Feasibility of High-Fidelity Simulator Models for Minimally Invasive Spine Surgery in a Resource-Limited Setting: Experience From East Africa

Abstract

Background:

Spine surgery is a rapidly evolving specialty with a continuous need to learn new skills. In resource-limited settings such as Africa, the need for training is greater. The use of simulation-based training is important in different stages of skill acquisition, especially for high-stake procedures such as spine surgery. Among the available methods of simulation, the use of synthetic models has gained popularity among trainers.

Method:

Twenty participants of a neurosurgery training course, most of whom (65%) were neurosurgery residents and fellows, were recruited. They had hands-on training sessions using a high-fidelity lumbar degenerative spine simulation model and hands-on theater experience. After this, they completed a survey to compare their experience and assess the effectiveness of the lumbar spine model in stimulating real patient and surgery experiences.

Results:

The participants were from four African countries, and the majority were neurosurgery residents. There were varying levels of experience among the participants in minimally invasive spine surgery, with the majority either having no experience or having only observed the procedure. All the participants said that the high-fidelity lumbar spine model effectively simulated real minimally invasive spine setup and real bone haptics and was effective in learning new techniques. Most of the participants agreed that the model effectively simulated real dura and nerve roots (95%), real muscle (90%), real bleeding from bones and muscles (95%), and real cerbrospinal fluid in the subarachnoid space. Among them, 95% agreed that the model is effective in lumbar minimally invasive spine training in resource-limited settings.

Conclusion:

With the development of new and better surgical techniques, the use of high-fidelity models provides a good opportunity for learning and training, especially in resource-poor settings where there is a paucity of training facilities and personnel.

Minimally invasive spine surgery (MISS) offers many advantages over standard surgical procedures, such as less blood loss, fewer surgical site infections, shorter hospital stays, and less demand for postoperative analgesia. In addition, a better outcome has been shown for obese patients. These advantages have led to the widespread adoption of MISS, which is now routinely used to treat many spinal pathologies. Especially in a setting with limited resources, it would be valuable if patients could benefit from these advantages. In low- and middle-income countries (LMICs), such as many countries in Africa, the need for efficient surgical procedures is significant because there are often not enough trained spine specialists to meet the basic surgical needs of the cohort.¹ However, one disadvantage of minimally invasive spine (MIS) techniques is that they have a learning curve that requires many procedures before the surgeon has sufficient experience to perform the procedure safely by himself without supervision. Therefore, there has been a continuous effort to develop safer and more efficient surgical training methods.

In recent years, simulation-based training has become popular across several medical specialties because of its several benefits.^2-5 These include the wider latitude to make mistakes during learning, especially in high-stake scenarios such as spine surgery with few real-life consequences.⁶ Simulation-based training has also become more important because restrictions on the work hours of trainees result in a decrease in theater time.⁶ In spine surgery, the use of surgical models is even more beneficial because of the delicacy of actual neural tissues and the potentially devastating effect of errors. These benefits of simulation-based training are likely to be highest in resource-limited settings where there is a wide disparity in the availability of training facilities but a greater need for training.⁷

The use of animal models such as porcine has been popular for decades because of the many similarities to human anatomy. However, for spine surgeries, especially MISS, some challenges exist in simulating intradisc procedures because of the narrow disk spaces in porcine.⁸ There is also the issue of cultural and religious acceptability of the use of porcine models, especially in Africa. Although visual reality is a favorably rated simulation model, haptic realism is a notable limitation. Human cadavers are the most realistic models in terms of anatomy and haptic reality; however, they are accompanied by the challenges of preservation, high cost, low availability, and strict ethical protocols for acquisition, transport, and preparation.

In addition, human cadaver models are often unable to effectively simulate the specific pathology and the reality of flowing blood and cerbrospinal fluid. There is also the belief by experts and trainers that safe surgical training and an effective learning curve should be a graduated process in which the use of models is an important stage.⁹ Synthetic models rank high in median tactile realism compared with other models.¹⁰ However, getting an improved synthetic model that is closely able to simulate the important realities of the fluid, colorless cerbrospinal fluid in the subarachnoid space, and flowing blood from spine bones for MISS training is challenging.

The use of a high-fidelity model has been recently reported to improve the working knowledge and technical skills of trainees in minimally invasive lumbar decompression.¹¹ However, it is unclear how these models can be implemented for surgical training in a resource-limited setting. To evaluate the feasibility of the use of high-fidelity simulator models for MISS spine surgical training, we conducted a feasibility study in simulating real-life patients for MISS training in a resource-limited setting.

