Secondary Logo

Journal Logo

Original Articles

Training and learning curves in minimally invasive pancreatic surgery: from simulation to mastery

Haney, Caelán M. MDa; Karadza, Emir MDa; Limen, Eldridge F. MDa; de Santibanes, Martin MDb; Kinny-Köster, Benedict MDa; Müller, Philip C. MDc; Bintintan, Vasile V. MDd; Kulu, Yakub MDa; Hackert, Thilo MDa; Müller-Stich, Beat-Peter MDa; Nickel, Felix MDa,∗

Author Information
doi: 10.1097/JP9.0000000000000050
  • Open


Minimally invasive pancreatic surgery

Pancreatic surgery belongs to the most delicate surgical fields with even modern benchmark cases from high-volume centers showing overall complication rates of up to 65.3%, severe postoperative complications (Clavien-Dindo ≥3) of 19% and a readmission rate of 16%.[1] The wide spectrum of malignancies and benign pathologies that require resection poses high technical requirements with surgeons having to adapt to diverse surgical environments.[2] However, morbidity and mortality have decreased significantly in experienced high-volume centers and standardized definitions of outcomes provide a good groundwork for comparing results.[3,4] Modern-day surgeons have been pushing the limits of how resectability of cancer is seen.[5–14]

An important factor that has gained publicity since the beginning of the 21st century is the case volume/outcome relationship in surgery. It has been shown that high-volume hospitals have significantly better short- and long-term outcomes compared to low-volume hospitals.[15–19] As to be expected, this is the case in open pancreatic surgery (OPS) as well as in minimally invasive pancreatic surgery (MIPS). In a recent study on the National Cancer Database, 22,013 patients were analyzed and hospitals that performed more than 6 minimally invasive pancreaticoduodenectomies (MIPDs) every year had significantly better outcomes than hospitals with a volume of less than 6 MIPDs. Interestingly, it has been proposed that the case volume necessary to perform MIPS safely must be even higher than for OPS. The justification for this higher number is that surgeons need the skills to operate both open and minimally invasive cases in order to adequately operate on minimally invasive cases that require intraoperative conversion. One recent study proposed a volume threshold of 22 minimally invasive cases per year as that was when the probability of experiencing a complication stopped decreasing.[20] Similarly, the recently published Miami Guidelines for minimally invasive pancreas resection propose a minimal volume of 20 MIPDs per year to uphold a high quality of surgery.[21]

Apart from the hospital volume, individual surgeon experience and skill has become an increasing focus. A study by Birkmeyer et al[22] that evaluated the skill of minimally invasive bariatric surgeons showed that there was an inverse correlation between both technical skill and annual procedural volume with postoperative complications. Technical skill was hereby measured with the help of an objective structured assessment of technical skills (OSATS) score for minimally invasive surgery (MIS) which has also been used to measure skill in randomized trials in pancreatic surgery.[23] Similarly, a study which compared pancreatic consultant surgeons with more than 10 years of experience with newly appointed consultant surgeons estimated that per 100 patients operated on by the newly appointed consultant surgeons there were 16.6 to 6.5 more postoperative pancreatic fistulae than if these patients had been operated on by the more experienced consultant surgeons.[24]

Both the hospital volume—outcome relationship as well as the individual surgeon's skill and volume—outcome relationship have come under increased scrutiny with the rising awareness that the learning curves of surgeons should not be seen as an excuse for complications experienced by patients.[25] The surgical community has also focused on maximizing safety during implementation of new surgical techniques as these potentially have higher complication rates than established procedures during the initial introductory phases. The IDEAL framework,[26–31] which structures this safe implementation of novel surgical techniques has evolved within the last 15 years and in its recent update the importance of learning curves for new surgical techniques was highlighted.[28]

History of minimally invasive pancreatic surgery

In the same period of time during which the IDEAL guidelines were developed, MIPS has become more popular. In some cases, these guidelines were instinctively or purposely followed. The first reports of laparoscopic pancreatic surgery were published in 1994 by Gagner et al.[32,33] From the beginning onwards, the learning curve was an important topic with the authors of the publication noting the decrease in operating time after several procedures. Since the beginning of the 21st century, laparoscopic pancreatic procedures has started to become more popular across the world with reports of larger cohorts being published from high-volume centers.[34–38] An analysis of the National Cancer Database showed that from 2010 to 2015 17.1% of all registered pancreaticoduodenectomies for pancreatic cancer were performed via minimally invasive access.[39] Overall, there was an increase from 12.2% in 2010 to 21.4% in 2015 of all registered cases being performed minimally invasively and a 73.8% increase of facilities performing MIPD. With the rising popularity of robotic surgery, this number is likely to have risen even more by now. Apart from studies reporting prospective or retrospective cohorts, case-matched studies and registry evaluations, recent years have seen the first randomized mono- and multicentric controlled trials in which laparoscopic pancreaticoduodenectomy (LPD) was compared to open pancreaticoduodenectomy (OPD).[23,40,41] After meta-analysis of these trials no significant differences in complications or postoperative mortality between the 2 surgical approaches were seen.[42] However, it must be stated that one of the 3 randomized controlled trials (RCTs) had to be stopped for safety purposes after interim analysis because of higher mortality in the LPD group compared to OPD. The possibility that this was learning curve associated was also considered by the authors of the expertly devised trial.[23]

With the dissemination of robotic-assisted operating systems especially in the United States, there has been a sharp increase of robotic assisted pancreatic resections. First cohorts were reported in 2010[43–45] and larger programs such as the Pittsburgh Medical Center which obeyed the IDEAL guidelines and adhered to a strict training program followed soon after.[46] It has to be noted, however, that to date no RCTs comparing robotic assisted pancreatic surgery to open or laparoscopic pancreatic surgery have been published and that the literature assessing robotic-assisted surgery (RAS) has been shown to suffer from strong industry bias.[47–49]

Advantages and limitations of minimally invasive surgery

In order to go into detail on training in MIPS, it is important to know the advantages and limitations of the different minimally invasive techniques (Table 1). Both robotic assisted pancreatic surgery and laparoscopic pancreatic surgery were developed with the goal of decreasing tissue trauma to the patient's body by minimizing incisions and therefore reducing wound surface and wound related disadvantages. Hereby, surgeons hope to reduce complications and to minimize the length of hospital stay. However, the 2 techniques have fundamental differences. Conventional laparoscopic surgery can nowadays be seen as a relatively inexpensive alternative to robotic surgery. The overall cost is about the same as in open surgery while training is easily available and can be performed inexpensively. However, the main limitation of laparoscopy is the initial difficulty in acquiring the necessary skills as the lack of 3D vision, the fulcrum and pivoting effect as well as the limited haptic feedback and degrees of motion makes the handling of the instruments more difficult to learn. Furthermore, the oftentimes ergonomically awkward operating positions especially in longer operations make the procedures physically demanding. This aspect gains importance for complex procedures such as MIPD which even in an open setting can have significant operative durations.

