Training outside of the operating room offers a structured and systematic educational opportunity, in addition to stress modulation. Because the operating room is a stressful place with many distractions, time constraints, concerns for the patient, equipment failures, and interpersonal issues [7, 8], stress modulation was found to enhance performance when fine motor control and complex cognition are required [9].

The effectiveness of simulation has been demonstrated primarily in lower level learners [10]. According to Fitts and Posner, the first two stages of motor skills acquisition can be sufficiently accomplished in simulation labs. These include the cognitive stage (intellectualization of the task) and the integrative (associative) stage (translation into appropriate motor behaviour), which are achieved by practice and feedback. Accomplishing both stages allows one to proceed to the third stage, the autonomous stage, which is mastered in the operating theatre and results in a smooth performance without cognitive awareness [11, 12].

Today, multiple models of simulation are available, though there is no consensus on which dexterity drill should be incorporated into simulation models for the acquisition of appropriate motor skills [2]. As pointed out by Aggarwal et al that it is not a matter of which simulator to use to acquire the skill, but rather the design of the laboratory-based skills training curriculum [13]. Trainees can evaluate their own performance by comparing their results to the standards associated with a particular simulator, and then working to minimize the difference with subsequent exercises (internal feedback). The use of expert evaluators is also important as they can provide external feedback to the trainees, including information about effectiveness and quality of the operating end product (Fig. 6.2). Procedure effectiveness can be evaluated using an objective assessment of various outcome measures, including goal and non-goal directed actions, forces and torques, operating time, etc. The quality of the operative end product can then be assessed by the end product analysis, which includes accuracy, error, tissue damage (e.g. water tightness of anastomosis) and with the histological outcome (total mesorectal excision for instance).

External feedback was found to be critical to the learning process [2, 14, 15]. In particular, summary expert feedback, which takes place after completion of the task, was found to be more efficacious than concurrent feedback, which occurs during completion of the task [16]. It has been reported that there is poor correlation between the procedure effectiveness and the end product analysis [17]. This indicates that a range of pattern of movements during the exercise can result in a similar quality of end product. Interestingly, the end product quality was not adversely affected by the surgeon’s fatigue [17]. However, this was investigated on simulated task on VR and it would be interesting to see the correlation between the quality of complex surgery such as laparoscopic total mesorectal excision and the end product. The end product analysis has also been found to be suited for skills assessment, despite its low reliability [2].

Deliberate practice is one of the components of the integrative (associative) phase of motor skills acquisition. It is most effective when distributed throughout many sessions, as opposed to one long single session, because the intervals between sessions allow knowledge of the new skill to be consolidated [18].

For simulation to be successful, the training must be recreated outside of the operating room for all students of surgery, including those with basic as well as advanced skills. A simulated learning environment can be easily controlled and adjusted to varying levels of difficulty. It can also occur in a more step-wise fashion than one performed in the operating room, where learning relies on random chance and opportunity [8] (Fig. 6.3) The step-wise model for learning a complex surgical procedure is based on the premise of building skills gradually using previous accomplishments, with more advanced skills built on a foundation of basic skills. Every complex surgical procedure can be broken down into several simple tasks that are required to complete most complex operations [19]. One of the first teaching modules for laparoscopic surgery was created by Rosser et al. and involved three basic task stations to teach a two-handed technique, coordination in handling tissue and manipulation of a sewing needle [20].

Further evolution of the programme led to abandoning the simplified peg exercise program and validating clinically significant exercise models, including vascular control, lesion excision, appendicectomy, mesh repair, perforation closure, and hand-sewn anastomosis [21]. (Fig. 6.4) The basic task analysis and its performance is the first step in the step-wise model of learning the complex surgical tasks. The second step of this process is frequently performed concurrently with the first step, and is often unrecognized. Called visual-spatial training, this step places emphasis on a three-dimensional relationship of anatomic structures and surgical manoeuvres and stresses the importance of proper knowledge of the key relationships of vital structures and dynamic anatomy during an operation. While building on these skills, the trainee can then proceed to the third step, which is practice of the set up and exposure (Fig. 6.3) Interestingly, this step is also under-appreciated by trainees, but valued by experts. In fact, mastering the art of set up and proper exposure enables the surgeon to avoid struggling with poor ergonomics and insufficient exposure during laparoscopic procedures. The sequential steps of the operation become much easier when a proper set up is used [8].

