With the increasingly spreading of laparoscopic surgery, simulation-based surgical training has been introduced into the surgical residency curriculum [1, 2]. Many studies have demonstrated the positive impact of simulation on technical skills for basic laparoscopic procedures [3,4,5,6,7], shortening the learning curve (LC) of young surgeons, and reducing the number of adverse events in the operating room [8, 9].

Laparoscopic liver resection (LLR) has been growing in terms of the number of procedures performed and their complexity [10]. However, LLR has developed very slowly due to various obstacles, which include a significant LC, especially to become familiar with liver mobilization, parenchymal transection techniques, and overall control bleeding, along with specific devices [11]. Recently, the Southampton Guidelines advocated that the laparoscopic approach should be considered as the standard practice for the lesions of the left lateral and anterior segments at the beginning of the LC [12]. Indeed, laparoscopy has become the standard surgical approach for left lateral sectionectomy (LLS) and for wedge resection of the anterior segments (WRAS), due to its anatomical accessibility [13,14,15].

Today, only few studies have focused on simulation-based teaching programs in laparoscopic liver surgery [16]. Most training models used for LLR were the animal model (swine liver) and the cadaver [17,18,19,20,21]. However, no standardized training programs have been validated for LLR, whereas specific assessment tools for training in advanced laparoscopic surgeries have recently been developed and validated [22,23,24,25,26,27]. No performance assessment scales have been reported so far in LLR. Such scales are needed to assess both the educational value of training programs and the technical skills of trainee surgeons, especially for “basic” LLR, such as WRAS and LLS. As a bridge between training and the operating room, a dedicated skills scale may be enabled to shorten the LC of young surgeons and could also be used as an operative roadmap in the operating room to improve patients’ safety.

Therefore, the purpose of this study was to identify the essential steps in WRAS and LLS, and to develop an objective scale for the assessment of technical skills during these procedures. The identification of procedural steps was performed by hierarchical task analysis (HTA) and then by expert consensus, using the Delphi methodology.

Material and methods

The present study did not involve human subjects or laboratory animals. Thus, approvement by institutional review committee is not needed.

Hierarchical task analysis

HTA [28] was adopted in the present study. The purpose was to identify the successive discrete steps that are required to complete a full laparoscopic WRAS (segments 2, 3, 4b, 5, and 6), and a LLS. Literature was searched in order to identify the different laparoscopic operative techniques [29,30,31,32,33]. A panel of video performed by liver laparoscopic surgeons illustrating these techniques was selected from one local database and one online database (WeBsurg). The aim was to develop a scale that could assess operative performances of trainee surgeons involving in a teaching program out of the operating room. Two experienced surgeons in both laparoscopy and hepatobiliary surgery (DF and DJB) and one surgical resident (TG) reviewed independently the videos. Each reviewer listed all the consecutive discrete steps required for completion of each procedure. Then, reviewers pooled together the results and elaborated a joint list of steps. A numerical scoring scale ranging from 1 to 3 was assigned to each selected step.

Creation of the laparoscopic liver skills scale (LLSS) by Delphi

All consecutive steps generated by the HTA were submitted to a panel of 15 international experts using the Delphi methodology. In order to present a technically and educationally relevant objective skills assessment tool, surgeons who were experts in both liver and laparoscopic surgery were invited to participate in the Delphi. Moreover, the invited experts should have prior publications on LLR, be key opinion leaders in the field of laparoscopic liver surgery and have an active involvement in simulation-based surgical training programs. They were invited to participate in the project via email. Experts from different geographic zones were recruited (Europe, US) in order to develop an internationally relevant scale. All participating experts were volunteers and informed consents were obtained.

The Delphi methodology is a systematic and interactive forecasting method used to obtain consensus among a panel of experts, who are consulted over several rounds. After each round answers are collected, analyzed, and submitted back in an iterative fashion to the group. Over the successive rounds, group opinion should converge toward consensus [34, 35].

