Table 3.
Clinical outcomes after robotic-assisted total knee arthroplasty.
| Study | Year | Studies (#) | System | Robotic cases (#) | Conventional cases (#) | Follow-up in months | Conclusions | Level of evidence | Grade of recommendation |
|---|---|---|---|---|---|---|---|---|---|
| Song et al47 | 2013 | 1 | ROBODOC | 50 | 50 | 41 | No differences in postoperative ROM, WOMAC scores, and HSS knee score. | 1 | A |
| Liow et al28 | 2014 | 1 | ROBODOC | 31 | 29 | 6 | At 6-month follow-up, there was no overall difference in terms of clinical outcome measures, except in SF-36 vitality scores, where the robot-assisted group reported higher vitality scores. | 1 | A |
| Liow et al88 | 2017 | 1 | ROBODOC | 31 | 29 | 24 | Both robotic and conventional TKA displayed significant improvements in majority of the functional outcome scores at 2 years. Despite having a higher rate of complications, the robotic-assisted group displayed a trend towards higher scores in SF-36 QoL measures, with significant differences in SF-36 vitality (p = 0.03), role emotional (p = 0.02) and a larger proportion of patients achieving SF-36 vitality MCID (48.4% vs. 13.8%, p = 0.009). No significant differences in KSS, OKS or satisfaction/expectation rates were noted. | 1 | A |
| Kim et al91 | 2020 | 1 | ROBODOC | 975 | 990 | 120 | At a minimum follow-up of 10 years, there were no differences between robotic-assisted TKA and conventional TKA in terms of functional outcome scores, aseptic loosening, overall survivorship, and complications. Considering the additional time and expense associated with robotic-assisted TKA, we cannot recommend its widespread use. | 1 | A |
| Kayani et al107 | 2018 | 1 | Mako | 40 | 40 | 1 | Robotic-arm assisted TKA was associated with reduced postoperative pain (p < 0.001), decreased analgesia requirements (p < 0.001), decreased reduction in postoperative haemoglobin levels (p < 0.001), shorter time to straight leg raise (p < 0.001), decreased number of physiotherapy sessions (p < 0.001) and improved maximum knee flexion at discharge (p < 0.001) compared with conventional jig-based TKA. Median time to hospital discharge in robotic TKA was 3.2 days compared with 4.4 days in conventional TKA (p < 0.001). | 2 | B |
| Khlopas et al76 | 2020 | 1 | Mako | 150 | 102 | 3 | This prospective, non-randomized, open-label, multicentre comparative cohort study found robotic TKA patients to have equal or greater improvements in 9 out of 10 of the Knee Society Scoring System components assessed at 3 months postoperatively, though not all findings were statistically significant. | 2 | B |
| Ren et al87 | 2019 | 7 | ROBODOC CASPAR |
315 | 262 | 16–120 | Seven studies with a total of 577 knees undergoing TKA were included. Compared with conventional surgery, active robotic TKA showed better outcomes in precise mechanical alignment (p < 0.05) and implant position, with lower outliers (p < 0.05), better functional score (WOMAC functional score) and less drainage (p < 0.05). | 2 | B |
| Siebert et al57 | 2002 | 1 | CASPAR | 70 | 50 | – | The mean difference between preoperatively planned and postoperatively achieved tibiofemoral alignment was 0.8 degrees (0–4.1 degrees) in the robotic group vs. 2.6 degrees (0–7 degrees) in a manually operated historical control group of 50 patients. The authors observed reduced postoperative soft tissue swelling in the robotic cohort. | 3 | B |
| Park et al42 | 2007 | 1 | ROBODOC | 32 | 30 | 45 | Roughly 70% of conventional TKA gives a MA alignment of less than ± 3° as compared to more than 90% with navigation TKA. Six of the 32 active robotic TKA procedures had short-term complications including superficial infection, patellar tendon ligament rupture, patella dislocation, supracondylar fracture and patellar fracture and common peroneal injury. | 3 | B |
| Song et al48 | 2011 | 1 | ROBODOC | 15 | 15 | 16 | Radiographic results showed significantly more postoperative leg alignment outliers of conventional sides than robotic-assisted sides (mechanical axis, coronal inclination of the femoral prosthesis, and sagittal inclination of the tibial prosthesis). Robotic-assisted sides had non-significantly better postoperative knee scores (HSS, WOMAC, side preference) and ROMs. | 3 | B |
| Clark et al30 | 2013 | 1 | OMNIBotic | 52 | 29 | 1 | Robotic navigation times were, on average, 9.0 minutes shorter compared to computer navigation. The average absolute intraoperative malalignment was 0.5° less in the robotic procedures compared to the computer-navigation procedures. Patients in the robotic TKA group tended to be discharged 0.6 days earlier compared to patients in the computer-navigated TKA. | 3 | B |
| Yang et al89 | 2017 | 1 | ROBODOC | 71 | 42 | 120 | Clinical outcomes and long-term survival rates were similar between the two groups. Regarding the radiological outcomes, the robotic TKA group had significantly fewer postoperative leg alignment outliers (femoral coronal inclination, tibial coronal inclination, femoral sagittal inclination, tibial sagittal inclination, and mechanical axis) and fewer radiolucent lines than the conventional TKA group. | 3 | B |