The final step in the step-wise model is the procedural component. Ideally, the student should be able to practice and master the full procedure in the controlled simulated setting. Additionally, while the first three steps can be accomplished with the use of low fidelity simulators, the fourth step can be accomplished with high fidelity inanimate physical models, virtual reality simulators, or cadavers. The procedure should be performed repeatedly in the simulated environment until proficiency is achieved. The term isoperformance was introduced by Jones to describe how learning by two different methods will transfer the same skills, albeit with different efficiency [22]. It is also important to remember that surgical training will always require operating on a real patient, thus the presence of thoughtful mentors to guide the trainee is essential. In fact, technical proficiency is only a single component of the mix; simulation enables the trainee to focus more on the other aspects of the “mix” during clinical exposure, such as obtaining higher-level skills to learn more complex steps of the operation or how to manage complications [8, 23].

Inanimate physical models, including box trainers and bench models, are safe, portable, reproducible, accessible, and readily accepted by novice trainees. They offer a low-fidelity environment for practicing basic, discrete skills and tasks, but not full operations [21]. They also provide true haptic feedback and allow the acquisition of skills that are transferable to complex laparoscopic tasks [24]. Together with the web-based study guidelines, the use of inanimate physical models has been incorporated into the curriculum of the FLS (Fundamentals of Laparoscopic Surgery) programme, endorsed by the American College of Surgeons and the Society of American Gastrointestinal and Endoscopic Surgeons [25, 26]. One of the main advantages of the inanimate physical model is the ability to exercise the dynamic coupling of hands, eyes, and the interposed camera, a skill that is crucial in the operating theatre and not achievable using virtual reality simulators [27]. Sroka et al. found that training to proficiency using the FLS simulator in the surgical residency curriculum has resulted in improved resident performance in the operating room [28].

Virtual reality (VR) simulators, for which the setup time is minimal can provide immediate feedback and metrics on error rates, precision, and accuracy [29]. The identification and subsequent management of errors is crucial to safe surgical practice [30]. Grantcharov et al. was able to detect a higher economy of motion and fewer errors while performing laparoscopic cholecystectomy, as well as shorter operative time after using the MIST-VR simulator [31] (Fig. 6.5). This has been confirmed by others [32]. One of the disadvantages of the high-fidelity VR simulator is the cost, though this can be outweighed by the benefits. Additionally, one has to remember that surgical training without use of a simulator is associated with significant expense. Bridges and Diamond investigated the cost of having general surgery residents present in the operating room and estimated the annual cost to be $53 million in the United States in 1997 [33]. Conversely, Aggarwal et al. calculated the transfer effectiveness ratio of a modestly expensive VR simulator to be 2.28 for laparoscopic cholecystectomy, translating into every hour spent on VR simulation, and reducing time to achieve proficiency in vivo by 2.3 h [34], thus limiting the costs of training in the operating room.

More recently, the high-fidelity environment was reproduced using the hybrid ProMIS simulator with synthetic anatomical tray and VR metrics (Fig. 6.6). Precise time measurement, instrument path length, and smoothness of movements can all be recorded for analysis [35]. A study comparing the ProMIS simulator with the cadaver model for laparoscopic left colectomy found that technical skills acquisition was better using the simulator. The main overall occurrence in both models was error in the use of retraction, while the specific occurrence in both models was bowel perforation [36]. Essani et al. found that simulated laparoscopic sigmoidectomy training affected the responsiveness of surgical residents with significantly decreased operating time and anastomotic leak rate [37] (Figs. 6.7, 6.8 and 6.9) Another report, however, found no correlation between the simulator-generated metrics (path length, smoothness of movements) and the content valid outcome measures (accuracy error, knot slippage, leak or tissue damage) [38] (Fig. 6.10).

Animate models (porcine and canine), have also been used in training to enable the trainee to practice on live animals, experience the quality of live tissues, and address haemostasis. Apart from ethical considerations, the main obstacles or deficiencies of this model are the differences in anatomy and the need for technical support.

Other options include human cadaveric models, which more closely reflect reality. Milsom et al. performed one of the first feasibility studies of cadaveric laparoscopic proctosigmoidectomy in 1994, designing a standardized technique of oncologic resection [39]. Studies have found participants in cadaver laboratories to be highly satisfied with the teaching value and reliability of the materials used [40, 41]. The main advantages of a cadaveric model include tissue consistency and preservation of anatomic planes, which are very important for the learning process [42]. Le Blanc et al. compared human cadavers and augmented-reality simulators for acquisition of laparoscopic sigmoidectomy skills, finding cadavers to be more difficult but better appreciated than the simulators [36]. Human cadavers were also found to be superior in laparoscopic colectomy training when compared to high-fidelity virtual reality simulators [43]. A recent study reported that colonoscopy training with deployment of stents for colonic strictures in a cadaver model has content, construct and concurrent validity [44]. The major difficulties encountered with cadaver models include their limited availability, high cost, ethical concerns, need for specialized facilities and personnel, andthe inability to exercise haemostasis.