Methodology

The study was performed during a neurosurgery training course in Tanzania co-organized by Weill Cornell Medicine, New York, and Muhimbili Orthopaedic Institute, Dar es Salaam. Twenty of the course attendees who were neurosurgery and orthopaedic residents and fellows from different training institutions in East and West Africa were recruited for the study. They had different levels of experience but are from hospitals that do not routinely offer minimally invasive spine procedures. Participation in the study was voluntary, and informed consent was obtained.

The participants had to attend lectures about MISS techniques during the course and participated in live patient theater MISS sessions including explanations about the workflow by experts as part of the training course. Afterward, the participants attended a training session on minimally invasive lumbar decompression using high-fidelity spine models where they performed a MISS decompression under the supervision of a spine surgeon with expertise in MISS.

The spine model used was a synthetic degenerative lumbar stenosis model (Real Spine, Realists Training Technologies GmbH, Leipzig, Germany), which was developed from the CT and MRI scan of an actual patient with L4/L5 stenosis. We acknowledge a conflict of interest in that the senior author is an investor in this company. The anatomical structures contained in the model imitate the structures in a real patient and included skin, paraspinal muscles, vertebra, ligamentum flavum, epidural fat, nerve roots, and dura¹¹ (Figure 1). The models also offer the option to simulate bleeding from the bone when cut or drilled, respectively, which could be increased or decreased by a control panel. This allowed us to adjust the difficulty to the training status of the participant. The dura contained clear fluid to simulate a cerebrospinal fluid leak in case of an injury and also allowed the training of dural repair procedures. The models are equipped with replaceable cartridges (Figure 2A) that allow for the replacement of only the model parts that the participant worked on during the operation, whereas the base plate and cover always remain the same. For the installation of the models, the company Realists provided a virtual tutorial that allowed the setting up of the model through one tutor surgeon.

Figure 1.

Tubular MIS training using the high-fidelity lumbar stenosis model. A, The lumbar stenosis simulator. B, The complete setup, combined with the Viseon MaxView visualization system during the initial lecture. C and D, Practical training sessions of the participants by the expert. MIS = minimally invasive spine

Figure 2.

A, Inner system of the Real Spine model showing the irrigation system that maintains clear fluid simulating the cerbrospinal fluid and red fluid that simulates blood. B, Viseon MaxView portable camera system and the image output.

Because operating microscopes are often not available in LMICs, we decided to use a portable camera system (Viseon MaxView) that attaches directly to the tubular retractor and is compatible with most commercially available monitors (HDMI input) (Figure 2B). Because of its straightforward applicability, we believed that the system is well suited for procedures in which the magnification of an operating microscope is helpful, but a microscope is not obtainable. In addition, it allows the procedure to be presented to a larger audience on a large monitor. In our study, a standard commercially available monitor with a resolution of 1920 × 1080 was used. The power supply of the camera and the models was provided by the standard power adapters; only the electrical power sockets were adapted to the local power connections through adapters.

Both the microscope and the Viseon MaxView visualization systems were used for the live theater session training, whereas only the Viseon MaxView was used for the MISS decompression training using the spine model.

Spine Model Training Setup

The spine models were placed on the table simulating the patient in a prone position. The tubular retractor was mounted on the table contralateral to the surgeon's position. Although in real-life surgeries the tubular retractor is placed under radiograph control or navigation guidance, these options were not available because of the limited radiation transparency of the model. Therefore, the retractor was placed under visual control by the mentor surgeon at the correct level. The procedures on the models were performed using standard lumbar MIS instruments. The participants were able to use a high-speed drill for bone removal. The setup is shown in Figures 1B and 2B. After a short repetition lecture and demonstration of a tubular MISS lumbar decompression workflow, the participants were asked to perform the procedure under the guidance of the tutors.

Procedure Workflow

In all cases, the procedure trained was a tubular MIS decompression. The workflow followed was based on the description of Ten-Step Minimally Invasive Spine Lumbar Decompression by Boukebir et al.¹² Because of the limited radiographic capabilities of the models when placing the tubular retractor, the retractor was placed over the inferior medial edge of the lamina by the tutor surgeon under visual control. The training procedure was based on the following steps:

• Soft-tissue dissection and identification of the inferior edge of the lamina and spinous process for the orientation;

• Partial removal of laminar bone using a high-speed drill and Kerrison rongeur until the medial ligamentum flavum can be accessed;

• Exposure of the cranial insertion of the ligamentum flavum;

• Removal of the ipsilateral ligamentum flavum starting from the medial portion;

• Medial angulation of the tube and tilting the table away from the surgeon to gain access to the contralateral ligamentum flavum;

• Contralateral drilling of the laminar bone using a high-speed drill (undercutting);

• Removal of the contralateral ligamentum flavum using Kerrison rongeurs;

• Identification of the contralateral exiting nerve root and decompression of the nerve root;

• Tilting the table back and completing ipsilateral decompression.