Table 1
Table 1:
Advantages and disadvantages of minimally invasive surgery

Some of these limitations of laparoscopic surgery are alleviated in RAS. Due to the 3D vision and the more intuitive handling of the instruments with natural movement and elevated degrees of freedom, learning can potentially come easier and more naturally. In addition, surgeons can supposedly operate in a more ergonomic position, a factor that is especially important when considering the length of operations in pancreatic surgery. However, one of the pitfalls is the total lack of haptic feedback. Especially in the initial learning phase, this can lead to unintended tissue tears due to the application of excessive force by the surgeon. With time, this lack is in part compensated as surgeons develop so-called visual haptics which in part replaces real haptic feedback. In line with this, the basic skills of open and conventional laparoscopic surgery show limited transfer to robotic surgery for basic skills and these have thus to be learned outside the operating room.[50] One further limitation of RAS is the high cost of the procedures due to the prices of the surgical systems and instruments which can only be used a limited amount of times. This also makes training more difficult because even if robotic training instruments with an extended lifespan are available, the costs are still extremely high in comparison to open and conventional laparoscopic surgery.

Training requirements in minimally invasive pancreatic surgery

The main goal of training in MIPS is shortening the learning curve and sparing patients learning curve associated morbidity and mortality associated with the adoption of new techniques. In the recent past, there has been a change in attitude away from Halsted's “See One, Do One, Teach One” and a push toward safer training curricula that can be taken step by step and are comprised of a combination of multiple training methods and modalities.[51] Surgeons should learn the psychomotor and cognitive skills that are necessary to avoid complications outside of the operating room before performing surgery on patients. As in clinical practice complications will still be encountered at some point, the training should also prepare surgeons with how to deal with these complications before they become catastrophes. Guidance from dedicated experts in the form of proctoring and mentoring are key elements to safely deal with the training and learning curve in MIPS.

How to train?

The ideal approach to safely beginning MIPS is nowadays considered to be a stepwise approach accompanied by a validated and structured multimodality training curriculum.[52–54] Especially for surgeons that are not experienced with MIS, acquiring basic skills before attempting advanced procedures such as MIPD is of utmost importance.[55] Surgeons that already have experience in MIS definitely have an advantage here, especially when it comes to laparoscopic surgery. Many of the initial steps can be taken simultaneously and the process can be seen as a pyramid with the basis providing the groundwork for achieving the goal of safely performing MIPS on patients (Fig. 1). In essence, the process can be split up into 3 phases.

Figure 1
Figure 1:
Pyramid of training for minimally invasive pancreatic resections with training phases 1 to 3 and phases of the learning curve (competency, proficiency, mastery).

Phase 1: During the first phase, the surgeon should learn the basics of the operating system, acquire basic skills and master procedure-specific skills in a simulated environment with an emphasis on biotissue drills. The cognitive skills for the operative steps and tactics should be worked on with the use of conventional textbooks, video libraries, procedure cards, surgical training apps. On-site visits to specialized centers are also recommended as well as dedicated operative courses.

Phase 2: In this phase, the surgeon should start with simpler procedures on the preferred minimally invasive system (conventional laparoscopy or robotics) in the anatomical region to gain experience. These so-called index procedures should include cholecystectomies and other upper gastrointestinal and hepatobiliary procedures, while then gradually adopting minimally invasive distal pancreatectomy (MIDP) and moving up to MIPD. Ideally, surgeons should visit and observe an experienced center with a minimally invasive hepatopancreatobiliary program. A fellowship also provides ideal ground to include the different phases. Surgeons should emphasize standardized techniques and benchmark their performance with the help of continued use of video libraries, biotissue drills, and performance measurements. In the absence of a fellowship, the surgeon should then have an experienced proctor for the first cases performed at their own institution to help set up a standardized technique and to avoid early and unnecessary complications.

Phase 3: In this phase, surgeons should start operating by themselves and should be accompanied by constant or intermittent mentoring to maintain a high standardization for different cases. In case of longer pauses in between cases or for optimization of the technique it is again recommended to use video libraries, video analyses with feedback, and biotissue drills regularly. Patient selection is of high importance and surgeons should opt to select simpler cases at the beginning of the learning curve and should only start taking on more complex cases once the initial learning curve has been completed.[56]

Phase 1

Knowledge of the system and operative setups: The initial step that is especially important in RAS is acquiring knowledge of the robotic setup and functioning. The most widespread systems are based on the concept of a master console from where the main surgeon works and the patient cart where the instruments and camera are inserted into the patient. Before operating, the surgeon should be able to control the system without problems and have the necessary knowledge and basic skill to avoid common errors that could occur. In addition, a well trained support team should be assembled that trains with the surgeon and has knowledge of the operating system to ensure proper setup and to avoid malfunctions and collisions. The members of the team should practice communicating with each other as a lot of the communication actually does not take place face to face but rather with the main surgeon at the console and the assistant surgeon and the scrub nurse at the patient cart. Therefore, especially the main surgeon and the assistant surgeon have to practice accurately describing the problems they encounter and how they can help each other without the main surgeon having to leave the console. Most companies that provide robotic operating systems offer basic courses and online courses on how to operate the systems with modules available for operating surgeons, patient side assistants, and scrub nurses.

