Introduction

A new laparoscopic training curriculum is being developed for the North-East Surgical School of the Netherlands (NESSN). The NESSN consists of the University Medical Center Groningen and all affiliated teaching hospitals. INtraoperative Video-Enhanced Surgical procedure Training (INVEST) is shown to have a positive effect on the completion of the early learning curve for surgical procedural training by both increased efficiency and increased effectiveness [1, 2]. Based on this, we aim to construct a laparoscopic curriculum that provides a safe, uniform, efficient, and procedure-specific training program for a series of laparoscopic gastrointestinal procedures, transferrable between hospitals throughout the region. This curriculum will consist of well-defined procedure specific key steps that are incorporated in INVEST video fragments and a validated assessment tool based on these key steps. This curriculum is targeted at residents who have successfully completed a preclinical training course including simulator and wet-lab courses.

The aim of this study was to determine expert consensus regarding essential steps for laparoscopic right hemicolectomy and laparoscopic sigmoid colectomy, using a Delphi methodology. The key steps identified in this study will be the basis for creating the INVEST video’s for both procedures. Eventually the goal is to create and validate a procedure-specific assessment tool.

While other training curricula in laparoscopic colorectal surgery have been validated and published [3], none have focused on dividing surgical procedures into well-identified segments, which can be trained and assessed separately. This enables the surgeon and resident to focus on a specific segment, or combination of segments, of a procedure. Furthermore, it will provide a consistent and uniform method of training for residents rotating through different teaching hospitals within our region. A structured, visually demonstrated training curriculum might also reduce the risk of miscommunication and therefore add to the safety of resident training. The goal of this study was not to reinvent well-established guidelines, but to determine consensus on these guidelines within our teaching region. Bethlehem et al. successfully determined consensus on the key steps in laparoscopic appendectomy and cholecystectomy within the same teaching region [4]. These key steps will form the basis for the INVEST video-assisted side-by-side training curriculum. Therefore, a deliberate choice was made to only include NESSN teaching staff as experts. When successful in the future, the curriculum can be expanded to other teaching regions.

Methods

Delphi methodology

In order to reach consensus on the essential procedural steps for laparoscopic right hemicolectomy and sigmoid colectomy, the Delphi method was used. The Delphi method is a well-established, anonymous, group process in which ideas are expressed to the participants in the form of a questionnaire [5, 6]. Responses to the items in the questionnaire are collected and analyzed along with added comments of the experts. This leads to the adding, revising, or dropping of items to be used in a subsequent round, until group consensus is reached [6].

Expert panel

The expert panel was selected to represent currently practicing surgeons within the region of the North-East Surgical School of the Netherlands, who are responsible for the training and education of our surgical residents in laparoscopic colorectal surgery. All experts have performed more than 100 laparoscopic colorectal procedures and have more than 5 years experience in the field. For this study, 22 experts in 12 hospitals were asked to participate via email. They were sent a link to an anonymous online questionnaire. Sixteen surgeons agreed to participate by filling out the questionnaire. Throughout a period of 3 months, reminders were sent in the form of a personal email and/or phone call.

Delphi round 1

A list of procedural steps of the laparoscopic right hemicolectomy and laparoscopic sigmoid colectomy was identified. The steps were compiled from surgical textbooks and current guidelines from the Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) [7], the European Association for Endoscopic Surgery (EAES) [8, 9] and the Association of Surgeons of the Netherlands (NVvH) [10]. Each step identified from these sources was included in the initial questionnaire (Tables 1, 2). Each expert was asked to rate the steps on a Likert scale from 1 “not important” to 5 “essential”, thus valuing the steps as more or less important parts of the total process. The opportunity to comment or clarify was offered at the end of the questionnaire.

Table 1 Delphi round 1: steps identified for laparoscopic right hemicolectomy
Full size table
Table 2 Delphi round 1: steps identified for laparoscopic sigmoid colectomy
Full size table

Delphi round 2

As the responders could not be identified from the non-responders in this anonymous online questionnaire, all experts were invited for round 2. Based on statistical analysis of round 1, steps were either excluded, considered a key step, or in need of reassessment in round 2. This time, the experts were asked to motivate their answer when they rated a step as 1 “not important” or 2 “sometimes important”. Feedback from the first round taken into consideration and two questions were rephrased. We decided in advance to stop after two rounds, whether consensus was reached or not, because identification of essential steps was not to be expected in subsequent rounds.