| Marchand et al108 | 2017 | 1 | Mako | 20 | 20 | 6 | The mean physical function scores for the manual and robotic cohorts were 9 ± 5 and 4 ± 5, p = 0.055, respectively. The mean total patient satisfaction scores for the manual and robotic cohorts were 14 points and 7 points, p < 0.05, respectively. The results from this study highlight the potential of the Mako to improve short-term pain, physical function, and total satisfaction scores. | 3 | B |
| Kayani et al85 | 2018 | 1 | Mako | 30 | 30 | – | Patients undergoing RA-TKA had reduced medial soft tissue injury in both passively correctible (p < 0.05) and fixed varus deformities (p < 0.05); more accurate femoral (p < 0.05) and tibial (p < 0.05) bone resection cuts; and improved macroscopic soft tissue injury (MASTI) scores compared to conventional TKA (p < 0.05). | 3 | B |
| Cho et al90 | 2019 | 1 | ROBODOC | 155 | 196 | 132 | All clinical assessments showed excellent improvements in both robotic and conventional TKA cohorts (all p < 0.05), without any significant differences between the groups (p > 0.05). The conventional TKA group showed a significantly higher number of outliers compared with the robotic TKA group (p < 0.05). The cumulative survival rate was 98.8% in robotic TKA and 98.5% in the conventional group (p = 0.563). | 3 | B |
| Cool et al95 | 2019 | 1 | Mako | 519 | 2595 | 3 | Overall 90-day episode-of-care costs were US$2,391 less for robotic TKA (p < 0.0001). Over 90% of patients in both cohorts utilized post-acute services, with robotic TKA accruing fewer costs than manual TKA. Savings were driven by fewer readmissions and an economically beneficial discharge destination. | 3 | B |
| Malkani et al86 | 2020 | 1 | Mako | 188 | 188 | 24 | Patients undergoing RA-TKA experienced a significant, 4.5-fold decrease in rates of MUA (p = 0.032). Given that MUAs can be a marker of knee stiffness following total knee arthroplasty, the lower rate indicates that study cohort patients had less knee stiffness and, therefore, greater initial postoperative ROM than the control cohort. | 3 | B |
| Marchand et al75 | 2019 | 1 | Mako | 53 | 53 | 12 | The RA-TKA cohort had significantly improved mean total (6 ± 6 vs. 9 ± 8 points, p = 0.03) and physical function scores (4 ± 4 vs. 6 ± 5 points, p = 0.02) when compared with the manual cohort. The mean pain score for the RA-TKA cohort (2 ± 3 points [range, 0–14 points]) was also lower than that for the manual cohort (3 ± 4 points [range, 0–11 points]) (p = 0.06). RA-TKA was found to have the strongest association with improved scores when compared with age, gender, and BMI. This study suggests that RA-TKA patients may have short-term improvements at minimum 1-year postoperatively. | 3 | B |
| Sultan et al46 | 2019 | 1 | Mako | 43 | 39 | 1.5 | RA-TKA patients had smaller mean differences in posterior condylar offset ratio which has been previously shown to correlate with better joint ROM at one year following surgery. In addition, these patients were less likely to have values outside of normal Insall-Salvati Index, which means they are less likely to develop patella baja, leading to restricted flexion and overall decreased ROM. | 3 | B |
| Khlopas et al76 | 2020 | 1 | Mako | 150 | 102 | 3 | At 4 to 6 weeks postoperatively and at 3 months, RA-TKA patients were also found to have larger improvements in walking and standing, standard activities, advanced activities, functional activities total score, pain with walking, total symptoms score, satisfaction score, expectations score when compared with manual TKA patients. | 3 | B |
| Bellemans et al43 | 2007 | 1 | CASPAR | 25 | – | 66 | Results demonstrate excellent implant positioning and alignment was achieved within the 1° error of neutral alignment in all three planes in all cases. Despite this technical precision, the excessive operating time required for the robotic implantation, the technical complexity of the system, and the extremely high operational costs have led the authors to abandon the robotic system. | 4 | C |
| Marchand et al80 | 2018 | 1 | Mako | 330 | – | – | All 132 knees with initial varus deformity of less than 7° were corrected to neutral (mean 1°, range –1–3°). A total of 82 knees (64%) with 7° or greater varus deformity were corrected to neutral (mean 2°, range 0–3°). However, roughly 30% of patients with severe deformity who were not corrected to neutral were still corrected within a couple of degrees of neutral. There were seven knees with 7° or greater valgus deformity, and all were corrected to neutral (mean 2°, range 0–3°). This study demonstrated that all knees were corrected in the appropriate direction within a few degrees of neutral, and no knees were overcorrected. | 4 | C |