Similarly, in laparoscopic surgery, surgeons should become familiarized with the different available instruments and laparoscopic techniques. Surgeons should be able to adequately instruct their camera assistant on how to change the angle and focus and should know how other assistants can effectively help them during the procedure. Adequately describing what the assistants should do is very important as this can very much alleviate the operative flow and prevent disruptions of the operation.

Essential for both laparoscopic and RAS is the knowledge of how to adequately position the patient. This is a skill that is extremely important in MIS and can both lead to extreme facilitation of the procedure when done right and if done wrong can produce unnecessary difficulties even to the most skilled surgeons.

Simulation: Virtual simulation and basic skills training: Before actually beginning to operate on patients or performing biotissue drills it is advisable to acquire basic skills using virtual simulators and inanimate drills on box trainers.[55,57–63] There is a wide array of simulators available for both RAS as well as laparoscopic surgery.[64] With these simulators, especially basic skills such as orientation, camera handling, and working with different instruments can be practiced. While many of these simulators have been validated for specific tasks,[65–68] most of the simulators have high acquisition costs. Furthermore, some of the simulators for RAS do not come as stand-alone systems but can only be used as piggyback systems in combination with a robotic console, thereby making their use impossible when the robotic system is used on patients.

Biotissue drills: Biotissue drills are an ideal way to practice procedure specific reconstructive skills. These are the creation of the pancreaticojejunostomy (PJ), hepaticojejunostomy (HJ), and gastro-/duodenojejunostomy (GJ). Especially for the PJ and HJ, biotissue drills have great importance as the porcine models mostly have a different shape of the pancreas and a smaller size of the pancreatic duct and training on human cadavers is difficult due to rapid tissue degradation and autophagy of the pancreatic tissue.[69–71] There are several models commercially available at reasonable prices (Fig. 2) and different training drills have been proposed by notable institutions.[72]

Figure 2
Figure 2:
Biotissue model drill for robotic-assisted and laparoscopic pancreaticojejunostomy and hepaticojejunostomy (Biotissue models: Big Bite Medical, Heidelberg, Germany).

Cadaver training: Training on human cadavers offers several advantages. The anatomy is the most realistic of all available simulations which in turn makes practicing on these models highly effective. Especially the resection phase can be trained on fresh frozen cadavers. However, these models are not universally available and the lack of perfusion in these models makes training for complications nearly impossible. Also, the reconstruction phases of MIPD can often not be trained adequately on the cadavers due to the fast decomposition and autophagia of the pancreas.

In vivo simulation: In vivo simulation is performed mostly on porcine models. Again, the resection phases can be practiced very well but as already mentioned the reconstruction of the pancreatic anastomosis proves difficult due to the different anatomy and size of the porcine pancreas. However, the porcine model offers good opportunity for training of the HJ and GJ. The disadvantage of this sort of training is the ethical aspect that always comes with working on live animal models.

Courses: Though still not very widespread, there are intensive hands-on courses for MIPS that offer a variety of training possibilities. These are organized by different surgical associations specialized on hepatopancreatobiliary surgery and surgical oncology (e.g., E-MIPS:; E-AHBPA:; ESSO: Especially for the beginning of the journey toward MIPS these courses are advisable as they offer valuable training opportunities and insights from dedicated experts and also give a good opportunity to actually view training setups and to learn to accurately use these to surgeons’ and patients’ benefit.

An important question in surgery is to judge when surgeons are ready to step up to the next phase in training, for example, here from the first phase which can be seen as a sort of preparative phase to the next phase in which the actual operations on patients begin. This should in the future be answered based on scientific evidence with the help of objective parameters that can be gathered by evaluating the learning process. Subjective parameters such as the trainees’ own confidence and the confidence their mentors and proctors put in them need also be taken into account in a structured and validated way. Regarding the evaluation of the training, this can be done with the help of validated OSATS scores and other skill evaluation forms as beforementioned. These can be used to evaluate the level of suturing, knot tying and other basic tasks.

An example for a training program that involved training of these basic skills is that proposed by the Pittsburgh group. The program involved a mandatory simulated training with both virtual modules and biotissue drills that the surgeons had to become proficient in. Only after this, surgeons were allowed to take a stepwise robotic-assisted clinical training to perform parts of the procedures under supervision and to gain confidence and skill with the help of mentoring surgeons.[73] After this fellowship program, most of the participating surgeons had completed at least several partial robotic-assisted pancreaticoduodenectomies (RAPDs) and one complete RAPD by themselves.

Phase 2

Index procedures: After the initial basic skills training, and simulations and when confidence has arrived that a successful MIPS program can be supported by surgeons’ institution, it is advisable to begin the implementation with so-called index procedures. These are procedures that take place in the same anatomical operating field but have a lower level of difficulty. For pancreatic surgery, ideal index procedures can be cholecystectomies, hepatobiliary, and upper gastrointestinal procedures including bariatric operations. With the help of these procedures, the surgeons can become acquainted with the operating field and required technical skills. These can be especially important as the reconstructive phases can be in part trained during these procedures as well and surgeons that are not very experienced in MIS can get used to the different operating circumstances.

On site viewing and proctoring: Partnerships with experienced centers can be very helpful for setting up an on-site viewing at a center that has expertise in MIPS. By doing this, experienced surgeons can convey their knowledge to the learning surgeons in live cases. After visiting the experienced institution, a proctoring program should be set up. The experienced surgeons then visit the new surgeons’ institution to help perform the initial cases, which facilitates the process of adopting this new technique. Ideally, the inexperienced surgeon should start their operating program with MIDP in order to gain more experience in the different approaches to the pancreas and the different operating circumstances. Once MIDP can be performed with high confidence, MIPD can be attempted, again under the guidance of a proctor and after visiting experienced institutions that perform the procedure routinely. Prior performance of index procedures that also involve extensive reconstruction requiring suturing and knot tying will aid to transition to MIPD. Hybrid procedures with reconstruction via minilaparotomy are an alternative route in the transition to minimally invasive pancreatoduodenectomy.[74–76] Further examples for training programs for MIPS that involved successful proctoring are the LAELAPS and LAELAPS-2 studies. These were nationwide programs performed to introduce laparoscopic distal pancreatectomy and pancreaticoduodenectomy (PD) in the Netherlands. Tailored toward already more experienced surgeons, they involved intensive on-site and off-site proctoring by international and national experts as well as mandatory video training.[77,78] After the training program for LDP the nationwide percentage of LDP increased rapidly. The follow-up training program LAELAPS-3 that aims at training surgeons for RAPD also involves virtual simulation and biotissue drills in addition to proctoring and video training but has not yet reached publication at this time.

