Abstract
Background
There has been a proliferation of gastrointestinal surgical fellowships; however, little is known regarding their association with surgical volume and management approaches.
Methods
Surveys were distributed to members of GI surgical societies. Responses were evaluated to define relationships between fellowship training and surgical practice with pancreatoduodenectomy (PD).
Results
Surveys were completed by 889 surgeons, 84.1% of whom had completed fellowship training. Fellowship completion was associated with a primarily HPB or surgical oncology-focused practice (p < 0.001), and greater median annual PD volume (p = 0.030). Transplant and HPB fellowship-trained respondents were more likely to have high-volume (≥20) annual practice (p = 0.005 and 0.029, respectively). Regarding putative fistula mitigation strategies, HPB-trained surgeons were more likely to use stents, biologic sealants, and autologous tissue patches (p = 0.007, <0.001 and 0.001, respectively). Surgical oncology trainees reported greater autologous patch use (p = 0.003). HPB fellowship-trained surgeons were less likely to routinely use intraperitoneal drainage (p = 0.036) but more likely to utilize early (POD ≤ 3) drain amylase values to guide removal (p < 0.001). Finally, HPB fellowship-trained surgeons were more likely to use the Fistula Risk Score in their practice (29 vs. 21%, p = 0.008).
Conclusion
Fellowship training correlated with significant differences in surgeon experience, operative approach, and use of available fistula mitigation strategies for PD.
Introduction
Over the last three decades, there has been a steady but certain trend toward subspecialization among trainees in general surgery, with more than three quarters of new graduates now pursuing fellowship training after core residency.1 Those interested in practicing hepatopancreatobiliary (HPB) surgery have traditionally pursued fellowship training in either surgical oncology or transplant surgery, with a limited number of these programs offering a designated HPB track. Less standardized apprenticeship training models at a variety of institutions have also offered a pathway to HPB surgery for surgical trainees in years past.3 More recently, formal HPB surgery fellowships—created and accredited through a collaboration between the Fellowship Council (FC) and Americas Hepato-Pancreato-Biliary Association (AHPBA)—now offer another pathway to specialization in HPB surgery.4
The curricula and foci among these three training programs vary considerably within North America, and there is even further heterogeneity globally. Little is known regarding the impact of fellowship training on the actual practice of HPB surgery. Two recent studies have assessed the relationship between fellowship training pathway and current operative caseload among HPB surgeons.3, 4 In order to further evaluate the influence of training pathway on HPB surgery, the present study examines practice patterns relating to a specific operation—pancreatoduodenectomy (PD)—that is a cornerstone of HPB surgery and is defined by numerous intraoperative strategies and management decisions. Many of these strategies have been evaluated in randomized controlled trials (RCT), which have established Level I evidence for best clinical practices. Therefore, examining the practice of PD offers insight into the question of whether fellowship training affects the implementation of best clinical practices in HPB surgery.
Materials and methods
Survey design and administration
The study protocol was approved by the Institutional Review Board (IRB) of the University of Pennsylvania. A web-based survey was designed and distributed to surgeons who perform pancreatic surgery through 22 international gastrointestinal surgical societies. The survey instrument was initially disseminated to the members of large international surgical societies, including the International Hepato-Pancreato-Biliary Association (IHPBA), the Society for Surgery of the Alimentary Tract (SSAT), and the Pancreas Club. Further support was also sought from the three major regional vessels of the IHPBA [the Americas Hepato-Pancreato-Biliary Association (AHPBA), the Asian-Pacific Hepato-Pancreato-Biliary Association (A-PHPBA), the European/African/Middle Eastern Hepato-Pancreato-Biliary Association (E-AHPBA)], as well as the national chapters that comprise these associations. The web-based survey instrument was made available in eight different languages, including English, Spanish, Portuguese, French, German, Italian, Japanese, and Mandarin Chinese. Given the membership numbers of the targeted surgical societies, it is estimated that the survey instrument was distributed to between 1500 and 2000 surgeons around the world, although a precise count is not possible due to the organizations' distribution conditions. Other results from this survey—focused on the global diversity of practice of PD—have been previously described.11
Surgeons were asked to report their region of practice, which was subsequently grouped into one of five regions (Asia/Australia, North America, South/Central America, Europe, or Africa/Middle East). Surgeons were also queried on a number of experience-based parameters, including age, years of experience as an attending surgeon, annual PD volume, and total career PD volume. The scope of each respondent's clinical practice was also characterized as was their route of training prior to becoming an independent, attending surgeon.
Questions regarding the practice patterns of individual surgeons were presented using a modified Likert scale: (i) never, 0%; (ii) occasionally, 1–25%; (iii) sometimes, 26–75%; (iv) frequently, 76–99%, and (v) always, 100%. Specific operative techniques and management decisions with known variation in practice were evaluated. Operative techniques included pancreaticogastrostomy (PG), dunking/invagination, duct occlusion, isolated Roux limb, anastomotic suturing techniques, trans-anastomotic stent placement, biological sealants (e.g. fibrin glue), autologous tissue patch use, and externalized drain placement. Management decisions referred to the administration of prophylactic somatostatin analogues (e.g. octreotide), and the practice of early drain removal [postoperative day (POD) ≤3] based upon drain amylase values.
Statistical analysis
Descriptive statistics are presented as frequencies for categorical variables, and as the mean ± standard deviation (SD) and median [interquartile range (IQR)] for continuous variables. Pearson's chi-squared test or Fisher's exact test, and independent Student's t-tests or analysis of variance (ANOVA) testing were used to analyze categorical and continuous variables, respectively. Non-parametric comparisons of continuous variables were assessed by Wilcoxon rank-sum tests or Kruskal–Wallis one-way ANOVA. P-values of ≤0.05 were considered statistically significant. All tests were two-sided. Statistical computations were performed utilizing SPSS version 22.0 (IBM Corp., Armonk, NY, USA) statistical software.
Results
Demographics and surgical experience
Surveys were completed by 889 practicing surgeons, representing six continents and eight languages. There were significant differences in the demographic profile and experience of surgeons based upon fellowship training (Table 1). There were regional differences in rates of fellowship training, with surgeons in South/Central America and Africa/Middle East significantly less likely to complete a fellowship (p < 0.001). Completion of fellowship training was associated with a primarily HPB or surgical oncology-focused practice, while surgeons who had not completed a fellowship were significantly more likely to practice general surgery primarily. Furthermore, fellowship-trained surgeons were significantly younger (p = 0.002) and less experienced than non-fellowship-trained surgeons (p < 0.001). However, they reported a greater median PD volume during the last calendar year (p = 0.030). In particular, respondents who had completed a transplant or HPB fellowship were more likely to have a high-volume annual PD practice (p = 0.005 and 0.029, respectively), defined by the established high-volume cutoff of 20 PD per year.7 Surgeons without fellowship training had a greater mean career PD volume (p = 0.033), but there were non-significant differences in the proportion of surgeons who had passed two reported learning curves for PD (≥507: p = 0.353; >608: p = 0.077).
Table 1.
