Local complications following pancreatic trauma

Gustavo Recinos, Joseph J. DuBose *, Pedro G.R. Teixeira, Kenji Inaba, Demetrios Demetriades

Los Angeles County Hospital, University of Southern California School of Medicine, Los Angeles, CA, United States

Injury, Int. J. Care Injured 40 (2009) 516–520

A R T I C L E I N F O

Article history:

Accepted 10 June 2008

Keywords:

Pancreas

Trauma

Complications

A B S T R A C T

Background: Major trauma to the pancreas is uncommon, but associated with significant overall

morbidity and mortality. A vast majority of these adverse outcomes can be attributed to the presences of

associated injuries. Among those patients who survive the initial injury, however, the subsequent

development of pancreas-related complications represents a significant source of adverse outcomes.

Methods and results: A total of 257 patients admitted from January 1996 to April 2007 were identified

from the trauma registry database at our institution. One hundred and eighty-three patients surviving

more than 48 h after admission were selected for analysis. These patients were grouped according to the

surgical management utilised to address their pancreatic injuries: either resection or operative drainage.

After exclusion of patients with associated vascular injuries, those undergoing drainage had lower rate of

associated hollow viscus injuries (51.9% vs. 69.9%; p = 0.016) and lower rates of associated solid organ

injuries (44.2% vs. 70.9%; p � 0.001). Patients undergoing drainage were noted to have a higher incidence

of pseudocyst formation (19.5% vs. 9.0%; OR: 2.47, 95% CI, 0.92–6.67; p = 0.068), but lower hospital

lengths of stay (18.7 � 18.5 vs. 33.8 � 63.5; p = 0.001). No difference in mortality was noted between the two

populations (5.7% vs. 3.0%; p = 0.700). After multivariate analysis pseudocyst formation was the only

complication that proved different between the two management groups, with patients undergoing operative

drainage more commonly developing this adverse sequela (OR: 2.93, 95% CI, 1.02–8.36; p = 0.041).

Conclusions: In the absence of vascular injury, the choice of surgical management did not affect adjusted

mortality or the overall occurrence of pancreas-related complications. Individuals treated with operative

drainage alone, however, were significantly more likely to develop a post-operative pseudocyst than their

resectional counterparts.

� 2008 Published by Elsevier Ltd.

Contents lists available at ScienceDirect

Injury

journal homepage: www.e lsev ier .com/ locate / in jury

Introduction

Trauma to the pancreas is an uncommon but potentially morbidinjury that constitutes a diagnostic and therapeutic challenge forthe trauma surgeon. Despite their rarity, these wounds areassociated with significant overall mortality and morbid-ity.1,5,12,14,22,24,27 A vast majority of these adverse outcomes canbe attributed to the presence of multiple associated organ andvascular injuries.1,2,5,9,10,13,17,22,24,28,30,31,32 Among those patientsthat survive initial injury, however, the subsequent developmentof pancreas-related complications remain an important source ofadverse outcomes.

* Corresponding author at: University of Southern California, Department of

Trauma Surgery and Critical Care, 1200 North State Street, Room 10-750, Los

Angeles, CA 9003-4525, United States. Tel.: +1 323 226 8112; fax: +1 323 226 8116.

E-mail address: jdubose@surgery.usc.edu (J.J. DuBose).

0020–1383/$ – see front matter � 2008 Published by Elsevier Ltd.

doi:10.1016/j.injury.2008.06.026

Methods

After Institutional Review Board approval, the trauma registrydatabase of the Los Angeles County + University of SouthernCalifornia Medical Center was retrospectively reviewed to identifyall trauma patients with ICD-9 codes for pancreatic injury admittedfrom January 1996 to April 2007 who underwent abdominaloperation. Patients who died within 48 h of arrival to the hospitalwere then excluded from consideration.

Patient variables collected included age, gender, mechanismof injury, systolic blood pressure (SBP), Glasgow coma score(GCS) on admission, abbreviated injury score (AIS), abdominalAIS, ISS, associated intra-abdominal injuries, and type ofoperative intervention (resection vs. drainage). The organ injuryscore for pancreas, developed by Moore et al., was used todetermine the grade of the pancreatic injury.21 The primaryoutcomes analysed were mortality, systemic complications(ARDS, Pneumonia, Renal Failure, Sepsis, DIC) and pancreas-related complications (local abscess, pancreatitis, fistula, pseu-

G. Recinos et al. / Injury, Int. J. Care Injured 40 (2009) 516–520 517

docyst formation, local haemorrhage and wound-related com-plications).