Fellowships: Fellowships in institutions with high case volumes in MIPS can offer a great opportunity to observe and perform more extensive numbers of procedures and thereby acquire a higher level of experience and skill. As MIPS is still not widespread, there is a limited number of these institutions and the fellowships are coveted but with the further spread of MIPS this is set to change.

Phase 3

In the last phase, the surgeon should have enough confidence to perform the viable cases minimally invasive and independently. However, even after proctoring and mentoring, a learning curve can be expected. Surgeons should gradually ease into performing cases in a minimally invasive fashion. Patient selection is of importance during the learning curve in both open and MIPS. Surgeons should feel confident in themselves and should also consult their senior colleagues, mentors, and proctors on whether they are ready to perform parts of the procedures or full procedures first under supervision and then by themselves. And even when already performing by themselves comfortably, they should eventually solicit the help of their senior surgeons, proctors, and mentors. This can be done by video analysis or on-site. Learning curves and sufficient experience need to be taken into account as well when attempting more complex cases. Therefore, again a stepwise approach should be taken with careful case selection with initially choosing patients with lower risk profiles for the new technique in which the surgeon is confident that they can perform the operation without problems due to the newly adopted technique. These initial cases should exclude patients with high BMI, comorbidities that are unfavorable for prolonged operative times and could therefore complicate the procedure and patients that have undergone extensive previous operations and therefore might not grant access via endoscopic route due to adhesions. Patients requiring operation due to chronic pancreatitis represent good initial patients as oncologic guidelines do not have to be adhered to and the PJ usually has lower risk of fistula. Going even further, cases with vascular involvement have been described to be performed in a minimally invasive fashion; however, this can be seen as mastery and should only be attempted when the surgeon has reached a high volume of previous standard cases while gradually taking on more complex cases.[56] To keep up a high level of skill, in times in which the surgeon does not operate many cases in a minimally invasive fashion, biotissue drills as described above should be performed regularly. Furthermore, it is advisable to build a video library of performed cases and review these regularly in order to improve skill and to adequately learn from already performed cases.[79] In addition to keeping up and improving the surgeon's skill, this library can help mentor surgeons at the same institution and develop their technique and can be used for exchange with other surgeons. This is nowadays facilitated by means of dedicated online platforms but also social media.

The whole process from biotissue drills to independently performing cases without proctoring should not be rushed. By maximizing the training before operating on the first patients these can be spared complications that are associated with the learning curve. Furthermore, most surgeons would argue that before venturing into performing MIPS, there should be sufficient experience in OPS. This will prove essential when surgeries have to be converted from MIPS to an open procedure.

An example for a training program that has progressed into far into the third phase is again that of the Pittsburgh group. They have now performed more than 500 RAPD in a single center and over time the number of vascular resections and generally more difficult cases has highly increased without worsening the number of serious complications or the duration of the operation.[80]

Learning curves

Even with extensive simulation training, index procedures and proctoring, as with all types of surgery, a certain learning curve in the operating room still seems inevitable. Multiple measures for the learning curve in MIPD have been proposed with the certain levels being reached at 10 to more than 100 cases (Table 2). Giving a specific number of cases after which the learning curve has been passed is difficult, as the learning curve is dependent of many different variables and definitions. These include:

  • Previous experience in OPS and other index procedures
  • Previous experience in MIS
  • Availability and use of tutors, proctors and mentors
  • Hospital procedural volume as well as individual procedural volume
  • Time between cases
  • Individual surgeon talent and skill.

Table 2
Table 2:
Learning curves in minimally invasive pancreaticoduodenectomy

Therefore, when interpreting the learning curves given in the literature, it is always important to examine the circumstances in which these observations were made, as some surgeons described already had extensive experience in MIS and pancreatic surgery, while other surgeons had less experience in MIS. Furthermore, many observations published in the literature were made at the beginning of the development of MIPS. These observations describe the evolution and establishment of an entirely new surgical technique. Consequently, these learning curves actually do not reflect the learning curve of an already evolved technique that is being acquired in established training setting rather than being the work of pioneers.

In addition to the learning curves being dependent of previous surgeon experience, to date there has been little standardization of how to describe the learning curve. Technical aspects such as operative time are aspects that are observed most frequently as surrogate parameters of surgical quality. Operative time itself has multiple limitations in the interpretation as it is influenced by a multitude of surgeon, team, and patient-related factors. On the other hand, it is universally available and easy to use and while it might be of limited value for an individual case it certainly has its value when used to observe and compare learning curves over higher case numbers. Factors with which the development of surgical proficiency can be described reach from intraoperative safety measures such as blood loss or intraoperative near-misses and complications[81] to postoperative parameters including length of stay and patient safety measures such as perioperative complications. Postoperative pancreatic fistulae and other complications can be seen as a technical outcome as well as they have been shown to greatly depend on technical skill and learning curves of surgeons and teams.[82] For oncologic surgery the markers of radicality including resection margins and lymph node yield can also be taken into account for the learning curve.

Hence, these observations can currently only be taken as rough estimates of how long learning curves will be for an individual surgeon and will differ according to the abovementioned factors. Nonetheless, it is worth examining the differing learning curves described in the literature. Hereby, we limit ourselves to analyzing the learning curve of MIPD. Learning curves of MIDP are generally shorter since the procedure is less complex and technically demanding.