Two rounds of the Delphi survey were conducted. Experts were asked to rate, using the Likert scale ratings from 1 to 5 (1: strongly disagree, 2: disagree, 3: undecided, 4: agree, 5: strongly agree) with the guidance of the following questions: “Do you think that three levels of assessment are enough? Do you think the number of subtasks is adapted? By subtask, do you think the description of the skill levels is appropriate? Do you think that the proposed subtasks cover all the skills required to perform WRAS and LLS?” Experts were invited to comment their answers in order to modify or add steps during the second round. Mean and standard deviation obtained for each step during the first round were presented to the experts during the second round. A rate of agreement (RoA) was calculated as a measure of consensus among the experts: [(Agreement—Disagreement)/(Agreement + Disagreement + Indifferent)] × 100. Steps that reached consensus during the second round were included in the final scale.

Statistical analysis

Means and standard deviations (SD) were calculated for all the subtasks. Cronbach’s alpha was calculated for internal consistency among the experts. The consensus was predefined using Cronbach’s alpha > 0.8 according to a global Delphi consensus study on defining and measuring quality in surgical training [36]. The subtasks were adopted when they were rated 4 or 5 on the Likert scale by 80% or more of the experts. Data were analyzed using SPSS version 20.0 (SPSS, Chicago, Illinois, USA).

Results

A draft of the key subtasks was created using the literature, video analysis, and direct observations of WRAS and LLS cases in 2 teaching hospitals. As the videos collected recorded the intra-abdominal camera view, some aspects of the full procedure were not visualized. Therefore, the reviewers additionally listed the steps assessing patient positioning, preoperative checklist, and abdominal access. After the semi-structured review with two experienced surgeons and a surgical resident in the final process of the HTA, 15 essential subtasks for full laparoscopic WRAS and LLS were extracted for evaluation in the Delphi survey (Table 1). All subtasks were scored from 1 to 3 points. A score of 3 points considered the specific skill as acquired, whereas a score of 1 point considered it as non-acquired.

Table 1 After the semi-structured review, 15 essential subtasks for full laparoscopic wedge resection of the anterior segments and left lateral sectionectomy were extracted for evaluation in the Delphi survey
Full size table

The Delphi survey was conducted between March 2019 and January 2020. Five of the 15 international experts agreed with our invitation and voluntarily participated in the Delphi survey. Participating experts (one American and four French experts) were experienced surgeons in laparoscopic liver surgery and involved in simulation-based surgical training programs. All experts completed the 1st and 2nd round survey. A total of 3 over 15 subtasks did not reach the predefined level of consensus (Table 2). Cronbach’s alpha was 0.78 after the 1st round survey. The main drawback identified in the experts’ comments was that the number of subtasks was insufficient and did not cover all the skills required to perform a full laparoscopic WRAS or LLS. Therefore, based on these comments, 13 subtasks were reformulated, 4 subtasks were added, and a revised skills scale was developed, including the division of the revised skills scale in three steps. This revised skills scale was submitted to the five experts who had completed the first round. Results of the second round are detailed in Table 3.

Table 2 Results of the Delphi 1st round on the list of surgical steps generated by the hierarchical task analysis
Full size table
Table 3 Results of the Delphi 2nd round on the list of revised surgical subtasks
Full size table

After the Delphi consensus was achieved with the results of the 2nd round survey (Cronbach’s alpha 0.84), 19 subtasks were adopted. The LLSS was finally created based on the selected subtasks resulting from the Delphi 1st and 2nd round surveys (Table 4). A score of 3 points considers the specific skill as acquired. Thus, the LLSS maximum score for the full scale is 57 points. For WRAS and LLS, procedure’s skills could be considered as acquired when the LLSS score is 45 and 51 points, respectively.

