Commentary: Simulation of Dural Repair in Minimally Invasive Spine Surgery With Use of a Perfusion-Based Cadaveric Model

Daniel Lubelski, Debraj Mukherjee, Nicholas Theodore

DOI: 10.1093/ons/opz111 Published: 23 May 2019


In their timely article, “Simulation of Dural Repair in Minimally Invasive Spine Surgery With Use of a Perfusion-Based Cadaveric Model”,1 the authors present a novel simulation course to teach their neurosurgery residents how to repair inadvertent durotomies during minimally invasive spine surgery. In this pilot program, fresh cadavers were obtained and cervical laminectomies were performed. The 8 included resident participants (post graduate years 1-7) were asked to repair a durotomy via minimally invasive tubular retractor systems, while a perfusion system simulated pulsations of fluid that mimicked pulsatile flow of cerebrospinal fluid. After a short instruction session and the simulation, the authors’ report that the mean closure time improved from 12 to 7 min, all achieved robust closure resistant to Valsalva maneuvers, and residents surveyed reported a high degree of authenticity and greater confidence with dural closure.

As with any simulation, critics can point out the disconnect that exists between a cadaveric model and real-world intraoperative situations. In the simulation, visualization is superior, exposure is optimized, and there is no bleeding obscuring the field. In the model, a dural opening is midline and linear, rather than irregular and lateral or even ventrolateral as is often seen. The trainees do not have to encounter nerve roots floating out of the dura, the pressure associated with managing changes in intraoperative monitoring signal, or the fear of permanently injuring someone.

It is important to recognize, however, that this disconnect between the intraoperative setting and the practice setting is precisely the goal. Having a simulation where residents can hone their technical skills, better understand the anatomy and anatomical variants, learn to deal with intraoperative obstacles, and experience handling the tissue, all in a semirealistic laboratory-based setting, is critical for resident education. Developing the dexterity and muscle memory under optimal conditions facilitates superior performance intraoperatively, and enables one to improvise when the conditions may be suboptimal. This has been demonstrated in both the general surgery and neurosurgery literature, in the education of medical students, junior trainees, and senior residents.2-4

From a surgeon-mentor perspective, these simulations are critical as well. Attending surgeons are tasked with training competent and safe surgical residents, ready for independent practice. This objective, however, is challenged by increasing litigation in health care, and concerns regarding patient safety, as well as economic pressure from hospital administrators for faster, more efficient care.5 Working with residents in a laboratory eases the attending’s anxiety about giving residents autonomy and allows for more intraoperative surgical coaching and education. This also facilitates tracking resident progress and understanding of a given resident’s ability level, with this time spent together likely translating into greater autonomy in the operating room setting as well.

Societal expectations have changed the medical education landscape from the days of William Halsted and Harvey Cushing. Increasing data collection, hospital metrics, patient satisfaction scores, as well as greater transparency have likely impeded resident operative autonomy. The Centers for Medicare and Medicaid (CMS) reimbursement model require the attending surgeons be scrubbed for all “critical” portions of a case and provide supervision commensurate with the skill level of the resident.6 Additionally, while case series have shown the safety and equivalent outcomes associated with overlapping surgery.7-9 media reports have led to government and hospital policy limiting these actions.10,11 Naturally, these changes have affected residents’ operative autonomy and training. Over the past decade there has also been much discussion regarding the Accreditation Council for Graduate Medical Education (ACGME) 80-h work-week restrictions that have also reduced residents’ patient exposure, operating time, and case volume.12-14

A survey of the United States Neurosurgery Program Directors15 has suggested that simulation will not replace conventional surgical training, but that it is an imperative resource needed to supplement resident training. Within neurosurgery, numerous institutions, in virtually all of the subspecialties have piloted simulation programs to help with resident training.16-20 Resident surveys have suggested that cadaver simulation training to be the preferred method,21 though cost limitations and access to anatomic specimens at some institutions may preclude exclusive use of this modality.

In spine surgery, the importance of simulation resident training is particularly important. Often, the indications for elective spine surgery are to improve quality of life rather than the life-saving nature of trauma and cranial neurosurgery. While always sought to be avoided, in spine surgery there is even greater emphasis and significantly less tolerance for any complication or adverse outcome by both patients and providers. Spinal anatomy can be challenging for junior and senior residents, and a substantial portion of the surgery is based on tactile feedback. This often translates into attending spine surgeons being reluctant to give residents intraoperative autonomy. Harrop and colleagues22 demonstrated the effectiveness of a novel artificial simulation model for posterior cervical instrumentation. Others have reported the success of haptic computer based tools,23 minimally invasive simulators,24 and cadaver-based training.25

At our institution, we have employed a multipronged approach. Residents attend monthly “approach-of-the month” cadaver based prosections to explore the relevant anatomy. Quarterly, the residents attend a cadaver-based dissection course with representation from multiple faculties of various subspecialties. These courses, typically on a Saturday morning, include didactics, cadaver based teaching, and hands-on experience for the residents. Several times per year, the residents attend industry-sponsored courses to trial novel devices and instrumentation systems. Lastly, to improve intraoperative-based training, our institution has begun apprenticeship rotations. For a month at a time, the resident works exclusively with a single surgeon both in clinic and in the operating room. This leads to improved rapport between the specific resident and surgeon, allows the resident to learn the given attending’s preferences, and enable the surgeon to provide directed feedback on a daily basis regarding areas of strengths and weaknesses, as well as ways to improve. Particularly at larger institutions, such as ours, a resident operates with many different surgeons, which may limit the attending’s ability to provide consistent directed feedback. The apprenticeship model allows the attending to feel more comfortable giving a resident progressive responsibility and operative autonomy based on his/her particular ability.

While there have been many prior articles looking at simulation in neurosurgery, the present article is unique in its focus on simulation in spine surgery, and particularly the repair of dural leaks. As the authors explain, this complication is anathema to attending surgeons, and can be potentially devastating for patients leading to pain, infection, revision operations, and reduced quality of life. By presenting the tools to develop this novel simulation model and demonstrating its effectiveness, the authors have shown how other institutions can incorporate this model into their own resident education programs. This model, coupled with similar simulation programs, has the potential to lead to improved resident training, increased autonomy, and superior outcomes while modernizing our resident training armamentarium.