Article Text
PDF
Abstract
Introduction Surgical simulation provides a zero-risk setting in which technical skills can be obtained through repetition. The feasibility and utility of simulated diagnostic cerebral angiography among neurosurgical residents and fellows was studied using an endovascular biplane angiography simulator.
Methods Ten neurosurgical residents and four endovascular neurosurgery fellows were recruited into a standardized training protocol consisting of a didactic, demonstration and hands-on learning environment using the Simbionix simulator. Participants were instructed to catheterize the right internal carotid artery, left internal carotid artery and left vertebral artery. The task was repeated five times.
Results All participants demonstrated improvement over the five trials. Residents performed actions that were perceived as potentially dangerous (n=8) while fellows performed the procedure with superior technique. Residents performed the task with an initial total procedure and fluoroscopy time of 6.6±4.3 min and 4.9±3.7 min, respectively, and improved on the fifth trial to 3.4±1.3 min (p=0.03) and 2.3±0.78 min (p=0.004), respectively. Residents approximated the efficiency of fellows for the third and fourth trial.
Conclusions Incorporating an endovascular simulator is feasible for training purposes in a neurosurgical residency program. This study provides objective documentation of the facilitation of technical angiography skill acquisition by the use of simulation technology.
- Neurosurgery
- education
- simulation
- angiography
- hemorrhage
- fistula
- arteriovenous malformation
- intracranial pressure
- aneurysm
- economics
- thrombectomy
- technique
- complication
- catheter
- balloon
- thrombolysis
- stroke
- stent
- stenosis
- malformation
- intervention
- embolic
- device
- coil
- atherosclerosis
- angioplasty
Statistics from Altmetric.com
- Neurosurgery
- education
- simulation
- angiography
- hemorrhage
- fistula
- arteriovenous malformation
- intracranial pressure
- aneurysm
- economics
- thrombectomy
- technique
- complication
- catheter
- balloon
- thrombolysis
- stroke
- stent
- stenosis
- malformation
- intervention
- embolic
- device
- coil
- atherosclerosis
- angioplasty
Introduction
The use of technology to simulate real-world scenarios for training purposes has a rich history in the airline, aerospace and military industries. The value of simulated surgery for the re-creation of invasive procedures has also been recently acknowledged. Surgical simulation provides a zero-risk setting in which skills can be acquired through repetition. Even the most crude simulation strategies that ‘mimic’ multistep procedures can aid the trainee in planning for a real-life procedure by allowing and facilitating the rehearsal of the steps involved. Anatomically advanced and technically accurate simulation systems that provide sensory (haptic) feedback can further aid the trainee in developing and refining the mechanical skills required for the procedure. Ultimately, the best metric for determining simulator effectiveness is an objective assessment of the efficacy of skill acquisition derived from simulation usage and the transfer of this information to actual real-life procedures.
Simulation is now widely accepted for enhancing the resident training experience in a wide range of scenarios, including surgical training of laparoscopic procedures,1–4 endoscopy,5 colonoscopy,6 ,7 thorascopy,8 cataract surgery,9 peripheral vascular endovascular interventions10 and airway management.11–13 These simulation systems replicate procedures that involve a two-dimensional visual display. The application of simulation in the realm of neurosurgical training and education has lagged somewhat compared with other specialties, probably due to the challenge inherent in replicating a complex three-dimensional reality. However, improvements in computer processing power, volume rending, graphics and haptics have facilitated the creation of sophisticated—albeit limited—simulation systems for neurosurgical applications.14
Angiography and the growing field of neurointervention are playing a rapidly evolving role in the care of neurosurgical patients. Neurointervention requires a unique skill set that is not transferable from open domains taught in neurosurgical training programs. This study is intended to provide some preliminary data on the feasibility and utility of simulated diagnostic cerebral angiography among neurosurgical residents and fellows using an endovascular biplane angiography simulator.
Methods
Neurosurgical residents in their postgraduate years (PGY) 1–5 and first and second year endovascular neurosurgery fellows were recruited into a standardized training protocol consisting of a didactic, demonstration and hands-on learning environment.