Demographics and experience of surgeons practicing pancreas surgery, stratified by fellowship training
| Variable | Overall | Any fellowship |
HPB fellowship |
Surgical oncology fellowship |
Transplant |
Other fellowship |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | ||
| Surgeons, N (%) | 889 | 141 (16) | 748 (84) | – | 431 (49) | 458 (51) | – | 619 (69) | 270 (30) | – | 695 (78) | 194 (22) | – | 682 (77) | 207 (23) | – |
| Region, N (%) | ||||||||||||||||
| Asia/Australia | 209 (24) | 26 (18) | 183 (25) | <0.001 | 61 (14) | 148 (33) | <0.001 | 169 (27) | 40 (15) | <0.001 | 158 (23) | 51 (27) | 0.337 | 143 (21) | 66 (32) | <0.001 |
| North America | 277 (31) | 33 (23) | 244 (33) | 175 (41) | 102 (22) | 180 (29) | 97 (36) | 227 (33) | 50 (26) | 231 (34) | 46 (22) | |||||
| South/Central America | 127 (14) | 14 (10) | 113 (15) | 80 (19) | 47 (10) | 55 (9) | 72 (27) | 102 (15) | 25 (13) | 79 (12) | 48 (23) | |||||
| Europe | 249 (28) | 66 (47) | 183 (25) | 102 (24) | 147 (32) | 200 (32) | 49 (18) | 190 (28) | 59 (31) | 206 (31) | 43 (21) | |||||
| Africa/Middle East | 21 (2) | 2 (1) | 19 (3) | 9 (2) | 12 (3) | 13 (2) | 8 (3) | 15 (2) | 6 (3) | 17 (2) | 4 (2) | |||||
| Scope of clinical practice, N (%) | ||||||||||||||||
| Pancreas only | 32 (4) | 11 (8) | 21 (3) | <0.001 | 16 (4) | 16 (4) | <0.001 | 19 (3) | 13 (5) | <0.001 | 28 (4) | 4 (2) | <0.001 | 25 (4) | 7 (3) | <0.001 |
| HPB | 486 (55) | 60 (43) | 426 (57) | 155 (36) | 331 (72) | 403 (65) | 83 (31) | 333 (48) | 153 (79) | 388 (57) | 98 (48) | |||||
| Surgical oncology | 195 (22) | 23 (16) | 172 (23) | 152 (35) | 43 (9) | 52 (8) | 143 (53) | 182 (26) | 13 (7) | 165 (24) | 30 (15) | |||||
| General surgery | 144 (16) | 42 (30) | 102 (14) | 89 (21) | 55 (12) | 120 (19) | 24 (9) | 127 (18) | 17 (9) | 83 (12) | 61 (30) | |||||
| Other | 30 (3) | 5 (4) | 25 (3) | 17 (4) | 13 (3) | 24 (4) | 6 (2) | 23 (3) | 7 (4) | 20 (3) | 10 (5) | |||||
| Age | ||||||||||||||||
| Mean (SD) | 46.9 (10.1) | 49.4 (11.3) | 46.4 (9.8) | 0.005 | 48.0 (10.3) | 45.8 (9.7) | 0.001 | 47.4 (10.3) | 45.7 (9.4) | 0.016 | 46.9 (10.4) | 46.5 (8.7) | 0.547 | 46.5 (10.1) | 47.9 (10.1) | 0.097 |
| Median (IQR) | 45 (39–54) | 50 (40–57) | 45 (39–53) | 0.002 | 48 (40–55) | 45 (38–52) | 0.001 | 46 (40–55) | 44 (38–51) | 0.012 | 46 (39–54) | 45 (40–52) | 0.923 | 45 (39–53.5) | 46 (40–55) | 0.088 |
| Years of experience | ||||||||||||||||
| Mean (SD) | 14.5 (10.6) | 18.1 (11.6) | 13.8 (10.3) | <0.001 | 15.8 (10.6) | 13.3 (10.5) | 0.001 | 15.0 (10.8) | 13.5 (10.1) | 0.061 | 14.7 (10.9) | 14.0 (9.4) | 0.39 | 14.0 (10.5) | 16.1 (10.8) | 0.013 |
| Median (IQR) | 13 (6–21.75) | 17 (9.5–27.5) | 12 (6–20) | <0.001 | 14.5 (7–24) | 10 (5–20) | 0.0002 | 14 (6–23) | 11 (6–20) | 0.087 | 14 (6–22) | 12 (6.75–20) | 0.884 | 12 (5–20) | 15 (7–24) | 0.009 |
| PD volume | ||||||||||||||||
| 2013 calendar year | ||||||||||||||||
| Mean (SD) | 18.1 (18.4) | 17.1 (21.0) | 18.3 (17.9) | 0.482 | 17.4 (20.1) | 18.7 (16.6) | 0.296 | 18.1 (18.5) | 18.1 (18.3) | 0.966 | 17.5 (18.1) | 20.3 (19.2) | 0.055 | 18.9 (18.9) | 15.5 (16.6) | 0.021 |
| Median (IQR) | 12 (6–25) | 10 (5–20.5) | 13 (6–25) | 0.03 | 11 (5–24) | 15 (8–25) | 0.0003 | 12.5 (6–25) | 12 (6–25) | 0.99 | 12 (5–24) | 15 (8–25.75) | 0.007 | 14 (7–25) | 10 (4–20) | 0.001 |
| Annual volume designation | ||||||||||||||||
| ≥20 – high-volume | 304 (35) | 39 (28) | 265 (36) | 0.068 | 133 (31) | 171 (39) | 0.029 | 212 (35) | 92 (35) | 0.887 | 220 (33) | 84 (44) | 0.005 | 246 (37) | 58 (29) | 0.027 |
| Career total | ||||||||||||||||
| Mean (SD) | 152.3 (215.9) | 205.9 (338.0) | 142.3 (183.1) | 0.033 | 170.9 (256.1) | 134.8 (168.3) | 0.015 | 160.2 (233.7) | 134 (167.4) | 0.102 | 155.2 (223.9) | 141.5 (184.9) | 0.434 | 155.1 (220.7) | 143.0 (199.7) | 0.481 |
| Median (IQR) | 79 (30–200) | 100 (35–204) | 71 (30–200) | 0.121 | 90 (25–206.5) | 70 (30–181.3) | 0.327 | 90 (30–200) | 60 (25–200) | 0.191 | 80 (27.5–200) | 75 (30–180) | 0.911 | 90 (30–200) | 60 (23–200) | 0.158 |
| Surpassed learning curve, N (%) | ||||||||||||||||
| ≥50 PD | 555 (64) | 92 (67) | 463 (63) | 0.353 | 269 (64) | 286 (64) | 0.954 | 395 (65) | 160 (60) | 0.185 | 431 (63) | 124 (65) | 0.679 | 434 (65) | 121 (59) | 0.116 |
| >60 PD | 468 (54) | 83 (61) | 385 (52) | 0.077 | 235 (56) | 233 (52) | 0.247 | 340 (56) | 128 (48) | 0.036 | 367 (54) | 101 (53) | 0.806 | 369 (55) | 99 (48) | 0.079 |
Operative technical approaches
Fellowship training was also associated with considerable variability in how surgeons perform the pancreaticoenteric reconstruction during PD (Table 2, Table 3). While there were no significant differences in the preferred anastomotic reconstruction—nearly 90% of respondents preferred pancreaticojejunostomy (PJ) over pancreaticogastrostomy (PG), regardless of fellowship training—surgical oncology trainees were more likely to vary their reconstructive approach depending on the circumstances of individual patients (p = 0.019). No significant differences were reported based on suture technique or suture type for creating the inner layer of the anastomosis, but HPB and surgical oncology fellowship training was associated with greater proclivity for absorbable suture use in the outer suture layer (p = 0.002 and 0.004, respectively). HPB-trained surgeons were less likely to use the dunking/invagination technique (p = 0.044), as opposed to those with surgical oncology or transplant training. Duct occlusion and isolated Roux limb reconstruction remained unpopular approaches among all respondents, regardless of fellowship training.
Table 2.