Statistical analysis and consensus

SAS version 9.2 (SAS Institute Inc., Cary NC, USA) was used for statistical analysis. Cronbach’s alpha was calculated for internal consistency between the experts. There are no established criteria for determining consensus using a Delphi methodology. The aim of this method is to reach consensus on relevance of each item. Means and 95 % confidence intervals (CI) were calculated for each step to identify relevant steps. The 95 % CI were used to quantify the variability of the experts’ responses. Rated on a Likert scale 1–5, the CI were between 1.00 and 5.00. A step was accepted as a key step if the lower confidence limit was ≥3.00. A step was excluded if the upper confidence limit was <3.50. All steps that did not meet the above-mentioned criteria were reassessed in round 2, as insufficient consensus was established in the first round.

In Delphi round 2, a cutoff point for consensus was predetermined. Consensus was established when at least 80 % of the respondents rated the step as ≥3. This step was then accepted as a key step. If the 80 % threshold was not reached, the step was excluded.

Results

Of the 22 experts who were asked to participate, 16 agreed. Of these 16 experts, 15 (94 %) completed the questionnaire for the laparoscopic sigmoid colectomy and 14 (88 %) completed the questionnaire for both procedures. Of the 14 who completed the first round, 13 (93 %) completed the second round. Cronbach’s alpha was calculated to be 0.79 for laparoscopic right hemicolectomy and 0.91 for laparoscopic sigmoid colectomy, showing high internal consistency between the experts.

In Delphi round 1 (Tables 1, 2), consensus was reached on 23 steps of the right hemicolectomy (two steps were excluded, 21 steps were accepted as key steps); the remaining eight steps were reassessed in round 2. For the sigmoid colectomy, consensus was reached on 33 steps (nine steps were excluded, 24 steps were accepted as key steps) and the remaining four steps were reassessed in round 2.

In Delphi round 2, of the eight reassessed steps of the right hemicolectomy, four steps were accepted as key steps, the other four were excluded (Table 3). For the sigmoid colectomy, all four reassessed steps were excluded (Table 4). A list of final key steps for both procedures is presented in Tables 5, 6.

Table 3 Delphi round 2: reassessed steps for laparoscopic right hemicolectomy
Full size table
Table 4 Delphi round 2: reassessed steps for laparoscopic sigmoid colectomy
Full size table
Table 5 Key steps identified for laparoscopic right hemicolectomy
Full size table
Table 6 Key steps identified for laparoscopic sigmoid colectomy
Full size table

Discussion

Statistical analysis showed good consistency between experts and between answers. On average, scores for individual steps were lower than expected. In part, this can be explained by how the Likert scale was labeled. Labeling steps as “1 not important, 2 sometimes important, 3 important, 4 very important, 5 essential”, resulted in many respondents frequently scoring “3 important”. We considered “important” the minimum threshold for a key step, in this study meaning a Likert score of 3 and above.

Some steps were excluded almost unanimously; others were excluded with a large variety of scores. Of the latter, many steps can be very important in only certain situations, i.e., they are important sometimes. Though these steps cannot be included in a list of key steps, they could be added to the INVEST video curriculum as optional steps for specific situations.

Steps excluded in round 1

Of the steps excluded by the experts in round 1, both for the right hemicolectomy and the sigmoid colectomy, three types can be identified. Firstly, practical and ergonomic steps concerned with positioning of the surgeon’s assistant and port placement. These steps may have been scored as “not important” because of personal preference, not necessarily influencing the quality of the procedure. Placement of ports is based on the experience and preference of the individual surgeon [9].

Secondly, a procedural set of steps was excluded: mobilizing the splenic flexure in the sigmoid colectomy. This typically represents the type of step that is very important in some situations, but not routinely required and therefore not a key step.

