1 Introduction

Robotics allowed a broader introduction of personalized alignment techniques in the OR due to increased technical precision.^1–3 Nevertheless, the topic of alignment is still challenging as a lot of different philosophies currently coexist. Beside measured resection techniques, such as unrestricted and restricted kinematic alignment (uKA and rKA), gap balanced techniques, such as inverse kinematic alignment (iKA) and functional alignment (FA) exist.^4–10 All these techniques are based on clear definitions, however, in daily practice surgeons often modify them to a more surgeon-specific alignment workflow. The reason for this modification is showing the complexity and limited success of alignment training to date.

A similar problem arises when it comes to the topic of training on balancing goals. Traditionally the aim was to create equally balanced extension and flexion gaps (Insall). A more recent approach mainly introduced by KA surgeons was restoring lateral flexion gap laxity.^10,11 McEwen et al. showed an increased patient satisfaction if this goal is applied. A third group of surgeons is aiming for a larger, but balanced flexion gap.¹² All these approaches have in common that these are systematic approaches, as they aim for one identical laxity goal in all types of patients independent of the individual laxity phenotype. Graichen et al. and others however, showed recently, that various laxity phenotypes exist and therefore argued for personalized reconstruction of gaps too.^13–15 Consequently this will add even more complexity to the topic of alignment training.¹³

Training of optimal alignment is performed by scientific and educational societies as well as by MedTech companies in order to assure process quality in theatres. Efficiency is another important aspect of this training as one persistent argument against personalized alignment with robotic surgery is increased process time. As theoretical training without practical parts is very challenging for most of the surgeons, additional practical training options have been discussed. One of them is cadaver training which has shown to be helpful in training of the robotic planning software as well as of the robotic execution. The limitation of this kind of training is that most of the cadavers show up with healthy and non-deformed knees. Furthermore, the individual bony phenotype of these cadavers is typically not known. To overcome these limitations, Graichen et al. introduced a TKA simulator some years back.¹⁴ This allows scientific training of different alignment workflows independent of MedTech companies.¹⁴ In consequence, it has become a widely accepted simulator, used by scientific and educational societies and Institutes. By integrating specific implant and insert designs, the simulator was also used by different MedTech companies to train surgeons on their specific implants and robots.

In a previous article an increase of decision quality and efficiency was described by using a TKA simulator.¹⁴ Both findings were based on early results obtained during the first courses. In the past 3 years many simulator training courses have been performed and the algorithms have been adapted according to the feedback of participants. The aim of this study was therefore to analyse (1) whether the mid-term results in terms of decision quality and efficiency have changed and (2) whether the limitations of previous training courses have been solved.

2 Material and methods

2.1

2.1 TKA simulator

All training courses included in this study used the Knee-Computational Alignment Trainer (CAT) simulator. This simulator allows for training of the following alignment workflows: Mechanical alignment (MA) – tibia first and femur first; Adjusted Mechanical Alignment (AMA) tibia first; Anatomical Alignment (AA); inverse kinematic alignment (iKA); Patient specific Technique (PST); unrestricted kinematic alignment (uKA), restricted kinematic alignment (rKA); functional alignment (FA)-tibia first and femur first. Fig. 1 represents the planning interface that participants would see during the resection plane assessment stage.

Close

Fig. 1 Knee-CAT planning interface for resection plane assessment.

Export to PPT

The following balancing goals are optional: 1) All gaps equal; the optimal gap size being 2 mm. 2) Lateral flexion gap laxity; meaning extension gap medial and lateral as well as medial flexion gap being optimal with 2 mm and simultaneously lateral flexion gap 4 mm or in KA/rKA up to 8 mm. 3) Flexion gap larger than extension gap, defined optimum extension gap being 2 mm and flexion gap 4 mm. 4) Medial Pivot goal: Medial extension gap 1.5 mm, lateral extension gap 2,0 mm, medial flexion gap 3.5 mm and lateral flexion gap 5 mm or larger. Fig. 2 represents an example of coronal resection plane adjustment.

Close

Fig. 2 Knee-CAT coronal resection plane adjustment.

Export to PPT

Seven symmetric and 1 asymmetric implant design with CR/PS or medial pivot inserts can be selected for training. The original simulation data with regards to resection thickness and gap data have been adapted to all implant specifications for the distal and posterior dimensions.

Individualized training courses have been designed with respect to the needs of the team organizing a specific course. Basic courses were created for training of basic principles in MA, such as joint line and posterior offset reconstruction, gap balancing, and sequential releasing steps. In specific advanced alignment courses three different alignment approaches can be selected and the impact of an alignment technique on HKA, gaps and resections can be compared.

3 Course evaluation

In this study all courses performed between January 2022 and April 2025 were analysed. The total number of courses and participants was assessed as well as the number of exercises performed by the participants. All performed exercises (were analysed in terms of all quality and efficiency parameters.

4 Results

In 35 courses 638 surgeons were trained at two levels: 20 courses were basic and 15 were advanced alignment courses. In a basic course, MA with principles of joint line reconstruction, and gap balancing was trained. In advanced alignment courses specific alignment philosophies were trained, predefined by the course administrator. The spectrum varies between KA, rKA up to iKA, FA and PST techniques.

