Introduction

Total Knee Arthroplasty (TKA) is a standard established procedure for patients with end stage (grade 4) arthritis [1,2,3,4]. Despite its success, literature reports dissatisfaction in about 20% of TKA patients [5,6,7,8,9,10,11]. The use of computer and robotic-assisted (RA) -TKA is on the rise. The utilization of RA-TKA facilitates precise bone excision and enables correct preoperative prediction of femur and tibial implant sizes [12, 13]. This technology also aids in obtaining the proper implant and limb alignment. Robotic-assisted TKA in comparison to traditional TKA has also been shown to reduce the incidences of iatrogenic bone and periarticular soft tissue injury [12, 14, 15]. The aforementioned reasons help in achieving higher levels of post-surgical patient satisfaction and desired clinical results. It also helps in improving the long-term implant survivorship [15,16,17].

The most important inhibition in the arthroplasty surgeon's mind about adapting the RA technology for performing TKA is the learning curve associated and consequently the extra time he/she going to spend doing the RA-TKA procedure. A few studies are available about the learning curve estimation, about the use of semiautomatic robot in performing RA-TKA [18, 19]. Cuvis joint robotic system (Curexo, Korea, supported by Meril Life Sciences, Pvt. Ltd, India) is fully automatic robotic TKA system. It is US-FDA approved and European CE mark, and Ministry of Food and Drug Safety of the Republic of Korea approved joint surgical robot [20]. It uses a pre-operative 3-dimensional (3-D) computerized tomography (CT) scan to create individual patient-specific bone model. The surgeon along with the software engineer plans the TKA procedure pre-operatively choosing accurate implant size/ position and limb alignment (Hip-Knee-Ankle axis) of 180°. In this system, after registration and confirmation of femur and tibia bony points, the fully automatic robot is attached to the patient and all femur and tibia cuts are done with the help of bone burr which mills the bone. To the best of our knowledge, no research has been conducted on the estimation of the learning curve for the fully autonomous Cuvis Joint robot.

The primary goal of the present study is to investigate the robot-specific operative time associated with RA-TKA using a fully automated Cuvis Joint robot. The secondary aim of the study was to assess the anxiety levels of the surgical team, the accuracy of the implant size/ position and alignment of the limb and to study post-operative complications if any.

Methods

This is a prospective, single-center study that included 40 patients undergoing fully automatic RA-TKA. The surgical team included experienced surgeons who regularly performed conventional TKA and were provided saw-bone training utilizing RA-TKA. The inclusion criterion was patients with symptomatic end stage arthritis requiring RA-TKA while the exclusion criteria were conversion of unicompartmental arthroplasty to TKA and conversion of high tibial osteotomy to TKA. The study had an ethics committee approval for the participating institute. Written informed consent was obtained from all patients enrolled for the study. All 40 patients were implanted with posterior stabilized high flexion Freedom total knee system (MAXX Orthopedics, Inc., Plymouth Meeting, Pennsylvania) utilizing a fully automatic Cuvis Joint robot. Routine pre-operative anteroposterior and lateral knee radiographs, hip-knee-ankle (HKA) scanograms along with pre-operative CT scans of the operated leg was performed in all the patients. The axial plane was utilized to conduct scans of the leg in the regions of hip, knee and ankle and the images were then transferred in the.jpg format. The pre-operative planning process relied on the mechanical alignment approach rather than the kinematic alignment concept. The desired alignment of the limb was neutral with an HKA axis of 180°. The surgical team along with the system software engineer did the surgical planning choosing the best fit femur and tibia implant achieving HKA angle of 180°. Particular attention was given to make sure that there was no femur implant overhang or notching and posterior condylar offset ratio is restored. As regards the tibial component, the native posterior tibial slope was restored making sure that there is no mediolateral or anteroposterior overhang. A mechanical alignment philosophy was used in planning the limb alignment. The HKA angle of 178° to 182° was deemed to be acceptable alignment. The surgical time specifically related to the active RA-TKA was defined as surgeon-dependent time (time for the insertion of the femoral and tibial registration pins + time for the bony registration) and surgeon-independent/ active robot-related time (time taken for docking the active robot to the patient + time taken for bone resection by bone milling performed by an active robot). The anxiety level of the surgical team (comprising 2 assistant orthopaedic surgeons headed by one senior surgeon) was assessed utilizing the Spielberger State-Trait Anxiety Inventory (STAI) questionnaire, a well-established and validated subjective assessment tool. This approach measures an individual's stress levels by considering the specific qualities that emerge from the clinical situation. The questionnaire designed by Marteau et al. consists of six items and is a 4-point-based system. The total scores on this questionnaire range from 6 to 24 [21]. Scores are directly proportional to stress levels; higher level stress is indicated by higher scores. Prior to the commencement of the surgery, all members of the surgical team took the STAI assessment. Accuracy of the implant size/ position and alignment of the limb was evaluated by the post-operative weight-bearing lateral and anteroposterior radiographs and full-length scanograms. An external observer, independent of the surgical team, evaluated the surgery duration, the surgeon's anxiety levels, and the precision of implant size and positioning, accuracy of limb alignment, as well as the occurrence of any post-surgical complications. Statistical analysis was carried out by student’s t-test for comparison of means and p-value < 0.05 was considered statistically significant (Fig. 1).