Lastly, three steps concerning safety were excluded: identification of the left and right ureter and the airleak test in the sigmoid colectomy. Of the latter, it could be argued that airleak testing can only be performed on a distal anastomosis and this step may be valued differently in a low anterior resection, though the expert panel listed no arguments for excluding the step. Recommendations on the identification of ureters are scarce. Identification of both ureters is advised in American guidelines, and identification of the right ureter in right hemicolectomy is advised on the EAES website [7, 8], whereas official EAES and NVvH guidelines do not address the issue [9, 10]. Our experts do not routinely perform this step. Ureter injuries are rare, but when they occur they are associated with significantly higher morbidity and mortality [11]. Risk of injury can be a legitimate reason not to open the retroperitoneum for the sole purpose of identifying the ureters, though in situations where this plane is opened for other reasons, good arguments can be made for exploration and identification. Consistent with this, the vast majority of our expert panel scored identification of the ureters as “sometimes important”. This step may be considered as an optional step in the video curriculum.

Steps excluded in round 2

For the right hemicolectomy, four steps were excluded after reassessment. Two of these steps were of a practical nature concerning port placement and positioning. A diagnostic laparoscopy is not routinely performed by our experts. Some experts commented that a quick inspection of liver and peritoneum will inevitably be part of the procedure, while no formal inspection of all quadrants is performed. This does not qualify as a diagnostic laparoscopy as such and cannot be included as a key step. Lastly, identifying the middle colic artery was excluded based on 23 % of experts scoring it “sometimes important”, mainly related to tumor localization.

For the sigmoid colectomy, all four reassessed steps were excluded. Again, the diagnostic laparoscopy was excluded for previously mentioned reasons, as was a practical step. Identifying and dividing the inferior mesenteric vein, and blunt dissection of the left colon to the level of the spleen, are only performed when there is insufficient length for an anastomosis.

The list of key steps identified in this study is less extensive than other published curricula and international guidelines. In a large international survey, Cheung et al. have shown that there is no consensus between experts in the adjacent field of laparoscopic TME surgery, and many differences seem to be related to experience and local protocol [12]. Differences between international guidelines and local practice illustrate the impact of local protocol and the value of obtaining consensus within a specific academic region. It can be speculated that every country, region, and even every hospital have its own protocol, and this Delphi consensus study could be expanded to different regions to develop a nationwide teaching curriculum in the future. Currently, local practice dictates our resident’ training curriculum. By reaching consensus on the laparoscopic right hemicolectomy and sigmoid colectomy, we can make the step toward a structured teaching curriculum that is exchangeable throughout the teaching region. It is important to stress that the steps eliminated by our experts for good reason are not necessarily unimportant steps. The procedural key steps determined in this study are a ‘minimum requirement’, and a procedure should not be limited strictly to these steps. Many steps that were excluded as key step may be very important depending on case-by-case variables.

In the traditional master-apprentice model of surgical training, demonstration by the supervising surgeon has the clear disadvantage that part of the procedure is lost to the trainee, who has to wait until the next procedure to perform the step himself. Intraoperative video demonstration minimizes the frequency of intervention by the supervisor and maximizes the actual operating time for the trainee. INtraoperative Video-Enhanced Surgical procedure Training (INVEST) is shown to have a positive effect on Objective Structured Assessment of Technical Skills (OSATS) items ‘knowledge of the procedure’, ‘time and motion’, ‘use of assistance,’ and OSATS sum score. INVEST significantly enhances technical and procedural skill development during the early learning curve for laparoscopic cholecystectomy, compared to the traditional master-apprentice model [1, 2]. No other laparoscopic training curriculum has used video fragments intraoperatively. Key steps identified in this study will be used to create INVEST video’s for both procedures. Eventually the goal is to create and validate a procedure-specific assessment tool, suitable for incorporation in the training and certification process of gastrointestinal surgeons.

Conclusion

The procedural key steps for laparoscopic right hemicolectomy and laparoscopic sigmoid colectomy were established, using a Delphi methodology with an expert panel of 22 surgeons within our teaching region. This will form the basis of a uniform, transferrable, and efficient video-assisted training curriculum, which is being developed.