Table 4 The Laparoscopic liver skills scale was finally created based on the 19 selected subtasks resulting from the Delphi 2nd round survey
Full size table

Discussion

The gradual introduction of simulation-based training programs using a dedicated rating scale could reduce the clinical impact of LC and improve patients’ safety [37]. We used a comprehensive method, including the HTA and Delphi methods to determine the essential subtasks of full laparoscopic WRAS and LLS. With those results, we developed the LLSS for measuring the skill set for the education of safe and secure laparoscopic WRAS and LLS procedures. The LLSS is composed of three main parts: patient positioning and intraoperative preparation (task 1 to 8), the core part of the WRAS and LLS procedure (tasks 9 to 14), and completion of procedure (task 15 to 19).

In this study, we inquired about each task in the Delphi extra survey that addressed the concept of “safety” procedure. That is why 19 tasks were selected in the LLSS, including some points of safety: initial check, patient—port and aid positioning, Pringle maneuver, intraoperative ultrasonography, intraoperative bleeding and biliary leakage management, management of open conversion, and final check. In that manner, the LLSS offers the advantage of being able to assess the level of skills acquired while highlighting important safety points during training program of WRAS and LLS procedures. Moreover, the LLSS emphasizes the need to become familiar with liver mobilization and parenchymal transection techniques, along with specific devices. Indeed, the proposed rating scale may be enabled to shorten the LC of young surgeons and could also be used as an operative roadmap reducing the number of adverse events in the operating room.

The advantages of the Delphi method in order to obtain consensus among experts are well described: the anonymous nature of the process prevents a dominant member of the group from influencing the group’s opinion. Furthermore, the questionnaire was completed by email and did not require for the experts to physically meet. As used in the present study, Von der Gracht et al. suggested that a RoA was an appropriate measure of consensus particularly when Likert scales were used [38]. Moreover, a two-round Delphi was conducted in order to limit the number of non-responders and to avoid forced consensus [39]. In this study, the notable limitations of the Delphi method include the fact that the selection of questions submitted to the experts were in part controlled by the Delphi facilitators and that interest of experts could diminish with consecutive rounds. Another concern of the present work is the limited number of experts who participated in the Delphi survey, that could impair the reproducibility of the study. Among the 15 international invited experts, only five agreed to participate and most of them came from French institutions. However, laparoscopy has become the standard surgical approach for WRAS and LLS, and laparoscopic LLS is a standardized step-by-step procedure. Hence, there could be a worldwide acceptance of the present LLSS. Furthermore, the usability, feasibility, and validity of this assessment scale still needs to be studied through a dedicated LLR training program. Validation of the LLSS should include comparison between experts and novices in laparoscopic liver surgery (construct validity) and comparison between operative outcomes and obtained scores in order to assess its educational value.

Recently, specific assessment tools for training in advanced laparoscopic surgeries have recently been developed [22,23,24,25,26,27] and have the advantages of identifying specific areas of the procedure that require improvement. The very detailed nature of these scales makes it an interesting tool for training purposes. However, specific assessment tools did not gain worldwide acceptance compared to the OSATS scale [40]. By identifying specific areas of the procedure that require improvement, these specific assessment tools and the present LLSS should facilitate constructive feedback and thus deliberate practice. It may also be used for research in surgical education.

The Southampton Guidelines presented clinical practice guidelines designed specifically to ensure the safety of LLR setting up [12]. Indeed, a progression in skill set and competency is required to safely perform LLR [16], whereas the number of complex resections in hepatobiliary surgery performed by surgeons at the end of their residency is estimated to be less than five procedures, predominantly by open surgery [41]. Furthermore, an international survey on minimally invasive training in registered standard hepatobiliary fellowship programs revealed that fellows performed on average nine LLR yearly, which is obviously far too less to meet today’s high-quality practice standards [42]. Recently, Halls et al. showed that after 46 procedures, the outcomes of the “early adopters” (i.e., surgeons who received laparoscopic training and developed their practice in the stage of technical “optimization” of LLR) were comparable to those achieved by the “pioneers” (i.e., surgeons who developed their practice in the earliest stage) after 150 procedures in similar cases [43]. As such, the LC in LLS could be dramatically reduced with both a specific training and an increased exposure of trainees to LLS. The Southampton consensus guidelines recommend fellowships in high-volume centers, proctored programs and courses to facilitate training [12].