Endovascular simulator
The Simbionix simulator (Simbionix USA Corp, Cleveland, Ohio, USA) was employed. Simbionix provides an interactive biplanar fluoroscopic display to perform both diagnostic and interventional procedures on a number of ‘patient’ case scenarios with unique vasculatures (figure 1). A broad selection of groin sheaths, diagnostic catheters and guidewires may be selected and the software incorporates the unique mechanical properties of each. While the behavior of the catheter in the vessel is simulated, actual catheter and wire manipulations are incorporated by motion tracking sensing capabilities in the hardware component of the simulator. The tracking system measures horizontal translation and roll of each tool at the fixed insertion point and these values are translated to tool length and roll in the simulated space. Each catheter and wire combination is defined by its set of biomechanical properties, such as shear and Young's modulus. Shear defines the resistance to torque forces (angulation and torquability) and Young's modulus defines the resistance of the tool to bending and stretching force (stiffness measure).
(A) Simbionix provides an interactive and realistic biplanar fluoroscopy to perform both diagnostic and interventional procedures on a number of ‘patient’ case scenarios with unique vasculatures. A broad selection of groin sheaths, diagnostic catheters and guidewires may be selected and the software incorporates the unique mechanical properties of each. (B) Patient vital signs are continuously monitored during the session.
To simulate the behavior of the catheters as they are navigated in the vasculature, the software then calculates the collision of the catheter and wires with the vessels. Each vessel applies a force vector to a catheter segment in order to maintain the position of the catheter within the intraluminal compartment. The magnitude of this force vector is an adjustable parameter that may have different values along the vessel. Combining the catheter properties with the vessel properties, the software determines the physical state that balances the outer force vectors (applied from the vessel on the tool) and the inner force vectors (the energy required to maintain the tool in its conformational shape rather than return to its free space configuration). In this manner, information such as angulation, friction and forward loading are all incorporated into the algorithm to produce haptic feedback and allow the catheters to navigate in a realistic fashion. Contrast can be administered during both simulated real-time fluoroscopy and digital subtraction angiography. Roadmap assistance for navigation can also be simulated.
Pre-task survey
Participants were asked the number of cerebral angiograms they had observed and in which they had directly participated. They were asked to rate their knowledge of the anatomy of the aortic arch, cervical carotid and vertebral arteries (scale 1–5) and their level of comfort on angiographic catheter selection technique (scale 1–5).
Didactic
Residents with limited prior experience with cerebral angiography were given a 4 min eight-slide instructional presentation. All received the standardized didactic by a single instructor, a neurosurgical resident (PGY7) with 1 year of diagnostic and neurointerventional fellowship level training (AMS) who had performed >500 cerebral angiograms. The presentation covered basic anatomy of the aorta and its daughter branches, with standard textbook illustrations and angiogram images. It also introduced the properties and geometry of the diagnostic catheter and guidewire to be used, in addition to basic technique of crossing the arch and selective catheterization. Lastly, the task of performing a four-vessel angiogram was described from a standard percutaneous common femoral artery approach. Participants were instructed to catheterize (in order) the right internal carotid artery, left internal carotid artery and left vertebral artery (LVA). The fellows with angiographic experience were not given the instructional presentation.
Demonstration and task
Each participant observed a single live demonstration (5 min) of the task during which they were free to observe the actual catheter and wire manipulations in addition to the simulated fluoroscopic images. A simulated 5F vertebral diagnostic catheter and a 0.035 inch soft standard-angled guidewire were employed. The starting point for the task was with the diagnostic catheter and wire intraluminally in the descending arch. Contrast administration and biplanar control was solely the responsibility of the instructor. The simulated ‘patient’ had a type I aortic arch with normally distributed daughter vessels (ie, not bovine) that arose from the dome of the arch (figure 2). The task was repeated five times by the participants and involved several independent steps to be performed in a standardized order (see appendix 1).
The training task involved performing a “four vessel” cerebral angiogram on a simulated patient with a type I aortic arch with normally distributed daughter vessels as demonstrated in this left anterior oblique view in schematic form (A) and with a three-dimensional cartoon overlaid on the simulated fluoroscopy view (B). This anatomical detail was not available to trainees who instead relied on knowledge of the relevant anatomy and behavior of the guidewire and catheter to navigate from the arch to the cervical vasculature (C, D). Trainees advanced the diagnostic catheter over the guidewire to cross the arch and then performed a selective catheterization of the right internal carotid artery (RICA). A selective RICA angiogram was then performed as shown in the non-digitally-subtracted view (E). Trainees were instructed to perform a roadmap control for selective catheterization of the left vertebral artery (F).