The lowest observed change was reported by Speicher et al.[83] The authors located at a high-volume center examined 56 LPD cases and showed a significant reduction of the operating time after 10 cases. After 50 cases they were also able to observe a significant reduction of blood loss. In a multicentric study which gathered cases 1029 cases of LPD, the Chinese authors were able to see a primary proficiency level after the first 50 cases and a secondary mastery level after 104 cases.[84] In RAPD, Boone et al[85] analyzed 200 cases performed from 2008 to 2014 performed by 4 main surgeons and presented a statistical improvement of blood loss and conversions after 20 cases, a reduction of postoperative pancreatic fistulae after 40 cases and a plateau in operative time after 80 cases. However, most likely the learning curve goes on for even longer. In a paper published in the Annals of Surgery, Shi et al[86] noticed changes in outcome even after 200 cases of RAPD and in their analysis of 500 RAPD Zureikat et al[80] noticed a plateau of operating room time after 240 cases. These 2 studies represent the largest reports of single institution RAPD cohorts up to date. Shi et al report 450 cases of RAPD performed by 3 surgeons and Zureikat et al report 500 cases performed by 5 surgeons. This only highlights the importance of consistent monitoring of outcomes and the possibility for constant improvement in this technically demanding procedure.

The comparison of the learning curves of MIPS with OPS is difficult due to the lack of sufficient data. This is because OPS is not a newly developing surgical technique and is therefore not subject to close scrutiny. Schmidt et al[87] and Tseng et al[88] both analyzed operative data of single centers that performed open PD. Tseng et al came to the result that blood loss, operative time, and length of stay declined after 60 cases. In their analysis of 1003 patients that underwent open PD, Schmidt et al predicted that beginning pancreatic surgeons would reach equivalent morbidity and mortality as experienced pancreatic surgeons after 20 and 60 cases. However, both of these studies represent data of high-volume expert centers. Therefore, the data have to be interpreted with caution as in low-volume centers the learning curve might well be longer.

Logically the later phases of the learning curve after competency can be seen as the phases where more difficult cases start to be performed (proficiency) and can be done with consistent and optimal outcome (mastery). This progress from easier to more complex cases can be seen in other specialties as well was described in a recent systematic review by Wehrtmann et al[56] who analyzed the learning curves in bariatric surgery and divided the learning curve into the 3 phases “competency,” “proficiency,” and “mastery.” The authors of the paper suggest a standardization of learning curve measurements by including a CUSUM analysis and predetermined outcomes in order to make comparing learning curves easier. This standardization would make the interpretation of single surgeon results feasible and allow adequate meta-analysis of learning curves, thereby also enabling a comparison between different procedures that can be used for the same illness such as open, laparoscopic and RAPD.

Furthermore, the outcome is very dependent on the hospital's case volume. In the recent randomized controlled LEOPARD-2 trial performed by the Dutch Pancreatic Cancer Group the inclusion criteria for surgeons participating in the trial was the completion of a dedicated training program which included simulation, biotissue drills, case observations, and proctoring.[78,89] In addition, a minimum of 20 previous LPDs and a hospital case volume of at least 10 LPDs and overall at least 20 PDs per year were required. However, after randomization the case volume of LPDs halved and the observed outcome was not as good as when observed during the training program. Because of safety concerns in the LPD group the trial was stopped prematurely since there was no benefit of LPD at the same time over open PD. In their publication, the authors discuss the halving of the hospital volume as one possible factor that might have negatively influenced surgical performance. This trial highlights the importance of only performing MIPS in hospitals that have a sufficiently high volume to upkeep a steady number of MIPS. The recently published Miami guidelines for MIPS advise that mortality in centers that perform less than 10 MIPD is higher and complications are higher in centers that perform less than 20 MIPD.[21] These case volumes should be seen as the bare minimum and higher case volumes should be aimed for. While surgeons nowadays are not obliged to follow these recommendations, the recent past has already seen a trend toward centralization[90] and the future will likely see a development away from the currently self-supervised decentralized surgical training and treatment and toward more structured certification programs in high-volume centers. To some part these certification programs already rely on hospital case volumes and this is likely to play a role in the future as well. However, it is likely that individual surgeon assessment and certification will play a more important role in the future in many countries. However, it remains to be seen which interested parties such as lawmakers, insurance companies, hospitals, patient advocacy groups, and surgical societies will be the driving forces behind these certification programs that must aim to improve patient safety. These certification programs will likely include mandatory training programs that will have to be based on valid training evidence which unfortunately is still scarce. Therefore, an effort should be made to develop valid, evidence-based training programs, and validated skills assessment scores that include the aforementioned steps. In addition, criteria should be developed on when trainees have reached a sufficient level of expertise for them to be able to advance from one phase to the next. It is our strong belief that surgeons must play the central role in developing and applying these training and certification guidelines and that we must provide the necessary scientific evidence.


MIPS training should be implemented with the goal of ensuring optimal introduction of the technique while minimizing the learning curve associated morbidity and mortality. This includes ensuring patient safety and adequate outcomes at all phases of the introduction of MIPS to a hospital or individual surgeon. This can be achieved by a stepwise approach that begins by building a strong foundation on basic minimally invasive techniques, biotissue drills, virtual as well as in vivo and ex vivo training. After that, dedicated proctoring and mentoring programs should be set up and fellowships can provide surgeons with an adequate case volume. Even after proctoring and fellowships learning curves are to be expected and careful patient selection is of utmost importance. Training in teams, video assessments, as well as exchange and feedback with dedicated experts can help further improve minimally invasive skills.[63,91] A prerequisite for safe MIPS implementation and to ensure optimal outcomes is a high hospital volume as well as individual surgeon procedural volume.



Author contributions


Financial support


Conflicts of interest

The authors declare no conflicts of interest.