Type of pancreaticoenteric reconstruction, stratified by fellowship training
| Variable | Overall | Any fellowship |
HPB fellowship |
Surgical oncology fellowship |
Transplant |
Other fellowship |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | ||
| Surgeons, N (%) | 889 | 141 (16) | 748 (84) | – | 431 (49) | 458 (51) | – | 619 (70) | 270 (30) | – | 695 (78) | 194 (22) | – | 682 (77) | 207 (23) | – |
| Preferred anastomotic reconstruction | ||||||||||||||||
| Pancreaticojejunostomy | 782 (89) | 124 (88) | 658 (89) | 0.826 | 383 (90) | 399 (88) | 0.638 | 543 (88) | 239 (89) | 0.184 | 611 (89) | 171 (89) | 0.063 | 599 (88) | 183 (89) | 0.151 |
| End-to-side-duct-to-mucosa | 595 (67.4) | 89 (63) | 506 (68) | 283 (66) | 312 (68) | 404 (66) | 191 (71) | 455 (66) | 140 (73) | 463 (68) | 132 (64) | |||||
| End-to-side-invagination | 106 (12) | 17 (12) | 89 (12) | 55 (13) | 51 (11) | 83 (14) | 23 (9) | 86 (13) | 20 (10) | 74 (11) | 32 (16) | |||||
| End-to-end-invagination | 60 (7) | 13 (9) | 47 (6) | 35 (8) | 25 (6) | 42 (7) | 18 (7) | 51 (7) | 9 (5) | 45 (7) | 15 (7) | |||||
| End-to-end-binding | 21 (2) | 5 (4) | 16 (2) | 10 (2) | 11 (2) | 14 (2) | 7 (3) | 19 (3) | 2 (1) | 17 (3) | 4 (2) | |||||
| Pancreaticogastrostomy | 86 (10) | 14 (10) | 72 (10) | 37 (9) | 49 (11) | 63 (10) | 23 (9) | 67 (10) | 19 (10) | 71 (11) | 15 (7) | |||||
| Duct-to-mucosa | 20 (2) | 3 (2) | 17 (2) | 10 (2) | 10 (2) | 11 (2) | 9 (3) | 20 (3) | 0 (0) | 16 (2) | 4 (2) | |||||
| Invagination/dunking | 66 (8) | 11 (8) | 55 (7) | 27 (6) | 39 (9) | 52 (9) | 14 (5) | 47 (7) | 19 (10) | 55 (8) | 11 (5) | |||||
| Ductal occlusion | 3 (0) | 1 (0) | 2 (0) | 2 (1) | 1 (0) | 2 (0) | 1 (0) | 3 (0) | 0 (0) | 1 (0) | 2 (1) | |||||
| Other | 12 (1) | 2 (1) | 10 (1) | 5 (1) | 7 (2) | 7 (1) | 5 (2) | 9 (1) | 3 (2) | 7 (1) | 5 (2) | |||||
| Pancreaticogastrostomy | ||||||||||||||||
| Never | 544 (62) | 82 (59) | 462 (63) | 0.198 | 264 (62) | 280 (62) | 0.816 | 379 (62) | 165 (62) | 0.451 | 420 (61) | 124 (65) | 0.327 | 419 (62) | 125 (63) | 0.559 |
| Occasionally (1–25%) | 224 (26) | 36 (26) | 188 (26) | 107 (25) | 117 (26) | 151 (25) | 73 (27) | 179 (26) | 45 (24) | 173 (26) | 51 (26) | |||||
| Sometimes (26–75%) | 33 (4) | 9 (4) | 24 (3) | 18 (4) | 15 (3) | 22 (4) | 11 (4) | 30 (4) | 3 (2) | 23 (3) | 10 (5) | |||||
| Frequently (76–99%) | 43 (5) | 5 (4) | 38 (5) | 18 (4) | 25 (6) | 35 (6) | 8 (3) | 33 (5) | 10 (5) | 33 (5) | 10 (5) | |||||
| Always | 32 (3) | 8 (6) | 24 (3) | 17 (4) | 15 (3) | 21 (4) | 11 (4) | 23 (3) | 9 (5) | 28 (4) | 4 (2) | |||||
| Same reconstruction on all cases | ||||||||||||||||
| No | 316 (36) | 51 (36) | 265 (36) | 0.918 | 156 (37) | 160 (35) | 0.654 | 204 (33) | 112 (42) | 0.019 | 255 (37) | 61 (31) | 0.153 | 247 (36) | 69 (34) | 0.543 |
| Yes | 567 (64) | 90 (64) | 477 (64) | 271 (64) | 296 (65) | 409 (67) | 158 (59) | 434 (63) | 133 (69) | 433 (64) | 134 (66) | |||||
Table 3.
The utilization of operative strategies for the pancreaticoenteric reconstruction, stratified by fellowship training
| Variable | Overall | Any fellowship |
HPB fellowship |
Surgical oncology fellowship |
Transplant |
Other fellowship |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | ||
| Surgeons, N (%) | 889 | 141 (15) | 748 (84) | – | 431 (48) | 458 (52) | – | 619 (70) | 270 (30) | – | 695 (78) | 194 (22) | – | 682 (77) | 207 (23) | – |
| Dunking or invagination | ||||||||||||||||
| Never | 325 (37) | 48 (35) | 277 (38) | 0.699 | 138 (33) | 187 (42) | 0.044 | 220 (36) | 105 (40) | 0.298 | 247 (36) | 78 (40) | 0.393 | 256 (38) | 69 (35) | 0.581 |
| Occasionally (1–25%) | 292 (33) | 43 (31) | 249 (34) | 144 (34) | 148 (33) | 198 (33) | 94 (36) | 230 (34) | 62 (32) | 221 (33) | 71 (36) | |||||
| Sometimes (26–75%) | 41 (5) | 8 (6) | 33 (5) | 22 (5) | 19 (4) | 30 (5) | 11 (4) | 35 (5) | 6 (3) | 33 (5) | 8 (4) | |||||
| Frequently (76–99%) | 122 (14) | 23 (17) | 99 (14) | 65 (15) | 57 (13) | 87 (14) | 35 (13) | 99 (15) | 23 (12) | 97 (14) | 25 (13) | |||||
| Always | 93 (11) | 17 (12) | 76 (10) | 54 (13) | 39 (9) | 73 (12) | 20 (8) | 68 (10) | 25 (13) | 67 (10) | 26 (13) | |||||
| Duct occlusion | ||||||||||||||||
| Never | 788 (91) | 121 (90) | 667 (92) | 0.740 | 385 (9) | 403 (90) | 0.365 | 546 (91) | 242 (91) | 0.913 | 622 (92) | 166 (88) | 0.36 | 608 (92) | 180 (91) | 0.918 |
| Occasionally (1–25%) | 62 (7) | 12 (9) | 50 (7) | 29 (7) | 33 (7) | 42 (7) | 20 (8) | 43 (6) | 19 (10) | 46 (7) | 16 (8) | |||||
| Sometimes (26–75%) | 7 (1) | 1 (1) | 6 (1) | 2 (1) | 5 (1) | 6 (1) | 1 (0) | 5 (1) | 2 (1) | 6 (1) | 1 (1) | |||||
| Frequently (76–99%) | 3 (0) | 1 (1) | 2 (0) | 1 (0) | 2 (0) | 2 (0) | 1 (0) | 3 (0) | 0 (0) | 2 (0) | 1 (1) | |||||
| Always | 3 (0) | 0 (0) | 3 (0) | 0 (0) | 3 (1) | 2 (0) | 1 (0) | 2 (0) | 1 (1) | 2 (0) | 1 (1) | |||||
| Isolated Roux limb | ||||||||||||||||
| Never | 591 (68) | 91 (67) | 500 (68) | 0.925 | 277 (66) | 314 (70) | 0.675 | 412 (68) | 179 (67) | 0.761 | 467 (69) | 124 (65) | 0.528 | 469 (70) | 122 (61) | 0.08 |
| Occasionally (1–25%) | 157 (18) | 23 (17) | 134 (18) | 81 (19) | 76 (17) | 111 (18) | 46 (17) | 118 (17) | 39 (21) | 109 (16) | 48 (24) | |||||
| Sometimes (26–75%) | 24 (3) | 5 (4) | 19 (3) | 14 (3) | 10 (2) | 17 (3) | 7 (3) | 17 (3) | 7 (4) | 20 (3) | 4 (2) | |||||