Total procedure and fluoroscopy times were recorded in minutes and seconds. Volume of contrast administered and complications were also recorded. Complications included arterial dissection, inadvertent right vertebral artery catheterization and advancing the catheter without a leading guidewire. Participants were allowed an unlimited amount of time to complete the task and were given feedback and suggestions by the instructor. The participants were not informed on what basis they would be scored.
Post-task survey
Following the completion of the task, residents were surveyed on the relative merits of the didactic versus the hands-on component of the teaching session. Fellows were asked to rate the accuracy of the simulator for both visual and mechanical/tactile properties (scale 1–10).
Statistical analysis
Numerical results are described as mean and standard deviations. Two-tailed p values were to compare performances of different groups of individuals. All analyses were performed with commercially available statistical software. A significance level of 0.05 was used to determine significance.
Results
Ten residents in postgraduate years 1–5 (M:F 9:1, mean age 28.2 years) and four fellows (M:F 3:1, mean age 33.3 years), two of which were in their first year (F1) and two in their second year (F2), were included in the study. All 14 participants successfully completed the training session in 1 h.
Pre-task survey
The residents reported a paucity of exposure to cerebral angiography prior to the training session. Among the resident participants, five (50%) had never observed a cerebral angiogram being performed, two had observed one, two had observed <5 and one had observed >10. Only one resident had performed (in part) a single cerebral angiogram. All the fellows had performed >100 angiograms. Fellows rated their knowledge of the anatomy of the aortic arch and cervical vessels and their comfort level with selective catheterization technique higher than residents.
Simulated cerebral angiography
All participants, regardless of prior angiography exposure, demonstrated improvement over the five trials as reflected in improved total procedure and fluoroscopy times (figure 3). There were no dissections, however residents performed actions that were perceived as potentially dangerous, such as erratic catheter movements into the LVA (n=4) and leading the catheter with a short segment of guidewire (n=4). All improved with direction on subsequent trials. One resident inadvertently catheterized the right vertebral artery and one resident did not inject contrast as the catheter was being withdrawn from the LVA to verify maintained anterograde flow. There was no difference among residents and fellows with regard to the distance the catheter was advanced without a leading guidewire (33.2±22.5 mm vs 27.9±22 mm; p=0.37). Fellows performed the procedure with superior technique (ie, no complications or potentially dangerous maneuvers).
Performance profiles for (A) residents designated by their postgraduate year (PGY) and (B) first (F1) and second (F2) year fellows. The bar graphs display the total procedure time (black) overlaid on the fluoroscopy times (gray) for each of the five test trials (x-axis). Time is expressed on the y-axis in minutes.
Fellows performed all the trials with a lower volume of contrast than residents (49±5.1 ml vs 55.7±6.1 ml; p<0.01). Since the volume of contrast injected for the selective angiograms was kept uniform, this reflected a more efficient navigation in the left subclavian artery in preparation of a roadmap image to visualize the origin of the vertebral artery for selective LVA catheterization. Residents performed cerebral angiography with an initial total procedure and fluoroscopy time of 6.6±4.3 min and 4.9±3.7 min, respectively. Residents demonstrated ongoing improvement over subsequent trials and performed the procedure with shorter procedure and fluoroscopy times on the fifth trial (3.4±1.3 min; p=0.03 and 2.3±0.78 min; p=0.004). Fellows performed the task on the initial trial with procedure (3.6±0.5 min) and fluoroscopy (2.7±0.7 min) times which were similar to the resident's fifth (best) trial times. However, fellows also improved their procedure and fluoroscopy times on the fifth trial (2.2±0.4 min; p=0.04 and 1.5±0.4 min; p=0.02, respectively).
Comparing the performance and learning curve of residents and fellows over the five trials showed that, although fellows were more efficient on the first two trials, the residents were able to close the gap and perform competitively on the third and fourth trials (figure 4). However, the performance of the residents appeared to plateau after the fourth trial while fellows were able to improve upon their preceding times on the final run. On the last trial, fellows completed the cerebral angiogram with fluoroscopy times 1 SD below the residents, although with the small sample size this difference did not reach statistical significance (p=0.07).