Patients were categorised according to the surgical strategyused for the management of their pancreatic injury into eitherresection or operative drainage groups. The two cohorts were thencompared for differences in baseline demographic and clinicalcharacteristics. Categorical variables were compared using x2- orFishers exact tests, and continuous variables were compared usingStudent’s t-test or Mann–Whitney rank sum of test.

After exclusion of patients with vascular injuries, differences inoutcome between the two groups were analysed using multi-variable analysis adjusting for mechanism of injury (Blunt vs.Penetrating), age (�55 vs. <55), GCS on admission (�8 vs. >8),systolic blood pressure (SBP � 90 mmHg vs. >90 mmHg), ISS (>15vs. � 15), abdominal AIS (�4 vs. <4), associated hollow viscus(small bowel, duodenal, colon) and solid organ injuries included inthe model. The adjusted odds ratio were then calculated using a95% confidence intervals and a statistically significance of <0.05.All statistical analysis was performed using SPSS for Windows�,Version 12.0 (SPSS Inc., Chicago, IL).

Results

During the study period a total of 257 patients with adequatedocumentation for diagnosis of pancreatic injury, outcomes andcomplications were identified. After exclusion of early mortalities(�48 h from admission), 183 patients remained for evaluation. Themajority of the patients were males, representing 84.2% (154/183)of the total study population. Mean age was 29 years, with a rangefrom 6 to 89 years of age. Pancreatic injuries following penetrating

Table 1Comparison of clinical and demographic characteristics according to procedure exclud

Total (n = 183) Dra

Age (years), mean � S.D. 29.3 � 12.7 28.

Age � 55 5.5% (10/183) 5.

Male 84.2% (154/183) 83.

Penetrating 75.4% (138/183) 71.

GCS � 8 3.3% (6/178) 3.

SBP < 90 8.8% (15/171) 8.

ISS mean � S.D. 21.6 � 11.1 20.

ISS > 15 72.1% (132/183) 67.

Abdomen AIS � 4 30.2% (55/182) 27.

Hollow viscus injury 59.6% (109/183) 51.

Stomach 31.1% (57/183) 26.

Small bowel 22.4% (41/183) 18.

Duodenum 10.4% (19/183) 12.

Colon 23.0% (42/183) 21.

Solid organ injury 55.7% (102/183) 44.

Abdominal vascular injury 15.8% (29/183) 16.

Aorta 0.5% (1/183) 1.

Cava 6.6% (12/182) 6.

Hepatic arteries 1.1% (2/183) 1.

Hepatic veins 0.5% (1/182) 1.

SMA 1.6% (3/183) 1.

SMV 2.7% (5/183) 1.

Portal 2.2% (4/183) 1.

Splenic vein 2.2% (4/183) 1.

Splenic artery 3.8% (7/183) 1.

Whipple 2.2% (4/183) 0.

Injury severity

Mild (grades I and II) 28.3% (50/177) 29.

Moderate (grade III) 55.9% (99/177) 54.

Severe (grades IV and V) 15.8% (28/177) 15.

S.D.: standard deviation; GCS: Glasgow coma scale; SBP: systolic blood pressure; AIS: ab

SMV: superior mesenteric vein.

trauma accounted for 75.4% (138/183) and blunt trauma for 24.6%(45/183) of pancreatic injuries (Table 1).

Associated vascular injury was documented in 74 (27.9%) of theinitially identified 256 patients. Forty-eight (60.8%) of these died,with the majority of these deaths occurring within the first 48 h(45 of 48). Among the 183 patients who survived at least 48 h afterpancreatic injury, 15.8% (29 of 183) sustained an associatedvascular injury. The mechanism of injury in this subgroup waspenetrating in 17.4% (24 of 138), and blunt in 11.1% (5/45). Themost common vascular injuries were those to the vena cava,occurring in 6.6% (12/183) of pancreatic injuries. (Table 1). Theoverall complication rate of patients with associated vascularinjury who survived at least 48 h after pancreatic injury was 76.9%(22 of 29). Pancreas-related complications occurred in 37.9% (11 of29) of these patients.