| Frequently (76–99%) | 40 (5) | 7 (5) | 33 (5) | 20 (5) | 20 (4) | 24 (4) | 16 (6) | 34 (5) | 6 (3) | 31 (5) | 9 (5) | |||||
| Always | 57 (7) | 10 (7) | 47 (6) | 26 (6) | 31 (7) | 39 (7) | 18 (7) | 43 (6) | 14 (7) | 41 (6) | 16 (8) | |||||
| Suture technique | ||||||||||||||||
| Single layer | 218 (25) | 43 (31) | 175 (24) | 0.081 | 106 (25) | 112 (25) | 0.928 | 156 (25) | 62 (23) | 0.500 | 164 (24) | 54 (28) | 0.230 | 168 (25) | 50 (24) | 0.910 |
| Double layer | 665 (75) | 98 (69) | 567 (76) | 321 (75) | 344 (75) | 459 (75) | 206 (77) | 526 (76) | 139 (72) | 510 (75) | 155 (76) | |||||
| Suture type – inner layer | ||||||||||||||||
| Absorbable monofilament | 623 (72) | 88 (67) | 535 (73) | 0.433 | 286 (69) | 337 (75) | 0.314 | 430 (72) | 193 (72) | 0.201 | 479 (71) | 144 (77) | 0.030 | 488 (74) | 135 (68) | 0.265 |
| Absorbable braided | 73 (8) | 12 (9) | 61 (8) | 40 (10) | 33 (7) | 51 (9) | 22 (8) | 68 (10) | 5 (3) | 49 (7) | 24 (12) | |||||
| Non-absorbable braided | 47 (5) | 9 (7) | 38 (5) | 22 (5) | 25 (6) | 30 (5) | 17 (6) | 37 (6) | 10 (5) | 34 (5) | 13 (7) | |||||
| Other | 37 (4) | 5 (4) | 32 (4) | 22 (5) | 15 (3) | 21 (4) | 16 (6) | 29 (4) | 8 (4) | 28 (4) | 9 (5) | |||||
| No inner layer | 84 (10) | 18 (14) | 66 (9) | 43 (10) | 41 (9) | 65 (11) | 19 (7) | 63 (9) | 21 (11) | 65 (10) | 19 (10) | |||||
| Suture type – outer layer | ||||||||||||||||
| Absorbable monofilament | 317 (36) | 49 (36) | 268 (36) | 0.838 | 125 (30) | 192 (42) | 0.002 | 235 (39) | 82 (31) | 0.004 | 234 (34) | 83 (43) | 0.004 | 229 (34) | 88 (43) | 0.010 |
| Absorbable braided | 98 (11) | 13 (9) | 85 (12) | 51 (12) | 47 (10) | 59 (10) | 39 (15) | 88 (13) | 10 (5) | 69 (10) | 29 (14) | |||||
| Non-absorbable braided | 360 (41) | 62 (45) | 298 (40) | 189 (45) | 171 (38) | 255 (42) | 105 (39) | 284 (42) | 76 (40) | 295 (44) | 65 (32) | |||||
| Other | 70 (8) | 10 (7) | 60 (8) | 42 (10) | 28 (6) | 38 (6) | 32 (12) | 58 (9) | 12 (6) | 58 (9) | 12 (6) | |||||
| No outer layer | 32 (4) | 4 (3) | 28 (4) | 15 (4) | 17 (4) | 23 (4) | 9 (3) | 21 (3) | 11 (6) | 23 (3) | 9 (4) | |||||
Fistula mitigation strategies
Surgeons were also queried as to their use of operative techniques and management strategies available to mitigate the risk of postoperative pancreatic fistula development (Table 4, Table 5). Overall, completion of any fellowship was not correlated with significant variation in the use of trans-anastomotic stents, but certain fellowships did affect surgeons' stenting preferences. In particular, HPB-trained surgeons were more likely to report routine use of stents (p = 0.007). Training under the HPB fellowship paradigm also correlated with always using internal trans-anastomotic stents, specifically (16 vs. 9%, p = 0.001).
Table 4.
The utilization of purported fistula mitigation strategies, stratified by fellowship training
| Variable | Overall | Any fellowship |
HPB fellowship |
Surgical oncology fellowship |
Transplant |
Other fellowship |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | ||
| Surgeons, N (%) | 889 | 141 (16) | 748 (84) | – | 431 (49) | 458 (51) | – | 619 (70) | 270 (30) | – | 695 (78) | 194 (22) | – | 682 (77) | 207 (23) | – |
| Trans-anastomotic stents | ||||||||||||||||
| Never | 233 (26) | 43 (31) | 190 (26) | 0.315 | 119 (28) | 114 (25) | 0.007 | 161 (26) | 72 (27) | 0.074 | 180 (26) | 53 (28) | 0.48 | 184 (27) | 49 (24) | 0.168 |
| Occasionally (1–25%) | 229 (26) | 30 (21) | 199 (27) | 116 (27) | 113 (25) | 148 (24) | 81 (30) | 182 (26) | 47 (24) | 164 (24) | 65 (32) | |||||
| Sometimes (26–75%) | 106 (12) | 21 (15) | 85 (12) | 56 (13) | 50 (11) | 70 (11) | 36 (13) | 89 (13) | 17 (9) | 88 (13) | 18 (9) | |||||
| Frequently (76–99%) | 175 (20) | 23 (16) | 152 (21) | 89 (21) | 86 (19) | 124 (20) | 51 (19) | 132 (19) | 43 (22) | 134 (20) | 41 (20) | |||||
| Always | 139 (16) | 23 (16) | 116 (16) | |||||||||||||
| 47 (11) | 92 (20) | 109 (18) | 30 (11) | 106 (15) | 33 (17) | 108 (16) | 31 (15) | |||||||||
| Type of stent used | ||||||||||||||||
| Internal | 477 (71) | 67 (66) | 410 (72) | 0.532 | 217 (67) | 260 (74) | 0.138 | 337 (72) | 140 (68) | 0.535 | 363 (68) | 114 (81) | 0.005 | 363 (71) | 114 (70) | 0.413 |
| External | 121 (18) | 20 (20) | 101 (18) | 64 (20) | 57 (16) | 83 (18) | 38 (19) | 108 (20) | 13 (10) | 94 (18) | 27 (17) | |||||
| Depends on patient | 76 (11) | 14 (14) | 62 (11) | 42 (13) | 34 (10) | 49 (10) | 27 (13) | 62 (12) | 14 (10) | 53 (10) | 23 (14) | |||||
| Biological sealants | ||||||||||||||||
| Never | 572 (65) | 85 (62) | 487 (66) | 0.164 | 265 (63) | 307 (68) | <0.001 | 400 (66) | 172 (64) | 0.927 | 445 (65) | 127 (66) | 0.85 | 426 (63) | 146 (72) | 0.23 |
| Occasionally (1–25%) | 181 (21) | 36 (26) | 145 (20) | 94 (22) | 87 (19) | 124 (20) | 57 (21) | 144 (21) | 37 (19) | 148 (22) | 33 (16) | |||||
| Sometimes (26–75%) | 49 (6) | 6 (4) | 43 (6) | 27 (6) | 22 (5) | 32 (5) | 17 (6) | 40 (6) | 9 (5) | 39 (6) | 10 (5) | |||||
| Frequently (76–99%) | 43 (5) | 9 (7) | 34 (5) | 32 (8) | 11 (2) | 29 (5) | 14 (5) | 32 (5) | 11 (6) | 36 (5) | 7 (3) | |||||
| Always | 34 (4) | 2 (1) | 32 (4) | 6 (1) | 28 (6) | 25 (4) | 9 (3) | 25 (4) | 9 (5) | 26 (4) | 8 (4) | |||||
| Autologous tissue patches | ||||||||||||||||
| Never | 538 (62) | 87 (63) | 451 (61) | 0.741 | 251 (59) | 287 (64) | 0.001 | 391 (64) | 147 (55) | 0.003 | 417 (61) | 121 (63) | 0.162 | 404 (60) | 134 (66) | 0.041 |