Graph comparing fluoroscopy times (min) for residents (n=10) and fellows (n=4) for each of the five task trials. Data are expressed as mean±SD.
Post-task survey
All residents rated the hands-on component of the training session as having a higher educational value than the didactic. However, all considered the didactic session to be of value and nine (90%) would not have replaced the time spent on the didactic portion with additional hands-on training.
Fellows with experience in real-life cerebral angiography rated the simulator highly. On a scale of 1–10 (highest), fellows rated the simulated task as 7 (range 6–8) on visual presentation and a 7.25 (range 7–8) on mechanical properties such as wire and catheter behavior and haptic feedback.
Discussion
Simulation of surgical procedures such as laparoscopy has been fully incorporated into the training of general surgery residents. In fact, general surgery training programs are required under the Accreditation Council for Graduate Medical Education to provide a simulation laboratory for trainees. The role of simulation in neurosurgical education is yet to be defined. There are two main components to a procedure that simulation can reinforce. First, the order of steps required to successfully perform any given procedure can be rehearsed under different scenarios. For example, one can perform A then B then C and await the outcome of this management algorithm. After the outcome is identified by the simulation system (adverse or desired), another treatment algorithm can be employed in a reiterative manner.15 ,16 This learning environment could be very beneficial and efficient, particularly for relatively rare yet vital occurrences in real-world patient encounters such as critical resuscitation.17 ,18 Not surprisingly, there is evidence that simulation of cardiopulmonary resuscitative efforts, for example, can improve performance and subjective perceptions of self-competence among trainees.17 ,19 An intensive curriculum-based simulation module can be as beneficial as 6 months of clinical ward experience.20
Second, simulation of technical procedures can also be of benefit by providing a no-risk environment for rehearsal of the mechanical skills required. Simulation has been shown to be effective for relatively simple procedures such as direct or peripherally inserted central venous access catheter placement and transnasal endoscopy.20–23 These findings reflect a transfer of skills acquired from the simulator to an actual patient encounter. It would be expected that the closer a simulation can mimic an actual procedure, the more readily the skills can be transferred. Strategies employing simulation into training must also incorporate periodic retraining of skills to avoid deterioration over time.24
We investigated the feasibility of employing a sophisticated endovascular simulator that was rated favorably for both visual display and mechanical/haptic feedback properties by experienced fellows for the training of residents in diagnostic cerebral angiography. The potential benefit of learning this invasive procedure—that requires fluoroscopy and cumulative radiation doses—on a simulator is that it involves a no-risk environment for both the patient and the trainee. Participants from all neurosurgery resident postgraduate levels were recruited into this pilot study. The residents did not have experience with selective catheterization technique and performed relatively poorly on the initial two trials. However, they demonstrated ongoing improvement over subsequent trials and closely approximated the performance of fellows by the third and fourth trial. In contrast, fellows with considerable experience in angiography performed well on the first trial and then continued to improve on successive trials. At this point it was observed that the limitation of any further progress was the speed with which the simulated x-ray sources could be manipulated rather than further speed or accuracy of catheter and wire skills. Fellows also performed the task with superior technique. This reflects a transfer of mechanical skills acquired in their actual angiography experience to the simulator. More importantly, it strongly suggests that skills acquired on the simulator should transfer readily to the actual angiography suite. Further study will be directed at testing the hypothesis that performance on the simulator correlates with performance in the angiography suite.25 Residents reported that both the hands-on and didactic components of the training session were valuable to the learning process.
In a similarly designed study, Fargen et al 26 recently reported their successful implementation of a simulation-based curriculum among seven neurosurgery residents. The study differed from the present study in several important ways. The two studies employed different simulation platforms. Although both aim to provide the same simulation experience, technical differences between the two systems would be expected. This study supports the use of the Simbionix simulator. Both studies involved a relatively small sample size (seven in the study by Fargen et al and 14 in the present study). In addition, our study includes the performance of fellows with substantial angiography training in addition to inexperienced residents, which serves to validate the simulation experience as well as provide a benchmark of performance. In this study participants were exposed to a single type I aortic arch to isolate catheter vessel selection skill acquisition with plans to incorporate more difficult arch configurations in future training sessions, while Fargen et al trained their resident participants in type I, II and III arches in the first training session. Lastly, Fargen et al provided a training experience that spanned 2 days with a faculty-trainee ratio of 4:7 while this study provided a more focused one-on-one experience lasting no more than 1 h with demonstrable results. Evidence from the two studies taken together suggests that there are varying methods of teaching angiography technique to residents that can be successful.