After the training sessions, a survey of the participants was performed using a 24-item self-administered e-survey. Participants were asked to compare their experience of the use of the high-fidelity spine model with the real patient in live surgeries to assess the ability of the spine model to simulate real surgery. The questionnaire gathered information about sociodemographic characteristics including age, sex, country of citizenship, country of training, age, specialty, position, handedness, and level of experience in spine surgery. The outcome measures were the ability of the high-fidelity spine model to simulate real patient and surgery characteristics (procedure setup, instrumentation used, workspace, and tissue haptics), effectiveness in training, the feasibility of use in a resource-limited setting, and realism, as assessed by the participants.

Results

It was possible to implement the described training setup without notable problems. The lounge of the operating tract of the Muhimbili Orthopaedic Institute served as the training location because it offered sufficient space for the equipment and for the audience. The experimental setup required a total of four power outlets, one for the model, the monitor, the camera, and the high-speed drill. The camera connected easily to the monitor through an HDMI port and consistently provided high-resolution images. The standard monitor used in our setup was adequate for training in all cases. There was no failure of the camera or the model during the study. In every case, a tubular unilateral laminotomy for bilateral decompression (ULBD) decompression was performed on the model.

The 20 participants were from four different African countries, 12 from Tanzania, 5 from Kenya, 2 from the Democratic Republic of the Congo, and 1 from Nigeria. They are all in training at an African hospital or completed their training in Africa. Of the participants, 13 are current residents, 4 are consultants, and 3 are physicians aiming to apply for neurosurgical residency. Seventeen were in neurosurgery specialty, whereas 3 were in general surgery. The average age of the participants was 35 years, and they were predominantly males with three females. All the participants have a right-dominant hand. The level of experience with open spine surgeries and minimally invasive spine surgeries varied among the participants. Only 10% (n = 2) of the participants were proficient enough with spine surgeries to perform procedures on their own. Forty-five percent (n = 9) said that they can perform open spine surgeries under supervision, whereas 20% (n = 4) of the participants had only observed spine surgeries without assisting, 20% had assisted (n = 4), and only one person had no prior experience with spine surgeries. However, with regard to MISS, 35% (n = 7) did not have any experience, whereas 25% (n = 5) said that they can perform MISS procedures under supervision. Ten percent (n = 2) have assisted, whereas 30% (n = 6) said that they had only observed (Figure 3).

Figure 3.

Level of previous experience with spine surgeries.

In the participants' assessment of the ability of the high-fidelity spine model to simulate real patients and surgery, all (100%, n = 20) participants said that the spine model effectively simulated the real surgery setup and allowed the good use of surgical instruments. All (100%, n = 20) participants said that the spine model effectively simulated real bone haptics and bone dust, and the majority (90%, n = 18) said that it effectively simulated real muscle haptics. Ninety-five percent (n = 19) said that the lumbar spine model effectively simulated dura and nerve roots, whereas the same percentage of participants agreed that the lumbar spine model effectively simulated bleeding from muscles and bones. Ninety percent (n = 20) of the participants were convinced that the lumbar spine model effectively simulated cerbrospinal fluid presence in the subarachnoid space.

All (100%, n = 20) participants believed that the high-fidelity lumbar spine model was an effective way of learning new techniques. Ninety-five percent (n = 19) agreed that surgical simulators may be effective in training MISS in resource-limited settings. The majority (90%, n = 18) of participants said that the lumbar spine model compares well with real patients, and all (100%, n = 20) said that sessions with the spine models will be useful in their future practice. All (100%, n = 20) participants believed that the use of the model in training would be feasible in their training institutions, and all participants (100%, n = 20) are interested in future training sessions involving the high-fidelity lumbar spine model for degenerative disease and other spine pathologies.

Discussion

The benefit of lumbar MISS over open surgery has been well demonstrated, and MISS techniques have become well established.^13,14 However, the global transfer of skills especially to resource-limited settings has been slow, and most developing countries, especially in Africa, still mostly depend on open spine surgical procedures for procedures that qualify and are better done with MIS. One of the challenges is due to the limitation of training facilities in such settings: the need is greater than the available facilities.¹⁵

Simulation-based surgical training has been useful in the acquisition of skills; however, there have been questions on the transferability of the skills acquired to the real patient and surgery.⁷ Synthetic spine models have been used for years but often only for learning pedicle screw placement.¹⁶ However, with the recent development of high-fidelity lumbar spine models, the applicability in performing almost an entire procedure such as tubular ULBD has been demonstrated. A recent study by Melcher et al¹¹ revealed that the use of the high-fidelity simulator for tubular ULBD increased the working knowledge and technical skills among orthopaedic and neurosurgical trainees.