Management of pancreatic trauma

The surgical management of pancreatic injuries was per-formed by drainage alone in 56.8% (104/183) of the cases and byresection in 43.2% (79/183). Overall, patients sustaining pene-trating injuries were more commonly treated by surgical drainagealone (53.6%; 74/138), while blunt mechanisms of injury weremore commonly managed by resection (66.7%, 30/45). Among the29 patients with associated vascular injury, drainage alone of thepancreatic injury was used in 17/29 and resection was performedin 12/29. Overall, only 2.2% (4/183) of patients required a Whippleprocedure during the study period. None of these four patientshad associated vascular injury, one died. All had sustained injuriesdue to penetrating mechanisms, no difference in the surgical

ing patients that died with in the first 48 h.

inage (n = 104) Resection (n = 79) p-Value

5 � 12.4 30.4 � 13 0.322

8% (6/104) 5.1% (4/79) 1.000

7% (87/104) 84.8% (67/79) 0.832

2% (74/104) 81.0% (64/79) 0.125

0% (3/100) 3.8% (3/78) 1.000

3% (8/96) 9.3% (7/75) 0.819

7 � 11.3 22.8 � 10.8 0.197

3% (70/104) 78.5% (62/79) 0.095

2% (28/103) 34.2% (27/79) 0.309

9% (54/104) 69.6% (55/79) 0.016

0% (27/104) 30.8% (30/79) 0.082

3% (19/104) 27.8% (22/79) 0.124

5% (13/104) 7.6% (6/79) 0.281

2% (22/104) 25.3% (20/79) 0.507

2% (46/104) 70.9% (56/79) <0.001

35 (17/104) 15.2% (12/79) 0.832

0% (1/104) 0.0% (0/79) 1.000

8% (7/103) 6.3% (5/79) 0.900

0% (1/104) 1.3% (1/79) 1.000

0% (1/104) 0.0% (0/79) 1.000

9% (2/104) 1.3% (1/79) 1.000

9% (2/104) 3.8% (3/79) 0.653

0% (1/104) 3.8% (3/79) 0.317

0% (1/104) 3.8% (3/79) 0.242

9% (2/104) 6.3% (5/79) 0.242

0% (0/104) 5.1% (4/79) 0.033

4% (30/102) 26.7% (20/75) 0.689

9% (56/102) 57.3% (43/75) 0.747

7% (16/102) 16.0% (12/75) 0.955

breviated injury score; ISS: injury severity score; SMA: superior mesenteric artery;

Table 2Comparison of clinical and demographic characteristics according to procedure excluding patients that died with in the first 48 h and vascular injuries.