| Sodhi et al74 | 2018 | 1 | Mako | 20 | 20 | – | Surgeon 1: First and last robotic cohort operative times were 81 and 70 minutes (p < 0.05). Mean operative times for the first 20 robotic-assisted cases and manual cases were 81 versus 68 minutes (p < 0.05). Mean operative times for the last 20 robotic-assisted cases and manual cases were 70 versus 68 minutes (p > 0.05). Surgeon 2: First and last robotic cohort operative times were 117 and 98 minutes (p < 0.05). Mean operative times for the first 20 robotic-assisted cases and manual cases were 117 versus 95 (p < 0.05). Mean operative times for the last 20 robotic-cohort cases and manual cases were 98 versus 95 (p > 0.05). The data from this study effectively create a learning curve for the use of robotic-assisted TKA. As both surgeons completed their total cases numbers within similar time frames, these data imply that within a few months, a board-certified orthopaedic joint arthroplasty surgeon should be able to adequately perform robotic TKA without adding any operative times. | 4 | C |
| Sodhi et al77 | 2019 | 1 | Mako | 3 | – | – | Three cases (femoral and tibial fracture malunion, proximal tibial fracture nonunion, healed tibial plateau fracture) of patients who underwent RA-TKA in the setting of preoperative extra-articular deformities were identified. Utilizing preoperative CT scans with a 3D plan for robotic-arm assisted surgery allowed for appropriate assessment of the deformity preoperatively and execution of a plan for a balanced and aligned total knee arthroplasty. The study reported excellent results utilizing RA-TKA in these complex cases. |
4 | C |
| Marchand et al78 | 2019 | 1 | Mako | 335 | – | 1 | For 98% of prostheses, RA-TKA software predicted within 1 implant size the actual tibial or femoral implant size used. The mean length of stay was found to be 2 days. No patients suffered from superficial skin infection, pin site infections or fractures, soft tissue damage, and no robotic cases were converted to manual TKA due to intraoperative complications. A total of 8 patients (2.2%) were readmitted; however, none were directly related to robotic use. The robotic software and use of a preoperative CT substantially helped with intraoperative planning and accurate prediction of implant sizes. | 4 | C |
| Khlopas et al83 | 2017 | 1 | Mako | 6 | 6 | – | During bone resections, the tibia in RA-TKA procedures did not require subluxation, which may reduce ligament stretching or decrease complication rates. Potential patient benefits for short-term recovery and decreased morbidity to reduce operative complications should be studied in a clinical setting. Since RA-TKA uses a stereotactic boundary to constrain the sawblade, which is generated based on the implant size, shape, and plan, and does not have the ability to track the patient’s soft tissue structures, standard retraction techniques during cutting are recommended. Therefore, the retractor placement and potential for soft tissue protection needs to be further investigated. RA-TKA has the potential to increase soft tissue protection when compared to manual TKA. | 5 | I |
| Parratte et al65 | 2019 | 1 | ROSA | 15 | 15 | – | 15 cadaveric specimens were used with 15 knees undergoing TKA with computer navigation and 15 knees undergoing robotic TKA with ROSA. The target angles obtained from the intraoperative planning were then compared to the angles of the bone cuts performed using the robotic system and measured with the computer-assisted system considered to be the gold standard. All angle mean differences were below 1° and standard deviations below 1°. For all 6 angles, the mean differences between the target angle and the measured values were not significantly different from 0 except for the femoral flexion angle which had a mean difference of 0.95°. The mean hip-knee-ankle axis difference was –0.03° ± 0.87°. All resection mean differences were below 0.7 mm and standard deviations below 1.1 mm. | 5 | I |
| Hampp et al81 | 2019 | 1 | Mako | 12 | 12 | – | Significantly less damage occurred to the PCLs in the RA-TKA versus the manual TKA specimens (p < 0.001). RA-TKA specimens had non-significantly less damage to the deep medial collateral ligaments (p = 0.149), iliotibial bands (p = 0.580), poplitei (p = 0.248), and patellar ligaments (p = 0.317). The remaining anatomical structures had minimal soft tissue damage in all manual TKA and RA-TKA specimens. These findings are likely due to the enhanced preoperative planning with the robotic software, the real-time intraoperative feedback, and the haptically bounded saw blade, all of which may help protect the surrounding soft tissues and ligaments. | 5 | I |
Note. TKA, total knee arthroplasty; (#), number; ROM, range of motion; KSS, Knee Society score; WOMAC, Western Ontario and McMaster Universities osteoarthritis index; SF-36, Short Form-36; RA-TKA, robotic-assisted TKA; PCL, posterior cruciate ligament; HSS, Hospital for Special Surgery; QoL, quality of life; MCID, minimal clinically important difference; OKS, Oxford Knee Score; MA, mechanical axis; MUA, manipulation under anesthesia; BMI, body mass index; CT, computed tomography