Training in endocrine surgery

A total of 24 national delegates (86% of 28) of the DES responded to the questionnaire on the training of surgical residents. Only half of the responders considered that 20 or more endocrine surgical procedures operations should be mandatory (Table 2). At the low end of this spectrum, one respondent indicated that 5 or less performed procedures would be acceptable.

A final written exam followed by an oral exam at the end of surgical residency is mandatory in 10 European countries. An oral exam only is mandatory in another 10 European countries whereas no mandatory exam exists in four European countries. A practical exam (e.g., OSCE) is organized in 12 European countries and a three-stage exam (written, oral, and practical) is mandatory in six European countries.

Sixteen national delegates (57% of 28) replied to the questionnaire on the training of fellows in endocrine surgery. Only three respondents considered that 200 or more endocrine surgical procedures should be performed during fellowship training in endocrine surgery, whereas 3 respondents considered 100–199 performed procedures to be sufficient. Three respondents agreed that 50–99 performed procedures would be a minimum (Table 2). Only 1 respondent indicated that less than 50 procedures performed would be acceptable, whereas 6 respondents skipped to answer this question. A written exam followed by an oral exam at the end of the fellowship in endocrine surgery is mandatory in three European countries, and an oral exam only is mandatory in one European country, whereas no obligatory exam exists in 12 European countries. A practical exam (e.g., OSCE) is organized in only one European country and it is a part of a three-stage exam (written, oral, and practical).

Programs for surgical residents contained thyroid procedures in the majority (n = 22) but not all European countries (n = 2). The minimum number of total/near-total thyroidectomies to be performed/assisted during residency varied between 0 and 40 (median 0.5)/0 and 50 (median 0) and these values were far below the values expected to be sufficient from the national delegates’ perspective: 0–50 (median 10) and 0–50 (median 20), respectively. The minimum number of hemithyroidectomies (or thyroid resections) to be performed/assisted during residency ranged from 0 to 50 (median 0)/0 to 50 (median 0) and these values were also below the values expected to be required from the national delegates’ perspective: 0–50 (median 15) and 0–50 (median 20), respectively. The minimum number of central lymph nodes dissections to be performed/assisted varied from 0 to 20 (median 0) and 0 to 20 (median 0), respectively. The national delegates’ expectations were slightly higher in this regard: 0–20 (median 5) and 0–50 (median 10), respectively. The minimum number of lateral lymph nodes dissections to be performed/assisted during surgical residency ranged from 0 to 10 (median 0) and 0 to 20 (median 0), respectively. The national delegates’ expectations were relatively consistent with the current minimum requirement for performed operations with 0–10 (median 2) but much higher for assisted lateral neck dissections with 0–20 (median 5).

With regard to fellows (Fig. 1), our survey revealed that the programs for fellowships in endocrine surgery contained thyroid procedures in all European countries which offer some form of fellowship in this area. The minimum numbers of total/near-total thyroidectomies to be performed/assisted during fellowship varied from 20 to 30 (median 30)/0 to 100 (median 30) and these values were somewhat below the values expected to be optimal from the national delegates’ perspective with 10–60 (median 30) and 20–100 (median 50), respectively. The minimum numbers of hemithyroidectomies (or thyroid resections) to be performed/assisted during fellowship varied from 20 to 40 (median 30)/10 to 50 (median 25) and theses values were also below the values expected to be optimal from the national delegates’ perspective with 20–60 (median 40) and 20–100 (median 50), respectively. The minimum numbers of central lymph node dissections to be performed/assisted varied from 5 to 20 (median 10)/10 to 30 (median 20), and the national delegates’ expectations were very similar in this regard with 5–30 (median 10) and 10–30 (median 20), respectively. The minimum numbers of lateral lymph nodes dissections to be performed/assisted during fellowship varied from 10 to 30 (median 20)/5 to 50 (median 10), and national delegates’ expectations were very similar for both performed operations with 5–30 (median 15) and assisted lateral neck dissections with 5–50 (median 15).

The current training of general surgical residents with regard to thyroid and other types of endocrine operations is highly variable, which may contribute to increased complication rates and number of reoperations [1]. The mean minimum number of thyroid operations to be done during surgical residency as shown by this survey was 5 performed and 13 assisted procedures, and these values were two-fold lower than what was recommended by the national delegates on average. This issue raises the question if a general surgeon with no special interest in thyroid surgery has received enough exposure during residency to be able to perform a total thyroidectomy safely.

Total thyroidectomy is an operation that always engenders controversy related to injury of the recurrent laryngeal nerves and the parathyroid glands. It has been shown that surgeons who have completed a well-designed training program and who have become proficient in total thyroidectomy as trainees will remain proficient despite practicing in a provincial center [2]. Achieving a low morbidity rate demands meticulous attention to operative technique and anatomical details [2].