Another challenge is the ability of the simulation methods to effectively simulate real patient pathology and surgeries. Based on the experience of the participants in this study, the high-fidelity degenerative lumbar spine model was able to effectively simulate the real patient's tissues such as muscles, bones, ligaments, epidural fat, dura, and nerve roots, with the bone simulation having the highest score among the participants. Tissues such as bones and muscles bleed during spine surgeries, and almost all the participants agreed that the spine model effectively simulated bleeding compared with the real patient. This is similar to the results observed by Stefan et al.¹⁷

Although the use of cadavers provides the most accurate anatomy, there is the limitation to providing accurate pathology, high cost, cumbersome logistics, and limited supply.

Our training setup was specifically designed to provide a way to deliver surgical training and improve patient care in LMICs. Our setup was straightforward and could be set up in under an hour of preparation time in a room of a suitable size. No additional equipment is needed to set up a training center with models and a MaxView camera, other than a TV, which is available in practically every surgical clinic, and four power outlets on the wall. In hospitals where MISS instruments are not available, a limited selection of instruments needs to be provided. However, our setup can be implemented even at a facility that does not have an expensive operating microscope. Furthermore, it was possible to assemble and operate the high-fidelity simulators we used without additional personnel after a virtual briefing from the company. For cadavers, additional personnel is needed in most cases to take care of the handling.

All the participants of this study were from resource-limited countries, and the majority agreed that the use of the high-fidelity lumbar spine model would be an effective tool for training MISS in resource-limited settings. The participants in our study were all physicians with varying spine surgical experience. With 90%, most participants did not yet feel confident enough to perform spine surgery independently. We believe that the target audience for training on spine surgery models are surgeons during their residency who have basic knowledge of the spinal anatomy, underlying anatomy, and surgical procedure. Physicians who do not have this basic knowledge so far are likely to acquire only basic skills during the model training, such as basic handling of a high-speed drill or spinal anatomy for which there are less expensive ways to train. On the other hand, surgeons who already have experience in tubular MISS are also likely to have less benefit from the models, as the haptics and surgical impression do not yet match real patients, despite the sophisticated technology. However, cadavers also share these two disadvantages in surgical training. Compared with cadavers, the models described here are less complicated to handle because they can be stored without refrigeration and chemicals. In addition, they are less complicated to transport because they can be disassembled and sent by mail to virtually any location. Because of the replaceable cartridges, the required space is even smaller compared with cadavers. In addition, the models can be used to display and train specific individual pathologies, an option that is not available with cadavers.

Another advantage of the models that are used is that the model can be disassembled after the procedure, and the trainee can assess the surgical result without the soft-tissue covering. This makes it easier to understand anatomical correlations and size concepts than with a cadaver, where this is only possible with a great effort and is practically never done.

However, our study also has some limitations. There may be an issue of cost, which is notable in low-resource settings as the high-fidelity model costs up to $1,000 and more. However, when compared with the total cost of available alternatives such as the acquisition, preparation, and storage of human cadavers, the models may probably be more cost effective.

Another limitation of our study setup is that the camera, monitor, high-speed drill, and model all require power to function properly. Although power was always available in a stable manner during our training course, this is not always guaranteed in LMICs. Therefore, based on our experience, we recommend conducting training courses at a center where the power supply is ensured by emergency generators.

Although the assessment by the participants was performed after hands-on experiences of both the high-fidelity lumbar spine model–assisted MIS training and real surgeries, they are subjective and may be prone to bias. Also, the sample size of 20 creates a limitation on the generalizability of the study.

Overall, however, our feasibility study showed that the training setup we developed allows for training tubular ULBD techniques in a resource-limited setting. We hope that this will improve surgical training in underresourced areas, thereby improving patient care in LMICs. To evaluate the actual training effect, our training model needs to be evaluated at additional hospitals and courses.

Conclusion

Our feasibility study showed that a training setup for tubular MISS can be established with comparatively little effort even under resource-limited settings. In a center with a stable power supply, a surgical skill training laboratory can easily be set up in any suitable room using high-fidelity simulators and a tube-mounted camera. In our opinion, the ideal target audience for the evaluated high-fidelity simulators are surgeons in training with already existing basic knowledge in anatomy and surgery because they benefit most from the simulators.

The use of the high-fidelity simulator model for minimally invasive degenerative lumbar decompression training is an effective method and a viable option to improve and increase training in resource-limited settings. It should be considered as part of the regular aspect of the training of neurosurgical and orthopaedic resident's training.