| Occasionally (1–25%) | 159 (18) | 20 (15) | 139 (19) | 79 (19) | 80 (18) | 108 (18) | 51 (19) | 119 (17) | 40 (21) | 118 (18) | 41 (20) | |||||
| Sometimes (26–75%) | 61 (7) | 10 (7) | 51 (7) | 32 (8) | 29 (6) | 43 (7) | 18 (7) | 46 (7) | 15 (8) | 48 (7) | 13 (6) | |||||
| Frequently (76–99%) | 67 (8) | 13 (9) | 54 (7) | 46 (11) | 21 (5) | 33 (5) | 34 (13) | 60 (9) | 7 (4) | 61 (9) | 6 (3) | |||||
| Always | 50 (6) | 8 (6) | 42 (6) | 15 (4) | 35 (8) | 33 (5) | 17 (6) | 40 (6) | 10 (5) | 41 (6) | 9 (4) | |||||
| Prophylactic somatostatin analogues | ||||||||||||||||
| Never | 357 (40) | 58 (41) | 299 (40) | 0.226 | 200 (47) | 157 (34) | <0.001 | 234 (38) | 123 (46) | 0.001 | 285 (41) | 72 (37) | 0.462 | 278 (41) | 79 (39) | 0.639 |
| Occasionally (1–25%) | 201 (23) | 27 (19) | 174 (24) | 96 (23) | 105 (23) | 135 (22) | 66 (25) | 155 (23) | 46 (24) | 149 (22) | 52 (26) | |||||
| Sometimes (26–75%) | 110 (13) | 25 (18) | 85 (12) | 57 (13) | 53 (12) | 72 (12) | 38 (14) | 88 (13) | 22 (11) | 89 (13) | 21 (10) | |||||
| Frequently (76–99%) | 97 (11) | 12 (9) | 85 (12) | 32 (8) | 65 (14) | 73 (12) | 24 (7) | 77 (11) | 20 (10) | 76 (11) | 21 (10) | |||||
| Always | 118 (13) | 19 (14) | 99 (13) | 41 (10) | 77 (17) | 100 (16) | 18 (7) | 85 (12) | 33 (17) | 88 (13) | 30 (15) | |||||
| Use the FRS to guide clinical practice | ||||||||||||||||
| No | 659 (75) | 110 (79) | 549 (74) | 0.184 | 335 (79) | 324 (71) | 0.008 | 449 (73) | 210 (78) | 0.120 | 511 (74) | 148 (76) | 0.548 | 512 (76) | 147 (72) | 0.272 |
| Yes | 224 (25) | 29 (21) | 195 (26) | 91 (21) | 133 (29) | 165 (27) | 59 (22) | 178 (26) | 46 (24) | 166 (24) | 58 (28) | |||||
Table 5.
Drain management strategies, stratified by fellowship training
| Variable | Overall | Any fellowship |
HPB fellowship |
Surgical oncology fellowship |
Transplant |
Other fellowship |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | No | Yes | p-value | ||
| Surgeons, N (%) | 889 | 141 (16) | 748 (84) | – | 431 (49) | 458 (52) | – | 619 (70) | 270 (30) | – | 695 (78) | 194 (22) | – | 682 (77) | 207 (23) | – |
| Prophylactic, intraperitoneal drainage | ||||||||||||||||
| Never | 124 (14) | 15 (11) | 109 (15) | 0.274 | 52 (12) | 72 (16) | 0.036 | 97 (16) | 27 (10) | 0.056 | 93 (14) | 31 (16) | 0.756 | 95 (14) | 29 (14) | 0.712 |
| Occasionally (1–25%) | 84 (10) | 9 (6) | 75 (10) | 34 (8) | 50 (11) | 58 (10) | 26 (10) | 65 (9) | 19 (10) | 66 (10) | 18 (9) | |||||
| Sometimes (26–75%) | 39 (4) | 6 (4) | 33 (4) | 26 (6) | 13 (3) | 23 (4) | 16 (6) | 33 (5) | 6 (3) | 32 (5) | 7 (3) | |||||
| Frequently (76–99%) | 114 (13) | 16 (11) | 98 (13) | 53 (12) | 61 (13) | 85 (14) | 29 (11) | 91 (13) | 23 (12) | 92 (14) | 22 (11) | |||||
| Always | 521 (59) | 94 (67) | 427 (58) | 262 (61) | 259 (57) | 350 (57) | 171 (64) | 407 (59) | 114 (59) | 394 (58) | 127 (63) | |||||
| Type of drain(s) regularly used | ||||||||||||||||
| Jackson–Pratt | 340 (39) | 51 (37) | 289 (39) | 0.800 | 160 (38) | 180 (40) | 0.262 | 247 (41) | 93 (35) | 0.001 | 260 (38) | 80 (42) | 0.079 | 269 (40) | 71 (35) | 0.033 |
| Blake | 301 (35) | 45 (33) | 256 (35) | 155 (37) | 146 (32) | 185 (30) | 116 (44) | 244 (36) | 57 (30) | 231 (34) | 70 (35) | |||||
| Penrose | 89 (10) | 17 (12) | 72 (10) | 47 (11) | 42 (9) | 62 (10) | 27 (10) | 76 (11) | 13 (7) | 57 (9) | 32 (16) | |||||
| Other | 121 (14) | 22 (16) | 99 (14) | 51 (12) | 70 (16) | 99 (16) | 22 (8) | 86 (13) | 35 (18) | 95 (14) | 26 (13) | |||||
| Number of drain(s) regularly used | ||||||||||||||||
| 1 | 252 (30) | 31 (24) | 221 (31) | 0.299 | 122 (30) | 130 (29) | 0.387 | 178 (30) | 74 (28) | 0.661 | 198 (30) | 54 (29) | 0.383 | 200 (31) | 52 (26) | 0.205 |
| 2 | 518 (61) | 83 (64) | 435 (60) | 253 (61) | 265 (60) | 352 (60) | 166 (63) | 397 (60) | 121 (64) | 393 (60) | 125 (63) | |||||
| 3+ | 59 (7) | 11 (8) | 48 (7) | 23 (6) | 36 (8) | 44 (7) | 15 (6) | 51 (8) | 8 (4) | 46 (7) | 13 (7) | |||||
| Varies each case | 26 (3) | 6 (5) | 20 (3) | 15 (4) | 11 (3) | 17 (3) | 9 (3) | 20 (3) | 6 (3) | 16 (2) | 10 (5) | |||||
| Remove drains using early amylase values | ||||||||||||||||
| No | 478 (55) | 68 (49) | 410 (56) | 0.154 | 267 (63) | 211 (47) | <0.001 | 319 (53) | 159 (59) | 0.075 | 369 (54) | 109 (57) | 0.538 | 365 (54) | 113 (57) | 0.586 |
| Yes | 394 (45) | 70 (51) | 324 (44) | 155 (37) | 239 (53) | 285 (47) | 109 (41) | 311 (46) | 83 (43) | 307 (46) | 87 (44) | |||||
| POD drain removal, median (IQR) | 5 (4–7) | 5 (4–7) | 0.867 | 5 (4–7) | 5 (4–7) | 0.002 | 5 (4–7) | 5.25 (4.5–7) | <0.001 | 5 (4–7) | 5 (4–7) | 0.359 | 5 (4–7) | 5 (5–7) | 0.059 | |
Other proposed strategies for ensuring the integrity of the pancreatoenteric anastomosis include biologic sealants and autologous tissue patches. As with trans-anastomotic stent placement, completion of any fellowship was not associated with significant differences in employment of these techniques; however, HPB fellowship-trained surgeons demonstrated greater propensity for the use of both biologic sealants and autologous tissue patches (p < 0.001 and 0.001, respectively). Respondents with surgical oncology training were also more likely to employ tissue patches than those who had not completed this training pathway (p = 0.003).