Interesting insight into this issue was recently provided by Phitayacorn et al. who published an expert consensus of general surgery residents’ proficiency concerning common endocrine operations. In this study, members of the American Association of Endocrine Surgeons (AAES) were surveyed about their opinions on resident proficiency with regard to common endocrine operations [3]. A total of 92% of the respondents operated with residents. On average, they believed that most of the steps of a total thyroidectomy for benign disease could be performed by a postgraduate year 4 surgery resident. Specific steps that were considered to require more training included decisions to divide the strap muscles or leaving a drain. Approximately 66% of the respondents felt that a postgraduate final-year surgery resident could independently perform a total thyroidectomy for benign disease but only 45% felt similarly for malignant thyroid disease. The annual endocrine volume of the respondents did not correlate with the beliefs concerning the residents’ autonomy. There was general agreement that a postgraduate final-year resident may not be proficient in advanced endocrine operations. Nevertheless, it was also felt that opportunities exist to improve the training for graduates that anticipate performing endocrine operations routinely [3, 4]. For more data on pertinent training issues on thyroid surgery see Table 3.

The literature on the training of neck dissection is sparse. According to one report, surgical residents perform on average less than 2 neck dissections. Fellows in endocrine surgery perform more neck dissections but the numbers of such procedures vary between about 6 and 15 [9, 13]. According to one report, 12 neck dissections are considered appropriate to achieve the required competence [13].

Only a minority of the European surgical residency programs include parathyroid procedures during the training of graduating residents. The numbers of procedures assisted/performed varied widely ranging from 0 to 40 for assisted focused parathyroidectomy (FP) (median 0), from 0 to 30 for performed FP (median 0), from 0 to 10 for assisted bilateral neck exploration (BNE) (median 0), and from 0 to 5 for performed BNE (median 0). This was in contrast with the personal views of the respondents. Most of them considered that graduating surgeons should assist and perform a higher minimum number of FP (19 out of 24 (79.2%) and 17 out of 24 (70.8%), respectively) and BNE (18 out of 24 (75.0%) and 15 out of 24 (62.5%), respectively). In the opinion of the respondents, the minimum number of parathyroid procedures that should be assisted and performed during residency varied widely ranging from 0 to 40 for assisted FP (median 10), from 0 to 30 for performed FP (median 10), from 0 to 30 for assisted BNE (median 7.5), and from 0 to 30 for performed BNE (median 5).

For the available fellowship programs, national delegates reported a wide range of experience in parathyroid surgery. The number of procedures assisted/performed varied widely ranging from 3 to 100 for assisted FP (median 7.5), from 10 to 50 for performed FP (median 20), and from 5 to 20 for assisted BNE (median 10), and were 10 for performed BNE (median 10). All the responding national delegates from countries in which there are no endocrine surgery fellowships programs available agreed that a post-residency training is needed to achieve the expected competency.

In the opinion of the respondents, the minimum number of parathyroid procedures that should be assisted and performed during an endocrine surgery fellowship varied widely ranging from 5 to 100 for assisted FP (median 20), from 10 to 60 for performed FP (median 25), from 10 to 30 for assisted BNE (median 20), and from 10 to 30 for performed BNE (median 15).

Using the search terms “parathyroidectomy AND training,” “parathyroidectomy AND residency,” and “parathyroidectomy AND fellowship,” 205 publications were identified on PubMed. However, only 12 of them were found to be relevant for this analysis [3, 5‐7, 9‐14, 18, 21]. The principal findings of the papers are summarized in Table 3.

All the studies included in the analysis were evaluating surgical training of US general surgery/otolaryngology residency programs or US endocrine surgery fellowship programs. No English literature was found on the training of general surgery/otolaryngology residents with regard to parathyroid surgery outside the USA.

In the USA, the data of the Residency Review Committee (RRC) of the Accreditation Council for Graduate Medical Education (ACGME) showed that there was a gradual but significant increase of the average numbers of endocrine surgical procedures from 1986 to 2008, parathyroidectomies in particular, performed by graduating general surgery residents (mean 4.1 parathyroid procedures in 1986 versus 9.6 in 2015) [5‐7, 9‐11, 14] (Table 3). Data of ACGME accredited otolaryngology—head and neck—surgery residency programs over 20 years (1996–2015) showed that the numbers of parathyroidectomies performed by graduating residents have steadily grown from 1.8 in 1996 to 16.3 in 2015 (an 800% increase) [3, 10, 14]. Zarebczan B. et al. [14] reported that most general surgery residents performed parathyroidectomies (and thyroidectomies) early during their training. Conversely, otolaryngology residents are performing most parathyroidectomies (and thyroidectomies) as chief residents (Table 3).