[1]. Sanchez-Velazquez P, Muller X, Malleo G, et al. Benchmarks in pancreatic surgery: a novel tool for unbiased outcome comparisons. Ann Surg 2019;270:211–218.
[2]. Huang L, Jansen L, Balavarca Y, et al. Resection of pancreatic cancer in Europe and USA: an international large-scale study highlighting large variations. Gut 2019;68:130–139.
[3]. Mollberg N, Rahbari NN, Koch M, et al. Arterial resection during pancreatectomy for pancreatic cancer: a systematic review and meta-analysis. Ann Surg 2011;254:882–893.
[4]. Hackert T, Schneider L, Buchler MW. Current state of vascular resections in pancreatic cancer surgery. Gastroenterol Res Pract 2015;2015:120207.
[5]. Bockhorn M, Uzunoglu FG, Adham M, et al. International Study Group of Pancreatic Surgery. Borderline resectable pancreatic cancer: a consensus statement by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 2014;155:977–988.
[6]. Strobel O, Buchler MW. ISGPS. Borderline resectable pancreatic cancer: ISGPS consensus statement. Chirurg 2014;85:1014.
[7]. Welsch T, Eisele H, Zschabitz S, et al. Critical appraisal of the International Study Group of Pancreatic Surgery (ISGPS) consensus definition of postoperative hemorrhage after pancreatoduodenectomy. Langenbecks Arch Surg 2011;396:783–791.
[8]. Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery 2017;161:584–591.
[9]. Hartwig W, Vollmer CM, Fingerhut A, et al. International Study Group for Pancreatic Surgery. Extended pancreatectomy in pancreatic ductal adenocarcinoma: definition and consensus of the International Study Group for Pancreatic Surgery (ISGPS). Surgery 2014;156:1–14.
[10]. Shrikhande SV, Sivasanker M, Vollmer CM, et al. Pancreatic anastomosis after pancreatoduodenectomy: a position statement by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 2017;161:1221–1234.
[11]. Shukla PJ, Barreto SG, Fingerhut A, et al. Toward improving uniformity and standardization in the reporting of pancreatic anastomoses: a new classification system by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 2010;147:144–153.
[12]. Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 2007;142:761–768.
[13]. Wente MN, Veit JA, Bassi C, et al. Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery 2007;142:20–25.
[14]. Arezzo A, Vettoretto N, Francis NK, et al. The use of 3D laparoscopic imaging systems in surgery: EAES consensus development conference 2018. Surg Endosc 2019;33:3251–3274.
[15]. Birkmeyer JD. Relation of surgical volume to outcome. Ann Surg 2000;232:724–725.
[16]. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med 2002;346:1128–1137.
[17]. Birkmeyer JD, Warshaw AL, Finlayson SR, et al. Relationship between hospital volume and late survival after pancreaticoduodenectomy. Surgery 1999;126:178–183.
[18]. Finlayson EV, Goodney PP, Birkmeyer JD. Hospital volume and operative mortality in cancer surgery: a national study. Arch Surg 2003;138:721–725. discussion 726.
[19]. Goodney PP, Stukel TA, Lucas FL, et al. Hospital volume, length of stay, and readmission rates in high-risk surgery. Ann Surg 2003;238:161–167.
[20]. Adam MA, Thomas S, Youngwirth L, et al. Defining a hospital volume threshold for minimally invasive pancreaticoduodenectomy in the United States. JAMA Surg 2017;152:336–342.
[21]. Asbun HJ, Moekotte AL, Vissers FL, et al. International Study Group on Minimally Invasive Pancreas Surgery (I-MIPS). The Miami international evidence-based guidelines on minimally invasive pancreas resection. Ann Surg 2020;271:1–14.
[22]. Birkmeyer JD, Finks JF, O’Reilly A, et al. Michigan Bariatric Surgery Collaborative. Surgical skill and complication rates after bariatric surgery. N Engl J Med 2013;369:1434–1442.
[23]. van Hilst J, de Rooij T, Bosscha K, et al. Dutch Pancreatic Cancer Group. Laparoscopic versus open pancreatoduodenectomy for pancreatic or periampullary tumours (LEOPARD-2): a multicentre, patient-blinded, randomised controlled phase 2/3 trial. Lancet Gastroenterol Hepatol 2019;4:199–207.
[24]. Roberts KJ, Boteon A, Marcon F, et al. Risk adjusted assessment of individual surgeon's pancreatic fistula outcomes. HPB (Oxford) 2019;22:452–460.
[25]. Altman LK. The doctor's world. When Patient's Life is Price of Learning New Kind of Surgery New York: The New York Times; 1992. [Accessed November 12, 2019]
[26]. Blazeby JM, Blencowe NS, Titcomb DR, et al. Demonstration of the IDEAL recommendations for evaluating and reporting surgical innovation in minimally invasive oesophagectomy. Br J Surg 2011;98:544–551.
[27]. Dimick J.B., Sedrakyan A., McCulloch P.. The IDEAL Framework for Evaluating Surgical Innovation: How It Can Be Used to Improve the Quality of Evidence. JAMA Surg 2019;154:685–686.
[28]. Hirst A, Philippou Y, Blazeby J, et al. No surgical innovation without evaluation: evolution and further development of the ideal framework and recommendations. Ann Surg 2019;269:211–220.
[29]. McCulloch P, Cook JA, Altman DG, et al. IDEAL framework for surgical innovation 1: the idea and development stages. BMJ 2013;346:f3012.
[30]. McCulloch P, Feinberg J, Philippou Y, et al. Progress in clinical research in surgery and IDEAL. Lancet 2018;392:88–94.
[31]. Pennell CP, Hirst AD, Campbell WB, et al. Practical guide to the idea, development and exploration stages of the ideal framework and recommendations. Br J Surg 2016;103:607–615.
[32]. Gagner M, Pomp A. Laparoscopic pancreatic resection: is it worthwhile? J Gastrointest Surg 1997;1:20–25. discussion 25–26.
[33]. Gagner M, Pomp A. Laparoscopic pylorus-preserving pancreatoduodenectomy. Surg Endosc 1994;8:408–410.