5 Discussion

The most important findings of this study were the following: Firstly, decision quality (51 %) and time efficiency (62 %) were significantly increased independent of the experience level of participants. Secondly, only 36–51 % of participants have used the simulator performing their homework training exercises. Despite this relatively low percentage of participants completing all exercises this study proves that simulator training is a very effective tool for alignment and robotic TKA training. However, a considerable number of participants seems overstrained by the complexity of digital training or may simply not be ready to move towards digital surgery for other reasons.

The improvement in decision quality (51 %) and efficiency (62 %) found in the present study is in agreement with an earlier experience with this simulation tool.¹⁴ Earlier results showed an almost doubled rate of successful exercise performance from 45 % to 87 % and a 40 % increase of efficiency but this may also be attributed to differences in selection or recruitment of course participants. An additional finding in the previous study was that 92 % of all participants described to be better prepared for later robotic surgery by preceding simulator training.

Simulator training is able to move a considerable part of the intra-operative robotic learning curve for all digital decisions outside the operating room, certainly beneficial for the patients. As previously shown simulator training is able to reduce cutting mistakes, balancing errors and time loss in the OR.^16–19

A disappointing finding throughout all courses, independent of the experience level, was the fact that around 50–65 % of delegates did not complete their assigned exercises. One possible reason besides lack of interest of the participants could be the challenging complexity of digital surgery and the simulator software. Reducing the simulation process to a pure decision training on planning screens might be overburdening, although in all cases a specific webinar and video training was offered in advance to get used to the tool. In particular, in beginners, such as residents and fellows, even the basics of TKA surgery, such as bone resections, balancing goals were unclear, demonstrating a failure rate of up 91 %. This is demonstrating that a lot of rather unexperienced participants are not ready for their first real robotic-assisted TKA surgery or definitely need a lot more of guidance throughout the entire decision process. One option to increase acceptance of the simulator is integrating the simulation tool in a broader learning experience consisting of step by step videos, alignment videos and case based discussions performed by well-known TKA experts. Another promising option is integrating the simulator tool in a virtual reality (VR) environment.¹⁹ It has been proven that with the support of VR/AR/ER quality and efficiency can also be improved.^20,21 Combining both modalities might help to make it more user friendly and accepted, in particular in the younger generation of surgeons. This would allow to train all parts of TKA surgery, and not only the parts of quantitative decisions. In such an ER environment all important steps of the conventional part of TKA surgery, such as approach, soft tissue management and wound closure including haptic feedback could be added. Such a comprehensive approach could potentially reduce the need for expensive and resource intensive cadaver lab trainings.²⁰ It is also an option to overcome the limitation of alignment training on non-arthritic knees in cadaver labs. Additionally, cadaver labs are limited to one or two knees while in simulator training a far larger number and broader range of deformities can be trained. Overall simulator training, particularly in combination with VR/AR/ER represents a highly attractive training option to ensure surgeons readiness for TKA surgery. Such developments may bear great potential for the future: By successfully completing a defined training course a certificate, similar to a pilot's license to fly a plane, could be obtained. This may help minimizing mistakes in early stages of the learning curve as well as it would reduce surgical time.^18,20

However, another reason for the high rate of participants not completing the course, in particular experienced conventional TKA surgeons might be the fact that they are not really interested in the transition towards digital surgery, leaving their comfort zone. This highlights the fact that this specific simulator training is not necessarily advisable for everyone. Therefore, a specific selection of participants prior to courses is extremely helpful to minimize the number of dropouts. This can help to reduce costs but also reduce frustration on both sides. Another aspect of dropouts in the experienced TKA surgeon group might be the unavoidable transparency of the software. With simulator training each decision is quantitative and every failing decision becomes transparent for the surgeon but also to the trainer. Nevertheless, without simulator training difficult balancing situations might happen then in the OR as the surgeons may not be ready to perform robotic surgery with personalized alignment in a reasonable time without mistakes having consequences in the real word.

6 Conclusion

Overall, simulator training has proven to work for TKA training and in particular for robotic alignment training. Decision quality and time efficiency are significantly improved and surgeons feel better prepared for the transition to digital personalized TKA. In order to improve acceptance and usage of simulators an integration in a digital training platform including different levels of training videos and case discussions might be the next step. Additionally, an integration in ER modules is a promising option, being particular helpful for younger surgeons. For the experienced, conventional TKA surgeons a clear preselection of interested candidates is a prerequisite to achieve high surgeon and trainer satisfaction. Not every surgeon may be ready for digital and personalized surgery and as long as its superiority is not clearly proven, this is not mandatory. However, we all should be open to new ideas as learning is a lifelong process.

Patient consent statement

Not applicable.

Guardian/patient's consent

Not applicable.

Data availability statement

Data is available on reasonable request by contacting the corresponding author.

Permission to reproduce material from other sources

Not applicable.

Author's contributions according to CRediT taxonomy

HG: conceptualization, formal analysis, writing-original draft, writing-review and editing; GMA: writing-review and editing, visualization, formal analysis; RvER: writing-review and editing; KG: writing-review and editing; MTH: writing-review and editing, formal analysis, methodology.

Ethical statement

Not applicable.

Funding statement

No funding was provided for this work.