Total (n = 154) Drainage (n = 87) Resection (n = 67) p-Value

Age (years), mean � S.D. 29.08 � 13.2 28.2 � 12.9 30.2 � 13.6 0.367

Age � 55 6.5% (10/154) 6.9% (6/87) 6.0% (4/67) 1.000

Male 82.5% (127/154) 82.8% (72/87) 82.1% (55/67) 0.914

Penetrating 74.0% (114/154) 69.0% (60/87) 80.6% (54/67) 0.103

GCS � 8 3.3% (5/152) 3.5% (3/85) 3.0% (2/67) 1.000

SBP < 90 8.2% (12/146) 6.1% (5/82) 10.9% (7/64) 0.291

ISS mean � S.D. 21.08 � 11.5 22.6 � 11.2 19.9 � 11.7 0.162

ISS > 15 68.2% (105/154) 62.1% (54/87) 76.1% (51/67) 0.063

Abdomen AIS � 4 28.1% (43/154) 23.3% (20/86) 34.3% (23/67) 0.131

Hollow viscus injury 59.1% (91/154) 50.6% (44/87) 70.1% (74/67) 0.014

Stomach 29.2% (45/154) 23.0% (20/87) 37.3% (25/67) 0.053

Small bowel 22.7% (35/154) 18.4% (16/87) 28.4% (19/67) 0.143

Duodenum 10.4% (16/154) 12.6% (11/87) 7.5% (5/67) 0.269

Colon 22.1% (34/154) 19.5% (17/87) 25.4% (17/67) 0.387

Solid organ injury 57.8% (89/154) 48.3% (42/87) 70.1% (47/67) 0.006

Injury severity

Mild (grades I and II) 28.9% (106/149) 30.6% (26/85) 26.6% (17/64) 0.591

Moderate (grade III) 59.1% (88/149) 56.5% (48/85) 62.5% (40/64) 0.459

Severe (grades IV and V) 12.1% (18/149) 12.9% (11/85) 10.9% (07/64) 0.710

S.D.: standard deviation; GCS: Glasgow coma scale; SBP: systolic blood pressure; AIS: abbreviated injury score; ISS: injury severity score.

G. Recinos et al. / Injury, Int. J. Care Injured 40 (2009) 516–520518

management was observed between patients with moderate orsevere pancreatic injuries.

Mortality and complications

The overall pancreas-related complication rate among survi-vors of initial pancreatic injury was 29.5% (54/183). Amongpatients in this group that had sustained a vascular injury, this ratewas 37.9% (11 of 29); compared to 27.9% (43 of 154) for patientswithout these associated injuries. Excluding patients with vasculartrauma, we found that patients who were managed by drainagealone were less likely to be severely injured (ISS > 15, 62.1% vs.76.1% p = 0.063) and had lower rates of associated hollow viscus(50.6% vs. 70.1%; p = 0.014) and solid organ injuries (48.3% vs.70.1%; p = 0.006) (Table 2). While the drainage group had a lowermean hospital LOS (18.7 vs. 33.8 days; p = 0.010) they also had ahigher rate of pseudocyst formation (19.5% vs. 9.0%; p = 0.068),when compared to patients undergoing resection. Although not

Table 3Outcomes and complications according to procedure excluding patients that died with

Total (n = 154) Drainage (n = 87

Death 4.5% (7/154) 5.7% (5/87)

Pancreas-related complications 27.9% (43/154) 31.0% (27/87)

Surgical site infection 14.9% (23/154) 11.5% (10/87)

Fistula 1.9% (5/154) 3.4% (3/87)

Dehiscence 3.2% (5/154) 1.1% (1/87)

Wound Infection 7.1% (11/154) 5.7% (5/87)

Intra-abdominal abscess 9.1% (14/154) 8.0% (7/87)

Pseudocyst 14.9% (23/154) 19.5% (17/87)

Systemic complications

ARDS 2.6% (4/154) 3.4% (3/87)

Pneumonia 4.5% (7/154) 2.3% (2/87)

Renal failure 2.6% (4/154) 3.4% (3/87)

Sepsis 3.2% (5/154) 3.4% (3/87)

DIC 1.95 (3/154) 2.3% (2/87)

Mean � S.D. (median) Mean � S.D. (median)

Hospital days 25.3 � 44.6 (15) 18.7 � 18.5 (13)

The p-values for categorical variables were derived from two-sided x2-test or Fisher’s

Mann–Whitney test.

statistically significant, a trend toward higher mortality wasobserved in patients undergoing drainage (5.7% vs. 3.0% p = 0.700)(Table 3).

After creating a multivariate logistic regression model adjustingfor all possible confounders, patients without associated vascularinjury who underwent drainage alone had higher odds ofdeveloping pseudocyst during their postoperative period (OR2.93, 95% CI, 1.02–8.36; p = 0.044) but no difference in mortality orlength of stay was observed (Table 4).

Discussion

Deaths directly attributable to pancreatic injury itself remainuncommon, occurring in approximately 3% of survivors of initialinjury.6,14,18,24,25,29,30 Pancreas-related local morbidity, however,is a much more frequent occurrence, with documented rates ashigh as 48%1,3,5,10,12–14,18,27,30 and the highest risks occurringamong patients with associated vascular injuries. 9,20 In our series,

in the first 48 h and patients with associated vascular injuries.