There are limitations to the endovascular simulation environment employed in this study. Catheter lumen maintenance including flushing and infusing heparinized saline to avoid hemostasis, thrombi formation and air introduction is not currently simulated. These represent a distinct set of skills that must be learned concomitantly to avoid thromboembolic complications during cerebral angiography. Percutaneous access to the common femoral artery is also not a component of the simulation at this time. Another limitation is that the task involved straightforward aortic arch anatomy. Tortuous, bovine or otherwise challenging arches were not employed. This was done to create the optimal environment for learning fundamental selective catheterization technique. Entirely different learning curves may be expected for aortic arch and cervical vasculature variants of varying degrees of difficulty. Despite these limitations, we have demonstrated significant improvements in procedural and fluoroscopic times among neurosurgical residents reflecting improved catheter technique for four-vessel angiography on a simulator over a short period of time. Residents who lacked formal training and experience in angiography were able to complete the task in a time similar to that of the experienced fellows after 3–4 trials. The next step will be to fully incorporate the endovascular simulator into the resident curriculum and follow aptitude and level of comfort as residents perform procedures in the angiography suite.
Conclusions
This study shows that incorporating an endovascular simulator is feasible for training purposes in a neurosurgical residency program and provides additional evidence that such training can facilitate the acquisition of skills obtained in real-world angiography.
Key message
There may be a role for endovascular simulation in the training of basic catheter technique and vessel selection among resident trainees.
Acknowledgments
The authors thank Christine Moore for editorial assistance and help in preparation of this manuscript for submission.
Appendix
-
1. Right internal carotid artery (RICA) angiogram
-
Cross the arch: guidewire over the arch to the proximal ascending arch, then catheter over the wire.
-
Pull the guidewire back into the catheter.
-
Pull the catheter back and torque until it engages the origin of the brachiocephalic artery.
-
Advance the guidewire from the brachiocephalic artery into the right common carotid artery and then past the bifurcation into the RICA petrous segment.
-
Advance the catheter over the wire and place the angled tip at the bend of the cervical-petrous segment junction.
-
Contrast injection (10 ml).
-
Pull the catheter back into the orifice of the brachiocephalic artery.
-
-
2. Left internal carotid artery (LICA) angiogram
-
A. Pull the catheter back into the aortic arch and then pull back and torque until it engages the origin of the left common carotid artery (LCCA).
-
B. Advance the guidewire into the LCCA and then past the bifurcation into the LICA petrous segment.
-
C. Advance the catheter over the wire and place the angled tip at the bend of the cervical-petrous segment junction.
-
D. Contrast injection (10 ml).
-
E. Pull the catheter back into the orifice of the LCCA.
-
-
3. Left vertebral artery (LVA) angiogram
-
A. Pull the catheter back into the aortic arch and then pull back and torque until it engages the origin of the left subclavian artery (LSCA).
-
B. Advance the guidewire into the LSCA past the origin of the LVA.
-
C. Advance the catheter over the wire into the LSCA.
-
D. Pull the catheter back while streaming contrast until the origin of the LVA is visualized.
-
E. Perform a roadmap angiogram.
-
F. Under roadmap control, advance the guidewire into the proximal V2 segment of the LVA.
-
G. Advance the catheter over the wire and place the angled tip at proximal V1s segment of the LVA.
-
H. Contrast injection (10 ml).
-
I. Stream contrast as the catheter is withdrawn from the LVA to the LSCA and then pull the catheter back into the distal descending aortic arch.
-
References
Footnotes
-
Disclaimer The authors have no financial interest in any of the materials discussed in this manuscript.
-
Competing interests None.
-
Ethics approval Ethics approval was obtained from Cleveland Clinic IRB.
-
Provenance and peer review Not commissioned; externally peer reviewed.