Routine intraperitoneal drainage was common among all surgeons, regardless of fellowship training, with only 14% of surgeons claiming to never employ drains. HPB fellowship-trained surgeons, in particular, were significantly less likely to use intraperitoneal drainage (p = 0.036). Irrespective of fellowship training, the majority of surgeons prefer placing two drains for prophylaxis. In terms of the type of drains employed, respondents with surgical oncology training preferred the use of Blake drains over Jackson–Pratt (JP) drains whereas surgeons from other training backgrounds favor JP drains (p = 0.001).
In terms of managing drains once they are employed, approximately 45% of respondents indicated that they rely upon drain amylase values to guide early (POD≤3) drain removal. HPB fellowship-trained surgeons were significantly more likely to favor this practice, with 53% reporting consideration of early drain amylase values in management decisions (p < 0.001). Globally, the mean preferred postoperative day of drain removal was POD 5 for all pancreas surgeons. HPB fellowship-trained surgeons preferred removing drains on a more aggressive timetable, whereas respondents from a surgical oncology background preferred drain removal at a later median POD (p = 0.002 and 0.0003, respectively).
Prophylactic somatostatin analogues are never administered by over 40% of respondents and always applied by approximately 13% of responding surgeons—regardless of whether they have completed fellowship training. Surgeons with HPB training showed a strong preference for the use of somatostatin analogues (p < 0.001), while respondents completing surgical oncology fellowships were less likely to administer them on a routine basis (p = 0.0003).
The utilization of fistula mitigation strategies has become increasingly dependent upon understanding risk stratification for the development of clinically-relevant fistulas. A highly validated tool for measuring aggregate fistula risk is the Fistula Risk Score (FRS).25 Although completion of any fellowship was not associated with utilization of the FRS to guide clinical care (26 vs. 21%, p = 0.184), HPB fellowship-trained surgeons were significantly more likely to use the FRS in their clinical practice (29 vs. 21%, p = 0.008).
Discussion
In recent years, several North American studies have demonstrated that nearly 80% of graduating general surgery residents have chosen to pursue fellowship training after residency.1, 12 Surgical trainees interested in HPB surgery, in particular, have sought fellowship training to augment the limited HPB operative exposure attained by the average general surgery resident. A recent study found that most surgical residents in the United States perform less than twelve pancreas, ten hepatic, and six complex biliary procedures during general surgery residency.2
Historically, the path to a HPB-focused surgical practice involved completion of an apprentice-like model or, more recently, completion of a surgical oncology or transplant surgery fellowship.3 A fourth pathway—completion of a HPB surgery fellowship—has been developed over the last decade, in response to increasing trends in subspecialization.4 Among these pathways, there may be significant variation in terms of programmatic emphases and operative exposure. Surgical oncology fellowship programs provide trainees with comprehensive exposure to cancer care, with rotations in medical and radiation oncology.5, 9 However, experience with complex upper GI cases, particularly for benign disease, is sometimes limited, and can be uneven across various training programs. Likewise, transplant fellowships equip graduates with excellent technical liver experience, but may not provide as comprehensive pancreas- or cancer-related exposure. HPB fellowships were created, in part, to provide a more comprehensive middle ground between these more traditional, and disparate, training curricula.6
The differences in curricula, foci, and operative experience among these three programs were highlighted in a recent consensus conference between the central governing bodies of each of the fellowship tracks providing HPB training in North America: the Society for Surgical Oncology (SSO), the American Society of Transplant Surgeons (ASTS), and the AHPBA.10 Case volume requirements across the training tracks vary significantly: HPB fellowships require 100 total procedures with at least 25 pancreas operations, whereas transplant fellowships require 50 total procedures, including 15 pancreas operations, and surgical oncology training programs mandate 240 total cancer operations irrespective of organ system.10 In North America, HPB fellows must pass a rigorous certificate process (devoid of a formal high stakes summative examination), with only 84% of fellows meeting requirements for the certificate over the past decade. Standards and structure of HPB fellowships are less well defined in other areas of the world. For instance, while not formalized structurally, in Europe there is a certification examination trainees must pass. The SSO also requires fellows to pass both qualifying and certifying examinations in order to complete the fellowship training process. The ASTS has recently introduced a certificate process with a certificate in HB or HPB surgery, but there is no required written exam.10 Each path offers different curricula (some formalized, some not) to achieve these endpoints.
Few studies have examined the influence of training pathway on the practice of HPB surgery. Warner et al. (2015) surveyed approximately 400 HPB surgeons about their fellowship and post-fellowship case volumes relating to pancreas and hepatobiliary procedures.3 They found that AHPBA and SSO training paradigms provided greater exposure to pancreas cases, which translated into more robust post-fellowship pancreas case volumes as compared to ASTS trained surgeons. No significant differences in post-fellowship hepatobiliary case volume were seen among the three training pathways.3 As part of a larger assessment of the HPB workforce in North America, Minter and colleagues (2015) also analyzed self-reported case volume stratified by training pathway. These authors found that completion of an SSO- or AHPBA-sanctioned fellowship correlated with a majority-HPB practice case mix, as compared to training in the ASTS paradigm.4
The present study builds upon the work of these prior investigations by examining the influence of fellowship training on practice patterns within a specific operation—pancreatoduodenectomy. PD was chosen as a benchmark for this study because it is a complex operation defined by a series of intraoperative strategies and management decisions that have been shown to vary significantly among pancreas surgeons.11 Furthermore, many of these management strategies have been studied in randomized controlled trials (RCTs), providing Level I evidence for best clinical practices in PD. Therefore, evaluating the practice of PD offer insight into what influence fellowship training may have on the implementation of best clinical practice in HPB surgery.
This investigation found significant variation in the demographics and clinical experience of surgeons based on fellowship training. Globally, surgeons practicing pancreatic surgery in North America, Europe, and Asia were far more likely than their colleagues in South America or Africa to complete fellowship training of any kind. This trend reflects both the differential availability of fellowships in these regions, as well as the Eurocentric and Americentric historical roots of these fellowship training paradigms.9, 13, 14 Fellowship-trained surgeons were also younger and less experienced than non-fellowship-trained surgeons, demonstrating the relatively recent proliferation of formal fellowship training programs, as well as the growing trend toward subspecialization in general surgery.1 Respondents with HPB fellowship training were significantly more likely to have a high volume (≥20) annual PD practice, which supports prior studies of the HPB surgical workforce.3, 4 However, fellowship training of any kind did not correlate with surgeons' attainment of the PD learning curve milestones (≥50 or >60) reported by Schmidt et al.7 and Tseng et al.8, respectively. This lack of correlation may be explained by the relatively recent introduction of fellowship programs, whose graduates have high annual volume but have not yet had time to achieve the learning curve.