) Resection (n = 67) Odds ratio (95% CI) p-Value

3.0% (2/67) 1.98 (0.37–10.5) 0.700

23.9% (16/67) 1.43(0.70–2.95) 0.327

19.4% (13/67) 0.54 (0.22–1.32) 0.172

0% (0/67) 0.97 (0.92–1.00) 0.258

6.0% (4/67) 0.18 (0.02–1.68) 0.168

9.0% (6/67) 0.62 (0.18–2.13) 0.534

10.4% (7/67) 1.33 (0.44–4.00) 0.607

9.0% (6/67) 2.47 (0.92–6.67) 0.068

1.5% (1/67) 2.36 (0.24–23.2) 0.449

7.5% (5/67) 0.29 (0.05–1.55) 0.240

1.55 (1/67) 2.35 (0.24–23.2) 0.633

3.0% (3/67) 1.16 (0.19–7.15) 1.000

1.5% (1/67) 1.56 (0.14–17.5) 1.000

Mean � S.D. (median) Mean difference (95% CI) p-Value

33.8 � 63.5 (19) 15.1 (0.93–29.2) 0.010

exact test; p-values for continuous variables were derived from Student’s t-test or

Table 4Adjusted odds ratio for mortality and complications (Drainage vs. Resection).

Adjusted OR (95% CI)a Adjusted p-valuea

Deaths 2.04(0.345–12.12) 0.431

Any local complication 1.66 (0.76–3.62) 0.199

Pancreas-related complications

Surgical site infection 0.62 (0.24–1.58) 0.313

Pseudocyst 2.93 (1.02–8.36) 0.044

Mean difference (95% CI)

Hospital days 20.15 (10.57–29.73) 0.113

a Multivariable analysis adjusting for age (�55 vs. >55), mechanism, ISS (�15 vs.

>15), hollow viscus injury and solid organ injury.

G. Recinos et al. / Injury, Int. J. Care Injured 40 (2009) 516–520 519

we found an overall pancreas-related complication rate of 29.5%(54/183).

A wide variety of factors may contribute to the occurrence ofpancreas-related complications following trauma, includingmechanism, location and grade of injury. It has been previouslyreported that pancreas-related morbidities occur more frequentlyfollowing gunshot.22,24 These adverse events also appear to be morecommon in the presence of associated vascular, hollow viscus orsolid organ injuries.9,22,24 The location and grade of pancreatic injuryalso represent additional significant risk factors,2,2,6,9,11,14,16,18,19,22

although their specific role in the occurrence of pancreas-relatedcomplications has been debated.24 On initial review of our data, wenoted that vascular trauma was associated with a higher rate ofadverse outcome among patients with pancreatic trauma. Otherauthors have reported similar findings.9,22,24 For this reason, weelected to conduct our multivariate review in a population that didnot include individuals with these associated injuries.

The most commonly reported pancreas-related complicationfollowing trauma is pancreatic fistula, reported in up to 38% ofcases.1,6,8,9,10,15,18,24,29,30 In our present study, we were able todocument a fistula rate of 1.9% among patients withoutconcomitant vascular injury (Table 3). Intra-abdominal abscessis also a common complication following pancreatic trauma,occurring in up to 34% of patients surviving initialinjury;1,5,6,9,18,24,30 most frequently occurring in the setting ofassociated solid organ,1 hollow viscus,5,6,24,30 or main pancreaticduct injury.24 In our experience, this complication occurred in 9.1%of patients with pancreatic injury and no associated vasculartrauma. (Table 3).

Other reported local complications include pancreatitis(4.3–23.1%),1,5,10,15,18,30,32 pseudocyst formation (1.9–22.0%),1,5,9,10,15,18,19,30 pancreatic hemorrhage requiring re-exploration (2.8–8.5%),5,30 and sepsis (23.1%).18 In our series, wefound that pseudocyst proved the most common pancreas-relatedmorbidity following injury without concomitant vascular trauma,occurring in 14.9% of patients. Other complications includeddehiscence (3.2%), wound infection 7.1% and pancreatitis (0.6%)(Table 3).