[34]. Gagner M, Palermo M. Laparoscopic Whipple procedure: review of the literature. J Hepatobiliary Pancreat Surg 2009;16:726–730.
[35]. Palanivelu C, Rajan PS, Rangarajan M, et al. Evolution in techniques of laparoscopic pancreaticoduodenectomy: a decade long experience from a tertiary center. J Hepatobiliary Pancreat Surg 2009;16:731–740.
[36]. Palanivelu C, Shetty R, Jani K, et al. Laparoscopic lateral pancreaticojejunostomy: a new remedy for an old ailment. Surg Endosc 2006;20:458–461.
[37]. Palanivelu C, Shetty R, Jani K, et al. Laparoscopic distal pancreatectomy: results of a prospective non-randomized study from a tertiary center. Surg Endosc 2007;21:373–377.
[38]. Senthilnathan P, Srivatsan Gurumurthy S, Gul SI, et al. Long-term results of laparoscopic pancreaticoduodenectomy for pancreatic and periampullary cancer-experience of 130 cases from a tertiary-care center in South India. J Laparoendosc Adv Surg Tech A 2015;25:295–300.
[39]. Torphy RJ, Friedman C, Halpern A, et al. Comparing short-term and oncologic outcomes of minimally invasive versus open pancreaticoduodenectomy across low and high volume centers. Ann Surg 2019;270:1147–1155.
[40]. Palanivelu C, Senthilnathan P, Sabnis SC, et al. Randomized clinical trial of laparoscopic versus open pancreatoduodenectomy for periampullary tumours. Br J Surg 2017;104:1443–1450.
[41]. Poves I, Burdío F, Morato O, et al. Comparison of perioperative outcomes between laparoscopic and open approach for pancreatoduodenectomy: the PADULAP randomized controlled trial. Ann Surg 2018;268:731–739.
[42]. Nickel F, Haney CM, Kowalewski KF, et al. Laparoscopic versus open pancreaticoduodenectomy: a systematic review and meta-analysis of randomized controlled trials. Ann Surg 2020;271:54–66.
[43]. Buchs NC, Addeo P, Bianco FM, et al. Outcomes of robot-assisted pancreaticoduodenectomy in patients older than 70 years: a comparative study. World J Surg 2010;34:2109–2114.
[44]. Horiguchi A, Uyama I, Miyakawa S. Robot-assisted laparoscopic pancreaticoduodenectomy. J Hepatobiliary Pancreat Sci 2011;18:287–291.
[45]. MacKenzie S, Kosari K, Sielaff T, et al. The robotic Whipple: operative strategy and technical considerations. J Robot Surg 2011;5:3–9.
[46]. Zureikat AH, Nguyen KT, Bartlett DL, et al. Robotic-assisted major pancreatic resection and reconstruction. Arch Surg 2011;146:256–261.
[47]. Criss CN, MacEachern MP, Matusko N, et al. The impact of corporate payments on robotic surgery research: a systematic review. Ann Surg 2019;269:389–396.
[48]. Patel SV, Yu D, Elsolh B, et al. Assessment of conflicts of interest in robotic surgical studies: validating author's declarations with the open payments database. Ann Surg 2018;268:86–92.
[49]. Probst P, Knebel P, Grummich K, et al. Industry bias in randomized controlled trials in general and abdominal surgery: an empirical study. Ann Surg 2016;264:87–92.
[50]. Kowalewski KF, Schmidt MW, Proctor T, et al. Skills in minimally invasive and open surgery show limited transferability to robotic surgery: results from a prospective study. Surg Endosc 2018;32:1656–1667.
[51]. Hogg ME, Besselink MG, Clavien PA, et al. Training in minimally invasive pancreatic resections: a paradigm shift away from "See one, Do one, Teach one”. HPB (Oxford) 2017;19:234–245.
[52]. Tam V, Borrebach J, Dunn SA, et al. Proficiency-based training and credentialing can improve patient outcomes and decrease cost to a hospital system. Am J Surg 2019;217:591–596.
[53]. Knab LM, Hogg ME. ASO author reflections: novel robotic training curriculum used in complex surgical oncology. Ann Surg Oncol 2018;25 (suppl 3):616–617.
[54]. Mark Knab L, Zenati MS, Khodakov A, et al. Evolution of a novel robotic training curriculum in a complex general surgical oncology fellowship. Ann Surg Oncol 2018;25:3445–3452.
[55]. Kowalewski KF, Garrow CR, Proctor T, et al. LapTrain: multi-modality training curriculum for laparoscopic cholecystectomy—results of a randomized controlled trial. Surg Endosc 2018;32:3830–3838.
[56]. Wehrtmann FS, de la Garza JR, Kowalewski KF, et al. Learning curves of laparoscopic Roux-en-Y gastric bypass and sleeve gastrectomy in bariatric surgery: a systematic review and introduction of a standardization. Obes Surg 2019;30:640–656.
[57]. Nickel F, Hendrie JD, Stock C, et al. Direct observation versus endoscopic video recording-based rating with the objective structured assessment of technical skills for training of laparoscopic cholecystectomy. Eur Surg Res 2016;57:1–9.
[58]. Romero P, Brands O, Nickel F, et al. Intracorporal suturing—driving license necessary? J Pediatr Surg 2014;49:1138–1141.
[59]. Romero P, Gunther P, Kowalewski KF, et al. Halsted's “See One, Do One, and Teach One” versus Peyton's Four-Step Approach: a randomized trial for training of laparoscopic suturing and knot tying. J Surg Educ 2018;75:510–515.
[60]. De La Garza JR, Schmidt MW, Kowalewski KF, et al. Does rating with a checklist improve the effect of E-learning for cognitive and practical skills in bariatric surgery? A rater-blinded, randomized-controlled trial. Surg Endosc 2019;33:1532–1543.
[61]. Kenngott HG, Wunscher JJ, Wagner M, et al. OpenHELP (Heidelberg laparoscopy phantom): development of an open-source surgical evaluation and training tool. Surg Endosc 2015;29:3338–3347.
[62]. Kowalewski KF, Hendrie JD, Schmidt MW, et al. Development and validation of a sensor- and expert model-based training system for laparoscopic surgery: the iSurgeon. Surg Endosc 2017;31:2155–2165.
[63]. Kowalewski KF, Minassian A, Hendrie JD, et al. One or two trainees per workplace for laparoscopic surgery training courses: results from a randomized controlled trial. Surg Endosc 2019;33:1523–1531.