Fig. 1
figure 1

provides observation where beyond case 18 i.e., cohort 2 operative time seems to be lower

Full size image

Results

For convenience, the patients were grouped into 4 cohorts of 1–10, 11–20, 21–30 and 31–40. The patient’s baseline characteristics are shown in Table 1 and there was no significant difference in the pre-operative patient's attributes in all four cohorts.

Table 1 Comparison of preoperative patient characteristics between all four patient group cohorts
Full size table

The robot-specific operative time (mean ± standard deviation) for the four patient cohorts was: cases 1–10: 27.3 ± 3.83, cases 11–20: 24.8 ± 4.05, cases 22–30: 24.9 ± 3.30 and cases 31–40: 25.5 ± 2.42, respectively (Table 2).

Table 2 Comparison of surgical time (surgeon dependent + surgeon independent) specifically related to the active robotic-assisted TKA
Full size table

The operative cases 1–10 took longer time than 11–20, 21–30 and 31–40 cases, p-value 0.0170, statistically significant. From cohort 2, there was no statistically significant difference as regards the operative time between cohort 2, 3 and 4 (p-value 0.9280 for cohort 2 -3 and p-value 0.4282 for cohort 3–4) (Table 3).

Table 3 Case-wise surgical time specifically related to the active robotic-assisted TKA
Full size table

Assessment of the anxiety levels (Fig. 2) revealed highest STAI scoring in cohort 1 (cases 1–10) than cohort 2 (cases 11–20) p-value 0.0003, reducing in cohort 3 (cases 21–30) p-value < 0.0001 (cohort 2 vs. cohort 3) and lowest with cohort 4 with p-value 0.0233 when compared to cohort 3 (Table 4).

Fig. 2
figure 2

Represents the case-wise presentation of the anxiety level of the surgeons specifically related to the active robotic-assisted TKA

Full size image
Table 4 Comparison of the surgical team’s anxiety levels according to the six-item questionnaire score
Full size table

Analysis of the post-operative radiographs revealed no oversizing or undersizing of the femur or tibia implant. There were no outliers on the post-operative scanograms as regards the HKA angle (acceptable range 180° ± 2°). One patient in cohort 3 developed a wound dehiscence which was managed by debridement of wound edges and secondary suturing.

Discussion

The main findings of our study are that robot-specific operative times were longer in the first 10 cases. The operative cases 1–10 took longer time than 11–20, 21–30 and 31–40 cases. Beyond case no 18, there was no statistically significant difference as regards the operative time between cohort 2, 3 and 4. This indicates an inflection point of the initial 18 cases (cohort 2) where the surgeon progresses from the learning to the proficiency phase of the active RA-TKA procedure. The surgical team anxiety level also progressively decreased from cohort 1 to 4, indicating surgical team becoming familiar with the adaptation of the active RA technology and thereby getting accustomed to the technique of active RA-TKA. As regards the secondary objective the cumulative active RA-TKA experience did not affect the accuracy of the implant size and positioning, achieved acceptable limb alignment (178–182°) with the absence of surgical complications.