The importance of surgical management choice in theoccurrence of pancreas-related morbidity is less welldefined.4,6,7,8,12,18,22–24,26–29,32 It is difficult to consider the useof resection vs. drainage as an independent predictive factor forthese morbidities, as in many instances the choice of interventionis dictated by the nature of the injury itself. In patients with lesssevere injury, however, several factors dictate managementchoice; including patient condition, injury location and severity,and status of ductal integrity.19,24,33

A direct comparison of drainage and resectional techniques isdifficult, as many injury specific factors likely contribute signifi-cantly to the choice of intervention. Despite this fact, several small,

retrospective comparisons have been attempted. In one series of 41patients with penetrating pancreatic injuries, Sorensen et al. notedsignificantly greater rates of pancreatic complications (52% vs. 16%,p = 0.04) in patients undergoing resection vs. drainage. In theirexamination, however, they noted that resection was morecommonly used for patients with higher grade injuries.26 Othershave suggested that the use of closed-drainage systems may morecommonly predispose patients to fistula formation than resectionalprocedures.8,24,32 Patton et al., noted that for distal injuries withindeterminant ductal status based on clinical criteria, the complica-tion rate was no different whether drainage alone or distal resectionwas used (27 vs. 33% p = 0.60).24

In our study we found that, in the absence of associated vascularinjury, the choice of surgical management following pancreatictrauma did not affect adjusted mortality or the overall occurrenceof pancreas-related complications. However, individuals treatedwith operative drainage alone were significantly more likely todevelop a post-operative psuedocyst than their resectionalcounterparts.

Our study was limited by the small number of patients and theretrospective design of our investigation. While the grade of theinjury could be ascertained from our review of medical records, thespecific location of the injury and ductal status was not clear in allcases. In addition, the specific rationale for management choicecould not be elucidated clearly. Factors which may havecontributed to these decisions may have included patientcondition, associated injuries, location of pancreatic injury andsurgeon preference. The relative contribution of these factors isdifficult to discern from a retrospective examination. It is alsoimportant to note that while the rate of pancreatic fistula in ourseries was much lower than that reported by previous investiga-tors, the absence of a uniform definition of this adverse event onour retrospective review may have resulted in under-reporting.

Our study does, however, highlight the significant rates ofpancreas-related complications associated with these injuries andcontributes to a growing body of literature suggesting that thesurgical management choice for the majority of pancreatic injuriesdoes not affect the rates of local complications. In the absence ofwell-validated guidelines, however, determining the optimalsurgical management of pancreatic trauma remains a significantsurgical challenge. A larger, prospective investigation may provequite useful in providing more definitive guidance regardingoptimal operative management of pancreatic injuries.

Conclusion

Pancreas-related complications following pancreatic traumaare common events, occurring in 29.5% of patients. Many factorsprobably contribute to the incidence of these adverse outcomes,including the choice of management technique. In our study, theuse of operative drainage alone was associated with a higher rate ofpost-operative pseudocyst formation, compared to resectionalcounterparts. The choice of operative intervention, however, didnot affect adjusted mortality or the overall occurrence of pancreas-related complications following pancreatic trauma.

Conflict of interest

None.

References

1. Akhrass R, Yaffe MB, Brandt CP, et al. Pancreatic trauma: a ten-year multi-institutional experience. Am Surg 1997;63(7):598–604.

2. Babb J, Harmon H. Diagnosis and management of pancreatic trauma. Am Surg1976;42(6):390–4.

G. Recinos et al. / Injury, Int. J. Care Injured 40 (2009) 516–520520

3. Berni GA, Bandyk DF, Oreskovich MR, et al. Role of intraoperative pancrea-tography in patients with injury to the pancreas. Am J Surg 1982;143(5):602–5.

4. Buccimazza I, Thomson SR, Anderson F, et al. Isolated main pancreatic ductinjuries spectrum and management. Am J Surg 2006;191(4):448–52.

5. Cogbill TH, Moore EE, Morris Jr JA, et al. Distal pancreatectomy for trauma: amulticenter experience. J Trauma 1991;31(12):1600–6.

6. Degiannis E, Levy RD, Potokar T, et al. Distal pancreatectomy for gunshotinjuries of the distal pancreas. Br J Surg 1995;82(9):1240–2.

7. Fabian TC, Kudsk KA, Croce MA, et al. Superiority of closed suction drainage forpancreatic trauma. A randomized, prospective study. Ann Surg 1990;211(6):724–8 [discussion 728–30].