A recent survey conducted among AAES members reported that the respondents on average believe that all the steps of parathyroidectomy for a well-localized adenoma could be performed by a postgraduate 4-year surgery resident under direct supervision by an attending surgeon [3]. In addition, most of the respondents (79%) reported that a postgraduate final-year resident from their institution could autonomously perform a parathyroidectomy in a patient with concordant imaging [3] (Table 3).

In the late 1900s, in residency programs with one or more endocrine surgeon(s) in the teaching faculty, the number of parathyroidectomies performed by graduating general surgery residents was significantly higher as opposed to residency programs where such competency was missing [6] (Table 3). In addition, it has been reported that graduated general surgeons who choose to continue training in endocrine surgery performed significantly more endocrine cases during residency than the average graduating surgical resident (26 versus 9) [13] (Table 3).

Data on fellows training are less accurate since even in the USA there is no database for recognized fellowships like that of RRC [11]. Fellowship programs provide self-reported estimates on their operative volumes in thyroid, parathyroid, adrenal, neuroendocrine pancreas, and gastrointestinal endocrine procedures [11]. US fellowship programs showed significant differences in the numbers of endocrine operative cases performed at each program ranging from 10 to 500 parathyroidectomies [11]. During the early 2000s, the AAES recognized that endocrine surgery fellows graduated with a median of 80–90 parathyroid procedures (range 35–200) [9, 13] (Table 3). These numbers exceed the estimated numbers of performed cases to be competent (50, range 10–100) [13]. Given the rapid changes in parathyroid surgery over the past decade, in terms of preoperative localization studies, intraoperative PTH monitoring, and focused approaches in localized disease, AAES fellowship programs teach congruent management strategies that include focused dissection for localized disease and four-glands exploration if multiglandular disease is suspected, with intraoperative parathyroid hormone use and the intent to cure patients at the first surgery [22].

Trainees are involved in 45–73% of endocrine surgical operations performed at hospitals participating in the National Surgical Quality Improvement Program (NSQIP) of the American College of Surgeons [12, 18]. Between 2005 and 2008, senior residents assisted 36.7%, junior residents 32.9%, and endocrine surgery fellows 8.2% of 9101 parathyroidectomies [12]. Trainee-assisted operations were usually associated with longer operative time [12], especially if assisted by junior or senior residents [12, 18]. No difference in complications rate was observed when a trainee was involved [12, 18] (Table 3).

There was a poor reply rate to the questions related to adrenal surgery. Only three respondents (from Poland, Slovenia, and Turkey) described their personal experience. Median (range) numbers were 5 (3–30) for open adrenalectomies performed, 5 (5–50) for open adrenalectomies assisted, 5 (2–30) for minimally invasive adrenalectomies performed, and 10 (5–50) for minimally invasive adrenalectomies assisted.

Nine respondents declared their personal views suggesting that during training one should be exposed to a median of 10 (2–30) open adrenalectomies performed and 10 (10–50) assisted and 10 (5–30) minimally invasive adrenalectomies performed and 10 (10–50) assisted.

The national requirements for becoming a fellow in endocrine surgery varied widely with a median of 10 (5–60) for performed adrenalectomies and 15 (10–100) for assisted adrenalectomies. The national delegates view was that a median of 20 performed and 20 assisted adrenalectomies should be required but once again their views varied widely (range 7–60 and 15–100, respectively). These figures are well over the workload of many units and reinforce the need to centralize training in adrenal surgery in a small number of units with workload of probably over 20 cases/year.

Data published on the current exposure of surgical residents to adrenal surgery is focused mainly on the North American situation but anecdotal evidence from most European countries is very similar. In a survey of 22 graduates from a general surgery residency program, over 50% did not feel competent performing several operations including adrenalectomy and this was an area considered to be in need of educational improvement at a program level (in addition to other operations such as abdominoperineal resection, transanal excision of tumors, transhiatal esophagectomy, and Whipple operations). Surprisingly, the analysis of their logbooks showed that increased case volume correlated with competence for adrenalectomy at very small numbers (3 vs. 1) though no established surgeon would accept that having been involved or having performed three adrenalectomies equates to competence of the procedure [23].