Determination of a learning curve for a particular surgical technique is extremely important [22]. Learning curve indicates the time taken to adopt a new surgical technique and the number of repetitions to be done by the surgeon to become comfortable with a particular technology/ surgical procedure. Assessment of the learning curve also helps in analyzing the complications/ difficulties faced by the surgical team during the period of learning [23, 24]. It also helps the operating surgeon to do his/her own risk vs. benefit analysis to decide about the use of a particular technology/procedure in the surgical practice. Orthopedic surgeons have always been inclined towards fast adoption of the new emerging surgical technology [25]. Between 2008 and 2015 there was a threefold increase in the adoption of computer-assisted technology [26]. Researchers, biomedical engineers, and orthopedic surgeons have continually looked for methods to improve patient-oriented outcomes measures. This is necessitated by the fact that approximately 20% of TKA patients are not satisfied with the post-TKA results. The utilization of RA-TKA enables the surgeon to do preoperative templating and planning, specifically in relation to the sizing and placing of the femur and tibial implants. This includes ensuring precise alignment of the limb with the 180° HKA axis. Robotic-assisted TKA carries out bone resections with sub-millimeter accuracy. The primary source of anxiety and tension for surgeons when considering the implementation of a novel surgical method is the apprehension regarding a potential increase in operative times and decrease in efficiency. Several studies have reported varying learning curves for the Mako, OMNIBotics, and NAVIO robotics systems in the context of RA-TKA [27, 28]. These studies have identified learning curves of 7, 7, and 12 cases for the respective systems [27, 28].

The robot-specific operative time in our study decreased from cohort 2 onwards. This is because the surgical and nursing team got familiar with the steps and workflow of the active RA-TKA procedure. The surgical time specifically related to the active RA-TKA consisted of surgeon-dependent time (time for the insertion of the femoral and tibial registration pins + time taken by surgeon for the bony registration) and surgeon independent/ active robot-related time (time taken for docking the active robot to the patient + time taken for bone resection by bone milling performed by an active robot). In our study, the most significant improvement happened with surgeon-dependent time indicating less time taken by the surgeon for insertion of registration pins and time taken for femur and tibia bony registration. The anatomical bony registration landmarks (femur and tibia) were the same in all the patients. Hence with incremental surgical experience, the surgeon progressively took less time for the bony registration. In relation to the duration of the surgeon's independent and active robot-assisted procedures, which encompasses the time required for docking the active robot to the patient and the time taken for bone resection through active robot-performed bone milling, a moderate enhancement was observed as the surgical and nursing team grew more familiar with the surgical technique workflow. The results of our investigation align with the research conducted by Sodhi et al. [19] and Mahure et al. [29]. In a study conducted by Sodhi et al. it was demonstrated that the learning curve for performing RA-TKA consisted of 20 cases [19]. Conversely, Mahure et al. found that active RA-TKA exhibited a somewhat shorter learning curve ranging from 10 to 20 cases [29].

Higher levels of anxiety and stress observed in the surgical team for the initial 10 RA-TKA cases are revealed in the high scores obtained for the STAI questionnaire. This is significant as elevated mental stress and strain affects the decision-making ability and impairs technical skills thereby resulting in suboptimal operative performance [21]. The findings of our study demonstrated a gradual decrease in surgical anxiety levels throughout cohort 1 to cohort 4, suggesting an increasing level of proficiency within the surgical team in terms of both technical aspects and the overall understanding of the workflow of active RA-TKA. Another important finding of our study is that the initial surgical learning curve and the high surgical team anxiety levels did not affect the secondary outcomes measures. The possible explanation for the same may be certain unique features of the active RA-TKA namely the pre-operative planning carried out on the 3-D CT scan generated bone model, the accurate tracking of the femur and tibia bone resection windows within stereotactic boundaries thereby minimizing the chances of bone resection errors and soft tissue injuries. These unique features might have helped in achieving high accuracy in implant size/ position as well as limb alignment.

Limitations

The primary constraint pertains to the surgical team participating in the study, as they possess prior familiarity with computer-assisted (CAS) TKA. Consequently, it is inappropriate to generalize their surgical time findings to surgical teams lacking exposure to CAS. One notable constraint of this study is being a single center investigation and a small sample size. To advance the current body of knowledge, it is imperative to conduct a multicenter investigation on the utilization of active RA-TKA. A third drawback of the study was the failure to include the additional expenses associated with the utilization of the active RA-TKA technology.

Conclusion

The use of fully automatic active RA-TKA led to increased operative time (specific to robot) as well as increased anxiety level of the surgeon in the first 17 and 18 cases, respectively. The study showed progressive improvement in the surgical anxiety scores and a reduction in the operating time from cohort 2, indicating the proficiency of the surgical team. The initial prolongation of surgical duration and the elevated levels of anxiety experienced by the surgical team did not have a significant impact on the precision of implant size and positioning, limb alignment, and post-surgical complications.