8. Farrell RJ, Krige JE, Bornman PC, et al. Operative strategies in pancreatic trauma.Br J Surg 1996;83(7):934–7.

9. Feliciano DV, Martin TD, Cruse PA, et al. Management of combined pancrea-toduodenal injuries. Ann Surg 1987;205(6):673–80.

10. Fleming WR, Collier NA, Banting SW. Pancreatic trauma: Universities of Mel-bourne HPB Group. Aust N Z J Surg 1999;69(5):357–62.

11. Graham JM, Mattox KL, Jordan Jr GL. Traumatic injuries of the pancreas. Am JSurg 1978;136(6):744–8.

12. Hagan WV, Urdaneta LF, Stephenson Jr SE. Pancreatic injury. South Med J1978;71(8):892–4.

13. Hendel R, Rusnak CH. Management of pancreatic trauma. Can J Surg1985;28(4):359–61.

14. Jones RC. Management of pancreatic trauma. Ann Surg 1978;187(5):555–64.

15. Keeling P, Calthorpe D, Lane B, et al. Blunt injury of the neck of the pancreas: areport of nine patients. Injury 1987;18(2):93–5.

16. Leppaniemi AK, Haapiainen RK. Pancreatic trauma with proximal duct injury.Ann Chir Gynaecol 1994;83(3):191–5.

17. Levison MA, Petersen SR, Sheldon GF, et al. Duodenal trauma: experience of atrauma center. J Trauma 1984;24(6):475–80.

18. Lewis G, Knottenbelt JD, Krige JE. Conservative surgery for trauma to thepancreatic head: is it safe? Injury 1991;22(5):372–4.

19. Lin BC, Fang JF, Wong YC, et al. Blunt pancreatic trauma and pseudocyst:management of major pancreatic duct injury. Injury 2007;38(5):588–93.

20. Lopez PP, Benjamin R, Cockburn M, et al. Recent trends in the management ofcombined pancreatoduodenal injuries. Am Surg 2005;71(10):847–52.

21. Moore EE, Cogbill TH, Malangoni MA, et al. Organ injury scaling. Surg Clin NorthAm 1995;75(2):293–303.

22. Nowak MM, Baringer DC, Ponsky JL. Pancreatic injuries. Effectiveness of debride-ment and drainage for nontransecting injuries. Am Surg 1986;52(11):599–602.

23. Olah A, Issekutz A, Haulik L, et al. Pancreatic transection from blunt abdominaltrauma: early versus delayed diagnosis and surgical management. Dig Surg2003;20(5):408–14.

24. Patton Jr JH, Lyden SP, Croce MA, et al. Pancreatic trauma: a simplified manage-ment guideline. J Trauma 1997;43(2):234–9 [discussion 239–41].

25. Smego DR, Richardson JD, Flint LM. Determinants of outcome in pancreatictrauma. J Trauma 1985;25(8):771–6.

26. Sorensen VJ, Obeid FN, Horst HM, et al. Penetrating pancreatic injuries, 1978–1983. Am Surg 1986;52(7):354–8.

27. Stone HH, Fabian TC, Satiani B, et al. Experiences in the management ofpancreatic trauma. J Trauma 1981;21(4):257–62.

28. Sukul K, Lont HE, Johannes EJ. Management of pancreatic injuries. Hepatogas-troenterology 1992;39(5):447–50.

29. Teh SH, Sheppard BC, Mullins RJ, et al. Diagnosis and management of bluntpancreatic ductal injury in the era of high-resolution computed axial tomo-graphy. Am J Surg 2007;193(5):641–3 [discussion 643].

30. Vasquez JC, Coimbra R, Hoyt DB, et al. Management of penetrating pancreatictrauma: an 11-year experience of a level-1 trauma center. Injury 2001;32(10):753–9.

31. Wynn M, Hill DM, Miller DR, et al. Management of pancreatic and duodenaltrauma. Am J Surg 1985;150(3):327–32.

32. Young Jr PR, Meredith JW, Baker CC, et al. Pancreatic injuries resulting frompenetrating trauma: a multi-institution review. Am Surg 1998;64(9):838–43[discussion 843–4].

33. Zhang SH, Wang SM, Li JW. Diagnosis and treatment of pancreatic trauma. ChinJ Traumatol 2005;8(5):303–5.