[64]. Schmidt MW, Kowalewski KF, Schmidt ML, et al. The Heidelberg VR Score: development and validation of a composite score for laparoscopic virtual reality training. Surg Endosc 2019;33:2093–2103.
[65]. Abboudi H, Khan MS, Aboumarzouk O, et al. Current status of validation for robotic surgery simulators—a systematic review. BJU Int 2013;111:194–205.
[66]. Autorino R, Haber GP, Stein RJ, et al. Laparoscopic training in urology: critical analysis of current evidence. J Endourol 2010;24:1377–1390.
[67]. Aydin A, Shafi AM, Shamim Khan M, et al. Current status of simulation and training models in urological surgery: a systematic review. J Urol 2016;196:312–320.
[68]. Fisher RA, Dasgupta P, Mottrie A, et al. An over-view of robot assisted surgery curricula and the status of their validation. Int J Surg 2015;13:115–123.
[69]. Zwart MJW, Foppen M, Hilst JV, et al. How to teach and train laparoscopic pancreatoduodenectomy. Ann Pancreat Cancer 2019;2:
[70]. Muller PC, Steinemann DC, Chinczewski L, et al. Transpapillary endopancreatic surgery: decompression of duct system and comparison of greenlight laser with monopolar electrosurgical device in ex vivo and in vivo animal models. Surg Endosc 2018;32:3393–3400.
[71]. Muller PC, Steinemann DC, Nickel F, et al. Transduodenal-transpapillary endopancreatic surgery with a rigid resectoscope: experiments on ex vivo, in vivo animal models and human cadavers. Surg Endosc 2017;31:4131–4135.
[72]. Tam V, Zenati M, Novak S, et al. Robotic pancreatoduodenectomy biotissue curriculum has validity and improves technical performance for surgical oncology fellows. J Surg Educ 2017;74:1057–1065.
[73]. Hogg ME, Tam V, Zenati M, et al. Mastery-based virtual reality robotic simulation curriculum: the first step toward operative robotic proficiency. J Surg Educ 2017;74:477–485.
[74]. Deichmann S, Bolm LR, Honselmann KC, et al. Perioperative and long-term oncological results of minimally invasive pancreatoduodenectomy as hybrid technique—a matched pair analysis of 120 cases. Zentralbl Chir 2018;143:155–161.
[75]. Keck T, Kuesters S, Wellner U, et al. Laparoscopic pylorus-preserving pancreatic head resection and hybrid open reconstruction via pancreatogastrostomy. J Gastrointest Surg 2011;15:373–377.
[76]. Wellner UF, Kusters S, Sick O, et al. Hybrid laparoscopic versus open pylorus-preserving pancreatoduodenectomy: retrospective matched case comparison in 80 patients. Langenbecks Arch Surg 2014;399:849–856.
[77]. de Rooij T, van Hilst J, Boerma D, et al. Impact of a nationwide training program in minimally invasive distal pancreatectomy (LAELAPS). Ann Surg 2016;264:754–762.
[78]. de Rooij T, van Hilst J, Topal B, et al. Dutch Pancreatic Cancer Group. Outcomes of a multicenter training program in laparoscopic pancreatoduodenectomy (LAELAPS-2). Ann Surg 2019;269:344–350.
[79]. Al Abbas AI, Jung JP, Rice MK, et al. Methodology for developing an educational and research video library in minimally invasive surgery. J Surg Educ 2019;76:745–755.
[80]. Zureikat AH, Beane JD, Zenati MS, et al. 500 Minimally invasive robotic pancreatoduodenectomies: one decade of optimizing performance. Ann Surg 2019;[Epub ahead of print].
[81]. Curtis NJ, Dennison G, Brown CSB, et al. Clinical evaluation of intraoperative near misses in laparoscopic rectal cancer surgery. Ann Surg 2019;[Epub ahead of print].
[82]. Hogg ME, Zenati M, Novak S, et al. Grading of surgeon technical performance predicts postoperative pancreatic fistula for pancreaticoduodenectomy independent of patient-related variables. Ann Surg 2016;264:482–491.
[83]. Speicher PJ, Nussbaum DP, White RR, et al. Defining the learning curve for team-based laparoscopic pancreaticoduodenectomy. Ann Surg Oncol 2014;21:4014–4019.
[84]. Wang M, Peng B, Liu J, et al. Practice patterns and perioperative outcomes of laparoscopic pancreaticoduodenectomy in china: a retrospective multicenter analysis of 1029 patients. Ann Surg 2019;[Epub ahead of print].
[85]. Boone BA, Zenati M, Hogg ME, et al. Assessment of quality outcomes for robotic pancreaticoduodenectomy: identification of the learning curve. JAMA Surg 2015;150:416–422.
[86]. Shi Y, Wang W, Qiu W, et al. Learning curve from 450 cases of robot-assisted pancreaticoduocectomy in a high-volume pancreatic center: optimization of operative procedure and a retrospective study. Ann Surg 2019;[Epub ahead of print].
[87]. Schmidt CM, Turrini O, Parikh P, et al. Effect of hospital volume, surgeon experience, and surgeon volume on patient outcomes after pancreaticoduodenectomy: a single-institution experience. Arch Surg 2010;145:634–640.
[88]. Tseng JF, Pisters PW, Lee JE, et al. The learning curve in pancreatic surgery. Surgery 2007;141:694–701.
[89]. de Rooij T, van Hilst J, Bosscha K, et al. Minimally invasive versus open pancreatoduodenectomy (LEOPARD-2): study protocol for a randomized controlled trial. Trials 2018;19:1.
[90]. Gooiker GA, Lemmens VE, Besselink MG, et al. Impact of centralization of pancreatic cancer surgery on resection rates and survival. Br J Surg 2014;101:1000–1005.
[91]. Nickel F, Jede F, Minassian A, et al. One or two trainees per workplace in a structured multimodality training curriculum for laparoscopic surgery? Study protocol for a randomized controlled trial—DRKS00004675. Trials 2014;15:137.

Laparoscopic surgery, Minimally invasive surgery, Robotic surgery, Training

Copyright © 2020 The Chinese Medical Association, Published by Wolters Kluwer Health, Inc. under the CCBY-NC-ND license.