Video analysis is one of the least examined and least understood training methods in laparoscopic surgery. This is ironic because laparoscopic surgery is often referred to as video surgery. In fact, use of a recording device in the operating room is almost a universal standard today. With video analysis, the trainee has the opportunity to review the recorded material individually or with the trainer, who can then provide the necessary critique. Additionally, material can be stored for future use as a reference tool during independent practice, particularly during a low case volume schedule. Recorded material can be also used for grading and progress evaluation of the trainee. Ideally, review of the video should take place within a few days (preferably 1–2) of the procedure being performed, so that it is still fresh in both the trainee’s and trainer’s memory and the intra-operative comments can be remembered.

In 2008, a systematic review of randomized and non-randomized data by Sturm et al. was inconclusive as to which skills learned during laparoscopic simulation were transferable to the operating theatre [45]. More recently, however, Sroka et al. was able to demonstrate that skills learned from the FLS program improved performance during laparoscopic cholecystectomy [28]. Likewise, a systematic review by Zendejas et al. of simulation-based laparoscopic surgery training was found to have significant benefits when compared with no intervention, and moderately more effective when compared to non-simulation intervention (e.g. video instruction) [46]. Despite the validation of many inanimate and VR simulators, there is only one study demonstrating a direct effect of simulation on improved performance in colorectal surgery. In the study by Palter and Grantcharov, a comprehensive curriculum consisted of a VR simulator, a cognitive training component and cadaver lab training. The curricular-trained residents were found to demonstrate superior performance during right colectomy when compared with conventionally trained residents [47].

Little is known about teaching the new techniques of colon resection with single-incision laparoscopic surgery (SILS) or natural orifice trans-luminal endoscopic surgery (NOTES). The only available report by Buscaglia et al. examined the usage of ProMIS simulator for training in NOTES sigmoidectomy, demonstrated a positive outcome for surgical endoscopists with a 42 % reduction in operating time [48] (Figs. 6.11, 6.12, 6.13 and 6.14).

According to recent reports, the use of laparoscopy in colorectal surgical procedures is gradually increasing, although it has been a very slow process, with varying adoption rates. In 2009, 50 % of all colon resections in the United States were performed laparoscopically [49]. This was up from 31.4 % in 2008, as reported by a different study only a year before [50]. Similarly, in the United Kingdom, laparoscopy was used in more than 40 % of all colectomies performed in 2013 (according to Hospital Episode Statistics) compared with 5 % of procedures performed in 2006.

Many factors are responsible for these increased adoption rates. During the early adoption phase of laparoscopy, there were concerns about oncological safety, due to reports of trocarsite recurrence [51]. This issue was subsequently eliminated with appropriate wound protection, tissue handling, and proper oncological dissection. Other factors responsible for slower adoption included the need for multi-quadrant dissection, advanced laparoscopic techniques (for intra-corporeal vessel control, large surface dissection, bowel transection, and anastomosis), difficult retraction and exposure, increased operating time, and the cost of the laparoscopic equipment.

The implementation of laparoscopic rectal dissection has been even slower than with colectomy. This is due mainly to tumour location within the rigid confines of the pelvis, difficult and unstable retraction, visualization, and poor ergonomics for the surgeon. In 2009 laparoscopic total mesorectal excision ranged from 12 % in the United Kingdom to 19.6 % in Canada and 26 % in Australia in 2008 [52, 53].

When considering potential candidates for laparoscopic colorectal surgery training, three groups should be identified. The first group is general surgery residents who possess basic laparoscopic skills, but are in the process of acquiring colorectal knowledge and advanced laparoscopic skills simultaneously, in order to perform colectomies. The second group is individuals who are trained in general surgery and are undergoing postgraduate training (colorectal surgery or minimally invasive surgery training). This group is adept in the basics of laparoscopy and has sufficient colorectal knowledge, but lacks experience with laparoscopic colectomies. Finally, the third group is surgeons who have never been trained in laparoscopy but have sufficient experience in open colorectal surgery.

Identifying to which group the trainee belongs is critical and will determine which training model is the most suitable for the trainee. As a result of the time constraints of surgical training mentioned earlier, the first group of trainees is likely to be able to complete only basic training in laparoscopic colectomy. Indeed, unless they engage in postgraduate training, it is unlikely that they will become proficient in advanced laparoscopic colectomies or rectal resections. Conversely, the second group of trainees, because they are involved in postgraduate training (colorectal surgery residents or minimally invasive fellows), is capable of developing the full set of advanced skills required for advanced colectomies and rectal dissection. For the third group, it is expected that they will achieve the same set of skills as the first group, albeit through a different training path.