The number of procedures required to gain competency in minor LLR ranged from 15 to 64 cases [16, 44, 45] through literature. This heterogeneity can be attributed to three factors that differed from one study to another: (a) the procedure itself (the LC of now standardized procedures, such as LLS seem shorter than that of a wedge resection or segmentectomy), (b) the main endpoint used to define LC, and (c) the surgeons’ experience, with surgeons who did or did not receive laparoscopic training and developed their practice in the stage of technical optimization of LLR. For the specific concern of LLS, Ratti et al. collected a total of 245 LLSs performed across four centers by experienced surgeons [46]. In this study, the operative time was chosen as the marker of LC and the cut-off point for LC was determined after 15 LLSs. The present LLSS is a specific assessment tool and has the advantage of identifying specific critical points of safety of WRAS and LLS (scored from 1 to 3 points), that should be achieved by trainees in a dedicated training program to be considered competent before performing it in the operating room. Indeed, trainees could be “ready to perform independently” WRAS or LLS when they obtain a maximal score (i.e., all tasks were scored 3 points); “ready to perform with supervision” when 1 to 4 tasks need guidance (i.e., 1 to 4 tasks were scored 2 points); and “not ready to perform” on a real patient when more than 4 tasks were scored 2 points or when 1 task was scored 1 point. Forthcoming studies should assess the number of procedures needed in each category to achieve proficiency in a simulated setting. However, this number may depend on several factors, namely, the trainee him/herself and the simulation tool used.

Laparoscopy is now increasingly used for more complex liver resections [10, 12, 16, 37, 44]. Nevertheless, a shift in real LC compared with an ideal one shows that the increase in difficulty was attempted only after the initial LC was completed, highlighting that surgeons have a perception of their degree of training before performing more complex procedures [16, 44, 47]. The LLSS was developed to assess the level of skills acquired during training program of WRAS and LLS procedures, that represent the first steps of the LC [12]. However, such critical points as safety points, and the need to become familiar with liver mobilization and parenchymal transection techniques are emphasized by the LLSS, which can be extrapolated as an operative roadmap to other complex procedures.

Using training models to simulate complex laparoscopic liver procedures has been poorly studied so far [16]. The main training models used are the swine model and the human cadaver, including the Thiel model [16,17,18,19], which is probably the closest to reality. Swine are commonly used in training and offer tactile feedback, facilitating the acquisition of fundamental skills. More recently, the implementation of a perfused ex vivo swine liver training model has increased the fidelity of the animal model by simulating intrahepatic bleeding [18, 19]. However, the anatomy of swine liver differs from those of humans. Human cadavers have identical anatomical conditions as real patients and enable lifelike surgery. Their use as a laparoscopic training model has already been reported in several studies [48,49,50]. Of note, the main drawback of formalin-fixed cadavers is tissue rigidity. In contrast, Thiel bodies are soft and flexible and maintain their natural color. These corpses remain more similar to in vivo conditions without releasing harmful substances into the environment [49,50,51,52,53]. Furthermore, vascular perfusion of Thiel-embalmed cadaveric tissues with colored solutions for endovascular procedures and flap raising education has been described [54, 55]. Recently, Rashidian et al. [17]] showed that Thiel bodies were considered significantly superior to pig models and even more useful than training under proctorship in the operating room. Unfortunately, in most of these studies, only the participants’ feedback was analyzed, and the overall methodological quality was poor. Nevertheless, these studies showed encouraging results in terms of teaching technical skills and educational value. The LLSS has the advantage of being able to assess skills acquisition regardless of the teaching model used. Furthermore, this scale can also be used as an assessment tool in the operating room, on real patients.

Conclusion

An assessment scale in laparoscopic liver surgery, the LLSS, was developed for the first time, using the Delphi methodology. Its validity will be assessed through a dedicated LLR training program.