The exposure to small numbers of adrenal procedures has been reported repeatedly over the last decades. A study that examined the operative experience of general surgery residents in endocrine surgery for the academic years 1986–1987 to 1993–1994 found this training to be inadequate due to low operative volume. In a review of more recent data from the Resident Statistic Summaries from 1994–1995 to 2003–2004, the average number of adrenalectomies per resident was 1.46 (for endocrine pancreas, the average was 0.14). The most common number of any of these procedures performed by US graduates was zero [11]. A more recent analysis of case-log data from 2004 to 2009 for American urology residents showed no improvement while for Canadian residents there was a slightly improved exposure [24]

Exposure to endocrine (adrenal) surgery and mentorship during residency programs are powerful factors that influence residents to pursue careers in endocrine surgery. In a recent survey, fellows performed significantly more endocrine surgery cases during residency than the average graduating chief resident and mentorship was a critical factor in fellows’ decisions to pursue endocrine surgery. Fellows graduated with a median (range) of 13 (0–60) laparoscopic adrenal operations (compared with 150 (50–300) thyroid operations) [13]. When questioned about the “ideal” exposure they would aim for about 30. Fellows in minimally invasive surgery (MIS) listed the case volume for adrenalectomy as insufficient (similar to the situation for laparoscopic gastric banding, colectomy, common bile duct exploration, gastrectomy, esophagectomy, splenectomy, hepatectomy, nephrectomy, and pancreatectomy) [25].

It is abundantly clear that adrenalectomy is clearly established on the list of topics for which training remains poor in consecutive generations of trainees.

There is a big variety concerning the numbers of operations for gastric, intestinal, and pancreatic procedures in the European countries. When it comes to training of surgical residents, the majority of operations performed on these organs (namely the stomach, the intestine, and the pancreas) are not performed on patients with NETs. This is a big difference as compared with thyroid gland, parathyroid gland, and adrenal gland procedures.

Our survey revealed that surgical residents according to the national surgical boards are required to perform between 0 and 15 gastric operations. However, the median number of required performed procedures was zero. Most national delegates felt that 4 (median) gastric operations should be performed by residents themselves. When asked for assisted gastric operations, the number was only slightly higher. According to the national delegates, 5 (median) operations were considered sufficient.

With regard to intestinal (small and large bowel) procedures, the numbers for required operations according to the national surgical boards were somewhat higher. In some countries, the minimum number of performed procedures was as high as 50. On average, however, only 5 (median) performed operations were required. The national delegates considered 12.5 (median) performed operations sufficient. Residents were required to assist 9 (median) intestinal operations while 10 (median) operations were considered appropriate by the national delegates.

In contrast to these numbers, pancreatic procedures are rather very rare. While the numbers for required performed (median 0) and assisted (median 0) operations were very low, slightly higher numbers (performed n = 2, assisted n = 5) were considered appropriate by the national delegates.

Adding the above numbers of gastric, intestinal, and pancreatic procedures, surgical residents should perform at least about 15 and assist another 15 GEP-NET operations according to the national delegates. The current number of required operations by the DES (performed n = 2, assisted n = 5) therefore appears very low.

As expected, the numbers considered to be appropriate for endocrine fellows were generally higher. While up to 30 performed and an additional 30 assisted gastric procedures were considered appropriate, the median number of appropriate procedures was 5 for performed and 8 for assisted operations. With regard to intestinal (small and large bowel) procedures, these numbers were slightly higher. The median numbers considered appropriate were 7 for performed and 10 for assisted operations. The median numbers were even relatively high for pancreatic procedures with 5 for performed and 20 for assisted procedures.

While the median numbers of operations considered to be appropriate overall were relatively high, the numbers that are required according to the national surgical boards were quite low in some countries. In addition, surgery of GEP-NETs is not part of a fellowship program in most European countries.

GEP-NETs are rather rare entities and their surgical treatment is not universally managed. Gastric NET may be surgically treated by general surgeons, upper-gastrointestinal surgeons, and endocrine surgeons. Enteric NETs may be operated on by general surgeons, lower gastrointestinal surgeons, colorectal surgeons, and endocrine surgeons. Pancreatic NETs may be in the same hands as gastric NETs. Since the surgical treatment of these tumors is distributed among various surgical subspecialties, any general statement is difficult to make. It has been shown that defining the right surgical fellowship program can be a challenge in itself [46].

Literature addressing training of surgical residents or fellows with regard to gastro-entero-pancreatic neuroendocrine tumors is almost non-existing. This is somewhat surprising owing to the fact that the surgical strategy on neuroendocrine tumors may differ somewhat from that used in their more common adenocarcinoma counterparts. This applies both to the resection of primary tumors and also to metastases [47]. Thus, there would be a lot to teach. Nevertheless, some literature exists on the training of both residents and fellows with regard to gastric, colorectal, and hepato-pancreato-biliary surgery in general and this will be addressed below. Of course, the procedures analyzed may differ a lot from those used in patients with GEP-NETs but the findings may give some important clues.