial in the RA-TKA group compared with C-TKA (MD = −0.93 [−1.20, −0.66], P < 0.00001), and the data showed homogeneity (P = 0.50, I² = 0%; Fig. 9).

Figure 9.

Forest plot of comparison: RA-TKA versus C-TKA, outcome: deviation from neutral mechanical axis.

Discussion

Our systematic review and meta-analysis of 21 RCTs comparing RA-TKA to C-TKA revealed several key findings. RA-TKA demonstrated significantly lower mechanical alignment outlier rates and less deviation from neutral mechanical axis compared to C-TKA. This suggests that robotic assistance may enhance surgical precision in achieving optimal limb alignment^[23]. However, we found no significant differences between the two approaches in functional outcomes as measured by WOMAC and OKS at various follow-up time points. This indicates that the improved alignment accuracy with RA-TKA may not necessarily translate to superior clinical outcomes in the short to medium term.

RA-TKA was associated with longer operative times but showed no significant difference in intraoperative blood loss compared to C-TKA. The extended surgical duration may be attributed to the additional steps required for robotic setup and intraoperative registration^[24]. However, the similar blood loss suggests that the robotic technique does not compromise hemostasis despite the longer procedure.

We observed higher postoperative KSS in the RA-TKA group, although the clinical significance of this small difference is questionable. The lack of significant difference in postoperative range of motion between the two techniques suggests that robotic assistance may not provide substantial advantages in this aspect of knee function.

These results could be explained by the learning curve associated with robotic technology, variations in surgical protocols across studies, and the possibility that conventional techniques, when performed by experienced surgeons, may achieve comparable outcomes to robotic assistance in many aspects of TKA^[25].

The absence of significant improvements in WOMAC and OKS scores, despite improved mechanical alignment, could be attributed to several factors. First, functional outcomes are multifaceted and may not directly correlate with mechanical precision alone. Soft tissue balance, muscle strength, patient-specific factors, and rehabilitation protocols might play equally critical roles in functional recovery. Additionally, experienced surgeons performing conventional TKA may already achieve near-optimal alignment, potentially minimizing the incremental benefit of robotic assistance in short to medium-term outcomes.

Our findings are largely consistent with those of earlier meta-analyses, with some notable differences. Ren et al (2019) similarly observed no significant alterations in functional outcomes between RA-TKA and C-TKA, aligning with our results on WOMAC and OKS^[26]. They also reported improved mechanical alignment with robotic assistance, corroborating our findings. However, unlike our study, they observed lower postoperative drainage volumes in the RA-TKA group.

Alrajeb et al (2024) reported results that closely mirror ours, particularly regarding the superiority of RA-TKA in restoring mechanical alignment^[6]. Their study also found no significant differences in functional scores and range of motion, consistent with our findings. However, they noted a trend toward better range of motion in RA-TKA, which was not observed in our analysis.

Ruangsomboon et al (2023) reported findings very similar to ours, including improved mechanical alignment and longer operative times with RA-TKA^[6]. They also observed no significant alterations in most patient-reported outcome measures, aligning with our results. However, their analysis showed lower blood loss in RA-TKA, which contrasts with our finding of no significant difference in this parameter. Given the common use of tourniquets in total knee arthroplasty, future studies should consider measuring hidden blood loss and total blood loss to provide a more comprehensive assessment of intraoperative bleeding, which was not possible within the scope of this meta-analysis.

The consistent finding across our study and previous meta-analyses of improved mechanical alignment with RA-TKA suggests that this technology offers a tangible benefit in terms of surgical precision. This could potentially lead to improved long-term outcomes and implant longevity, although longer follow-up studies are necessary to confirm this hypothesis.

The lack of significant differences in short to medium-term functional outcomes between RA-TKA and C-TKA implies that the clinical benefits of improved alignment may not be immediately apparent. This raises questions about the cost-effectiveness of robotic systems, given their substantial initial investment and ongoing maintenance costs^[27]. Healthcare systems and institutions considering the adoption of this technology should weigh these factors carefully.

The longer operative times associated with RA-TKA have implications for surgical efficiency and resource utilization. As surgeons gain more experience with robotic systems, these times may decrease, but the initial learning curve should be considered when implementing this technology^[28].

Our findings suggest that while robotic assistance may offer some advantages in TKA, particularly in terms of alignment accuracy, it may not be superior to conventional techniques in all aspects of the procedure and its outcomes. This underscores the importance of surgeon skill and experience, regardless of the technique used^[29].

While our analysis included studies with follow-up periods ranging from 6 weeks to 10 years, most studies concentrated on short to medium-term outcomes (typically up to 2 years). We define long-term outcomes as those assessed beyond 5 years post-surgery. The potential long-term benefits of improved mechanical alignment, such as reduced implant wear and extended prosthesis survival, remain speculative and warrant dedicated longitudinal studies specifically designed to evaluate these parameters. However, we did not conduct a long-term analysis in this study due to insufficient available data.

Our study has several strengths, including a comprehensive search strategy, inclusion of only randomized controlled trials, and a focus on clinically relevant outcomes. The large number of included studies and patients enhances the generalizability of our findings. Additionally, our analysis considered various follow-up time points, providing insights into both short and medium-term outcomes.

However, our study also has limitations. The heterogeneity in outcome measures and follow-up periods across studies made some comparisons challenging. The absence of long-term follow-up data in most studies restricts our capability to assess the probable long-term benefits of the improved alignment achieved with RA-TKA. Additionally, the rapid evolution of robotic technology means that some earlier studies may not reflect the capabilities of current systems.

In conclusion, our meta-analysis suggests that RA-TKA offers better mechanical alignment accuracy compared to C-TKA, but this does not translate to superior short to medium-term functional outcomes. The technology is associated with longer operative times but does not significantly impact intraoperative blood loss. Based on these findings, we recommend that the decision to adopt robotic assistance in TKA should be made cautiously, considering factors such as cost, learning curve, and the specific needs of the surgical team and institution. Future research should spotlight on long-term outcomes, cost-effectiveness analyses, and the potential benefits of RA-TKA in complex cases or revision surgeries.

Ethical approval

Not applicable.

Consent

Not applicable.

Sources of funding

None.

Author’s contribution

O.M. and M.M. equally contributed to this paper and joined first authorship is proposed. Study concept and design: K.Q., S.Q. Acquisition of data: O.M., M.M., K.Q., U.I. Analysis and interpretation of data: O.M., M.M., K.Q., U.I. Drafting of manuscript: A.K., A.W., Z.A. Critical revision of manuscript: A.K., A.W., Z.A., S.Q.. Final approval: all authors.

Conflicts of interest disclosure

None.

Research registration unique identifying number (UIN)

INPLASY database ID: INPLASY202490120.

Guarantor

Sohail Quraishi.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Data availability statement

No other datasets were generated during and/or analyzed during the current study. All the information is available with the manuscript.

Assistance with study

None.

Presentation

None.