The specific operative approaches within the practice of PD were also examined, focusing on the influence of fellowship training on best practices established by RCTs. The method of pancreaticoenteric reconstruction during PD has been the subject of a number of RCTs and meta-analyses in recent years, with most studies supporting the superiority of the PG anastomotic technique.15, 16 However, these data are far from convincing and, despite this evidence, nearly 90% of respondents preferred PJ to PG regardless of fellowship training. Interestingly, a large, multicenter RCT—the results of which were published after completion of the data acquisition for this study—supports the preference of the respondents. The RECOPANC trial demonstrated higher rates of postoperative bleeding events with PG and equivalent rates of Grade B/C postoperative pancreatic fistula (POPF) between the two anastomotic techniques.17
Other aspects of the pancreaticoenteric anastomosis analyzed in the present study include type of suture used and utilization of the dunking/invagination technique. This analysis found that HPB and surgical oncology training was associated with a tendency to use absorbable suture in the outer layer of the anastomosis. However, in the lone study assessing suture type in PD, Andrianello and colleagues found that non-absorbable outer layer sutures were associated with fewer Grade C POPFs and reduced incidence of severe post-pancreatectomy hemorrhage.18 With regard to the dunking/invagination technique generally employed for small pancreatic ducts, the current study found that HPB trained surgeons were less apt to employ this anastomotic technique despite evidence for its benefit. Indeed, Berger et al. demonstrated in a dual-institution RCT that invagination of the pancreatic remnant into the jejunum reduced the rate of POPF as compared to the standard duct-to-mucosa anastomosis.19
The present study also evaluated surgeons' use of other operative techniques that putatively reinforce pancreatic anastomotic construction including trans-anastomotic stents, biologic sealants, and autologous tissue patches. Respondents who had completed HPB fellowship were more likely to routinely use internal stents; however, Level I evidence has shown no benefit to internal stents as compared with no stents.20 Indeed, only external stents have been shown in RCTs to reduce the incidence of POPF, overall morbidity, and hospital stay.21 HPB-trained surgeons also displayed a strong predilection for the use of both autologous round ligament tissue patches and biologic sealants in solidifying the pancreaticoenteric anastomosis. No Level I evidence exists for the use of autologous tissue patches,22 and the RCTs investigating the use of biologic sealants have not shown any reduction in POPF incidence.23, 24
Significant variation by fellowship training in surgeons' postoperative drain management strategies was also evident. Though prophylactic intraperitoneal drain placement was routine practice for most respondents, irrespective of fellowship training, HPB-trained surgeons were less likely to always employ this management approach. Level I evidence backs up this practice, with a multi-institutional RCT demonstrating a benefit to the selective use of drains, particularly in those patients with moderate or high risk for an anastomotic leak based upon Fistula Risk Score (FRS) assessment.25, 26, 27 Recent studies have also supported the analysis of early (POD ≤ 3) amylase values in guiding drain removal and mitigating fistula risk.28, 29, 30 Despite this evidence, less than half of pancreas surgeons in the present study reported employing this management strategy. Only HPB trained respondents were significantly more likely to rely upon early drain amylase values in guiding drain removal.
The final management decision examined was prophylactic somatostatin use—one of the more controversial fistula mitigation strategies in the practice of PD. Prophylactic somatostatin analogues are never administered by over 40% of pancreatic surgeons and always applied by approximately 13% of surgeons, regardless of whether they have completed fellowship training. HPB-trained surgeons show a strong preference for the use of somatostatin analogs, while surgical oncologists are far less likely to administer them on a routine basis. There are data from multiple RCTs that support each approach. Octreotide, the most thoroughly scrutinized analog, has been the subject of eight high-accrual RCTs with conflicting results. Although earlier studies showed benefit to this fistula mitigation strategy, contemporary investigations have suggested that octreotide may potentiate the development of clinically relevant POPF in patients with certain risk factors.31 Further clouding the picture is a recently published single-institution RCT in which Allen and colleagues described a marked reduction in POPF with the use of pasireotide, a newer somatostatin analog.32 Thus, no consensus exists in the literature regarding this management strategy and, with this void, it appears that fellowship training may have a great influence in a surgeon's choice to use somatostatin analogs.
Employment of the above fistula mitigation strategies has been increasingly dependent upon fistula risk stratification. A highly validated tool for measuring aggregate fistula risk is the Fistula Risk Score (FRS), a 10-point scale based upon gland texture, pathology, pancreatic duct diameter, and intraoperative blood loss.25 A recent multicenter, prospective clinical trial found that risk stratification using the FRS can be used to identify low-risk patients for whom drains can be safely omitted, as well as those patients that will likely benefit from early drain removal.27 Although completion of any fellowship was not associated with utilization of the FRS to guide clinical care, HPB fellowship-trained surgeons were significantly more likely to use the FRS in their clinical practice.
This study has several limitations that warrant consideration. Most notably, this survey instrument relies upon surgeons to accurately self-report their operative experience and practice of PD. Particularly, the operative volumes, both annual and career total, could be subject to overestimation by respondents. The survey population was also limited to members of surgical societies, which could conceivably miss a sizable number of surgeons practicing pancreas surgery. It is likely, however, that these are practitioners who perform PD with lower volumes in the scope of a general surgical practice. Furthermore, this study included pancreas surgeons around the world, rather than limiting the sample to a specific region of the world which may have more consistent practice patterns, such as the prior investigations performed in North America. Whereas the three focused training paradigms discussed (HPB, surgical oncology, and transplant) have codified training standards and curricula in certain regions of the world, they are less well defined in other regions and introduce a level of heterogeneity to this data set (i.e. what constitutes a transplant fellowship in Canada may not be perfectly equivalent to a transplant fellowship in South Africa). As the first global study of the influence of fellowship training on the practice of HPB surgery, the broad geographic inclusion criteria is a strength of the present study. Nonetheless, it could be argued that the trends reported in the practice of PD may be reflective of known regional variations in PD practice,11 rather than a direct consequence of fellowship training. However, when the respondents were stratified by region, the same trends in fellowship training and PD practice as described above held true (data not shown).
In summary, there appear to be significant variations in the demographic profile, operative experience, and practice patterns between fellowship and non-fellowship-trained pancreas surgeons around the world. Some of these findings are not surprising, and self-fulfilling, given certain temporal facts. This study also demonstrated heterogeneity in the implementation of various operative and postoperative fistula mitigation strategies, which correlated significantly with fellowship training background. While many of these operative decisions were supported by Level I evidence, other practices reported by surgeons from various fellowship backgrounds were contradicted by the highest level evidence.
In light of the heterogeneity of HPB surgery training paradigms, there has been a growing call for establishment of common training and certification standards for HPB surgeons in North America.14, 33 Ongoing global collaborations through organizations such as the IHPBA can provide a forum for stakeholders and surgeons around the world to evaluate and modify current training in pancreatic surgery. It is hoped that the present study—demonstrating significant variations in clinical practice among graduates of different training pathways—can add further insight to this ongoing conversation and help ensure a competent, standardized HPB workforce in the coming years.
Funding
None.
Conflicts of interest
None declared.
Footnotes
This study was presented at the 12th World Congress of the International Hepato-Pancreato-Biliary Association, 22 April 2016, Sao Paulo, Brazil.