Regardless of the group, it is important to realize that each is comprised of individuals with different learning potential and, thus, will require individual learning curves.

One of the primary reasons behind the slow adoption of laparoscopy into colorectal surgery has been referred to as a “steep learning curve”. However, this description of the learning process is not valid for two reasons. Firstly, the term “learning curve” should be replaced by “proficiency-gain curve”, which more accurately describes the process of increasing levels of technical and non-technical proficiency, rather than simply ‘learning’, which implies a purely cognitive process [54]. Secondly, the term “steep” is a misnomer because it implies rapid acquisition of skills during the time period, quite contrary to what is observed in practice. Therefore, it is more accurate to reason that a “long proficiency-gain curve” is why the adoption process of laparoscopic colorectal surgery has been slower.

There have been many metrics used to describe the learning (proficiency-gain) curve of laparoscopic colorectal surgery. The most commonly used metrics include operative time, conversion rate, complication rate, and readmission rate [55–57]. The main limitations of these individual metrics, however, are their non-transferability throughout different hospitals and individual surgeons [58]. A recent systematic review and multi-centre analysis of multiple metrics, utilizing the risk adjusted CUSUM methods performed by Miskovic et al. estimated the learning curve in laparoscopic colorectal surgery to be between 88 and 152 cases [59]. This number differs from 60 found by Tekkis et al. in 2005 and is in sharp contrast to the range of 11-15 cases identified by Simons et al. in 1995 [57, 60]. This discrepancy can likely be explained by the increasing complexity of the laparoscopic procedures and the application of laparoscopy to more challenging clinical scenarios. Recently, Mackenzie et al .looked at the influence of mentored training on the proficiency-gain curve. They found that 40 cases were required in order for supervised fellows to achieve confidence in laparoscopic colectomy [61].

The goal of every training programme is to produce a competent trainee who can individually perform procedures in a safe manner. This competency is difficult to measure prospectively. Indeed, the ultimate method of evaluation used to be the retrospective analysis of clinical outcomes, which were often negative, and included mortality and morbidity data. It is important to recognize that competence is multifactorial and dependent on surgical skills, cognitive factors, personality traits, and decision-making [62].

The proficiency gain should be closely monitored during the training period. Thus far, various tools have been utilized, including the OSATS (Objective Structured Assessment of Technical Skills). Unfortunately, the value of this tool in the assessment of advanced laparoscopic procedures (e.g. colectomies) has been restricted, due to the ceiling effect and the learning curve of the assessor [63]. The Global Assessment Score (GAS) tool on the other hand, has been found to effectively assess and monitor the proficiency gain. This tool evaluates generic task steps for laparoscopic colorectal resection in a formative way, allowing for the identification of areas during the operation that may be more difficult to master and thus require more practice [64]. Another tool designed to evaluate technical competency in a summative manner is the Competency Assessment Test (CAT). Designed by a reiterative expert consensus (Delphi) method, the CAT evaluates the process of performance (instrument use, tissue handling) as well as the end product of the procedures. It has been validated for examining and credentialing use [65].

While maintaining competency is crucial, little has been made available with regard to the individual retention of learned skills. In fact, while institutions with higher case volumes have shown improved outcomes, some argue that it is the number of hours spent on deliberate practice by the individual surgeon that ultimately determines the level of expertise [66, 67]. However, others contend that tracking outcomes is likely to be the only reliable way to assess competency [68].

Safe and accurate operating can also be assessed using the observational clinical human reliability analysis (OCHRA). This concept, proven in other laparoscopic procedures, is based on video analysis of errors made during procedures and allows for identification of underlying performance-shaping factors [69].

According to Gagne, an essential component of the external conditions for learning motor skills is the provision of feedback as close to the time of performance as possible [76]. This feedback and constructive criticism are only possible by mentoring and close supervision, both of which impact the trainee’s performance. In fact, intra-operative instruction has been found to decrease the rate of errors in a randomized laparoscopic suture study [5]. Additionally, in a meta-analysis of outcomes of 6064 patients, Miskovic et al. found that trainees with an appropriate level of supervision generated the same complication, conversion, and mortality rates as expert laparoscopic colorectal surgeons. This is extremely important in the context of the modern world where patient safety is paramount. In the same study, the authors compared the outcomes of non-mentored and mentored trainees and found that the experienced trainer can further aid in intra-operative decision-making and the comprehension of anatomy [62]. Case selection for training purposes is also indirectly related to supervision, with laparoscopic sigmoid resection being the easiest [77]. Similarly, appropriate patient selection for training purposes revealed that male sex, past surgical history, obesity, high ASA class, and colorectal fistulae were associated with higher conversion rates [62].