References
- 1.Borman K.R., Vick L.R., Biester T.W. Changing demographics of residents choosing fellowships: longterm data from the American Board of Surgery. J Am Coll Surg. 2008;206:782–788. doi: 10.1016/j.jamcollsurg.2007.12.012. [DOI] [PubMed] [Google Scholar]
- 2.Sachs T.E., Ejaz A., Weiss M., Spolverato G., Ahuja N., Makary M.A. Assessing the experience in complex hepatopancreatobiliary surgery among graduating chief residents: is the operative experience enough? Surgery. 2014;156:385–393. doi: 10.1016/j.surg.2014.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Warner S.G., Alseidi A.A., Hong J. What to expect when you're expecting a hepatopancreatobiliary surgeon: self-reported experiences of HPB surgeons from different training pathways. HPB. 2015;17:785–790. doi: 10.1111/hpb.12430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Minter R.M., Alseidi A., Hong J.C. Training in hepatopancreatobiliary surgery: assessment of the hepatopancreatobiliary surgery workforce in North America. Ann Surg. 2015;262:1065–1070. doi: 10.1097/SLA.0000000000001096. [DOI] [PubMed] [Google Scholar]
- 5.Schweitzer R.J., Edwards M.H., Lawrence W. Training guidelines for surgical oncology. Cancer. 1981;48:2336–2340. doi: 10.1002/1097-0142(19811115)48:10<2336::aid-cncr2820481033>3.0.co;2-t. [DOI] [PubMed] [Google Scholar]
- 6.Rassadi R., Dickerman R.M., Dunn E.L. Hepatopancreaticobiliary (HPB) surgery: what is the right fellowship for the right training? J Surg Educ. 2008;65:186–190. doi: 10.1016/j.jsurg.2007.11.012. [DOI] [PubMed] [Google Scholar]
- 7.Schmidt C.M., Turrini O., Parikh P. Effect of hospital volume, surgeon experience, and surgeon volume on patient outcomes after pancreaticoduodenectomy: a single-institution experience. Arch Surg. 2010;145:634–640. doi: 10.1001/archsurg.2010.118. [DOI] [PubMed] [Google Scholar]
- 8.Tseng J.F., Pisters P.W., Lee J.E. The learning curve in pancreatic surgery. Surgery. 2007;141:694–701. doi: 10.1016/j.surg.2007.04.001. [DOI] [PubMed] [Google Scholar]
- 9.Balch C.M., Coit D.G., Berman R.S. 2015 James Ewing lecture: the 75-year history of the Society of Surgical Oncology – Part II: the transitional years (1966–1990) Ann Surg Oncol. 2016;23:358–364. doi: 10.1245/s10434-015-4863-y. [DOI] [PubMed] [Google Scholar]
- 10.Jeyarajah D.R., Berman R.S., Doyle M. Consensus conference on North American training in hepatopancreaticobiliary surgery: a review of the conference and presentation of consensus statements. Am J Transplant. 2016;16:1086–1093. doi: 10.1111/ajt.13675. [DOI] [PubMed] [Google Scholar]
- 11.McMillan M.T., Malleo G., Bassi C. Defining the practice of pancreatoduodenectomy around the world. HPB. 2015;17:1145–1154. doi: 10.1111/hpb.12475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Stitzenberg K.B., Sheldon G.F. Progressive specialization within general surgery: adding to the complexity of workforce planning. J Am Coll Surg. 2005;6:925–932. doi: 10.1016/j.jamcollsurg.2005.06.253. [DOI] [PubMed] [Google Scholar]
- 13.Kaufman D.B., Ascher N.L. Quo vadis, my transplant fellow: a discussion of transplant surgery fellowship training activity in the United States and Canada: 1991–1997. Transplantation. 1998;65:269–272. doi: 10.1097/00007890-199801270-00023. [DOI] [PubMed] [Google Scholar]
- 14.Raptis D.A., Clavien P. Evaluation of hepato-pancreato-biliary (HPB) fellowships: an international survey of programme directors. HPB. 2011;13:279–285. doi: 10.1111/j.1477-2574.2010.00283.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Menahem B., Guittet L., Mullri A. Pancreaticogastrostomy is superior to pancreaticojejunostomy for prevention of pancreatic fistula after pancreaticoduodenectomy: an updated meta-analysis of randomized controlled trials. Ann Surg. 2015;261:882–887. doi: 10.1097/SLA.0000000000000806. [DOI] [PubMed] [Google Scholar]
- 16.Liu F.B., Chen J.M., Geng W. Pancreaticogastrostomy is associated with significantly less pancreatic fistula than pancreaticojejunostomy reconstruction after pancreaticoduodenectomy: a meta-analysis of seven randomized controlled trials. HPB. 2015;17:123–130. doi: 10.1111/hpb.12279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Keck T., Wellner U.F., Bahra M. Pancreatogastrostomy versus pancreatojejunostomy for reconstruction after pancreatoduodenectomy: perioperative and long-term results of a multicenter randomized controlled trial. Ann Surg. 2016;263:440–449. doi: 10.1097/SLA.0000000000001240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Andrianello S., Pea A., Pulvirenti A. Pancreaticojejunostomy after pancreaticoduodenectomy: suture material and incidence of post-operative pancreatic fistula. Pancreatology. 2016;16:138–141. doi: 10.1016/j.pan.2015.11.004. [DOI] [PubMed] [Google Scholar]
- 19.Berger A.C., Howard T.J., Kennedy E.P. Does type of pancreaticojejunostomy after pancreaticoduodenectomy decrease rate of pancreatic fistula? A randomized, prospective, dual-institution trial. J Am Coll Surg. 2009;208:738–747. doi: 10.1016/j.jamcollsurg.2008.12.031. [DOI] [PubMed] [Google Scholar]
- 20.Winter J.M., Cameron J.L., Campbell K.A. Does pancreatic duct stenting decrease the rate of pancreatic fistula following pancreaticoduodenectomy? Results of a prospective randomized trial. J Gastrointest Surg. 2006;10:1280–1290. doi: 10.1016/j.gassur.2006.07.020. [DOI] [PubMed] [Google Scholar]
- 21.Hong S., Wang H., Yang S. External stent versus no stent for pancreaticojejunostomy: a meta-analysis of randomized controlled trials. J Gastrointest Surg. 2013;17:1516–1525. doi: 10.1007/s11605-013-2187-4. [DOI] [PubMed] [Google Scholar]
- 22.Iannitti D.A., Coburn N.G., Somberg J. Use of the round ligament of the liver to decrease pancreatic fistulas: a novel technique. J Am Coll Surg. 2006;203:857–864. doi: 10.1016/j.jamcollsurg.2006.08.021. [DOI] [PubMed] [Google Scholar]
- 23.Lillemoe K.D., Cameron J.L., Kim M.P. Does fibrin glue sealant decrease the rate of pancreatic fistula after pancreaticoduodenectomy? Results of a prospective randomized trial. J Gastrointest Surg. 2004;8:766–772. doi: 10.1016/j.gassur.2004.06.011. [DOI] [PubMed] [Google Scholar]
- 24.Martin I., Au K. Does fibrin glue sealant decrease the rate of anastomotic leak after a pancreaticoduodenectomy? Results of a prospective randomized trial. HPB. 2013;15:561–566. doi: 10.1111/hpb.12018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Callery M.P., Pratt W.B., Kent T.S. A prospectively validated clinical risk score accurately predicts pancreatic fistula after pancreatoduodenectomy. J Am Coll Surg. 2013;216:1–14. doi: 10.1016/j.jamcollsurg.2012.09.002. [DOI] [PubMed] [Google Scholar]
- 26.McMillan M.T., Fisher W.E., Van Buren G. The value of drains as a fistula mitigation strategy for pancreatoduodenectomy: something for everyone? Results of a randomized prospective multi-institutional study. J Gastrointest Surg. 2015;19:21–31. doi: 10.1007/s11605-014-2640-z. [DOI] [PubMed] [Google Scholar]
- 27.McMillan M.T., Malleo G., Bassi C. A multicenter, prospective trial of selective drain management for pancreatoduodenectomy using risk stratification. Ann Surg. June 24, 2016 doi: 10.1097/SLA.0000000000001832. Epub. [DOI] [PubMed] [Google Scholar]
- 28.Kawai M., Tani M., Terasawa H. Early removal of prophylactic drains reduces the risk of intra-abdominal infections in patients with pancreatic head resection: prospective study for 104 consecutive patients. Ann Surg. 2006;244:1–7. doi: 10.1097/01.sla.0000218077.14035.a6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bassi C., Molinari E., Malleo G. Early versus late drain removal after standard pancreatic resections: results of a prospective randomized trial. Ann Surg. 2010;252:207–214. doi: 10.1097/SLA.0b013e3181e61e88. [DOI] [PubMed] [Google Scholar]
- 30.McMillan M.T., Malleo G., Bassi C. Drain management after pancreatoduodenectomy: reappraisal of a prospective randomized trial using risk stratification. J Am Coll Surg. 2015;221:798–809. doi: 10.1016/j.jamcollsurg.2015.07.005. [DOI] [PubMed] [Google Scholar]
- 31.McMillan M.T., Christein J.D., Callery M.P. Prophylactic octreotide for pancreatoduodenectomy: more harm than good? HPB. 2014;16:954–962. doi: 10.1111/hpb.12314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Allen P.J., Gonen M., Brennan M.F. Pasireotide for postoperative pancreatic fistula. N Engl J Med. 2014;370:2014–2022. doi: 10.1056/NEJMoa1313688. [DOI] [PubMed] [Google Scholar]
- 33.Minter R.M. Hepatopancreatobiliary surgery fellowships: how many do we need? JAMA Surg. 2016;151:213–214. doi: 10.1001/jamasurg.2015.4601. [DOI] [PubMed] [Google Scholar]