Opinions of Practicing Surgeons on the Appropriateness of Published Indications for Use of Damage Control Surgery in Trauma Patients: An International Cross-Sectional Survey

Accepted Manuscript Opinions of Practicing Surgeons on the Appropriateness of Published Indications for Use of Damage Control Surgery in Trauma Patients: An International Cross-Sectional Survey Derek J. Roberts, MD, PhD, David A. Zygun, MD, MSc, Peter D. Faris, PhD, Chad G. Ball, MD, MSc, FACS, Andrew W. Kirkpatrick, MD, MHSc, FACS, Henry T. Stelfox, MD, PhD, Indications for Trauma Damage Control Surgery International Study Group PII:

S1072-7515(16)30213-7

DOI:

10.1016/j.jamcollsurg.2016.06.002

Reference:

ACS 8377

To appear in:

Journal of the American College of Surgeons

Received Date: 15 May 2016 Revised Date:

1 June 2016

Accepted Date: 1 June 2016

Please cite this article as: Roberts DJ, Zygun DA, Faris PD, Ball CG, Kirkpatrick AW, Stelfox HT, Indications for Trauma Damage Control Surgery International Study Group, Opinions of Practicing Surgeons on the Appropriateness of Published Indications for Use of Damage Control Surgery in Trauma Patients: An International Cross-Sectional Survey, Journal of the American College of Surgeons (2016), doi: 10.1016/j.jamcollsurg.2016.06.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Opinions of Practicing Surgeons on the Appropriateness of Published Indications for Use of Damage Control Surgery in Trauma Patients: An International CrossSectional Survey

RI PT

Derek J Roberts, MD, PhD1,2,3,4, David A Zygun, MD, MSc2,5, Peter D Faris, PhD6, Chad G Ball, MD, MSc, FACS1,4,7, Andrew W Kirkpatrick, MD, MHSc, FACS1,4, and Henry T Stelfox, MD, PhD2,3,8, Indications for Trauma Damage Control Surgery International Study Group 1

Department of Surgery, University of Calgary, Calgary, Alberta, Canada Department of Community Health Sciences, University of Calgary, Calgary, Alberta, Canada 3 Department of Critical Care Medicine, University of Calgary, Calgary, Alberta, Canada 4 Regional Trauma Program, University of Calgary and the Foothills Medical Centre, Calgary, Alberta, Canada 5 Division of Critical Care Medicine, University of Alberta, Edmonton, Alberta, Canada 6 Alberta Health Sciences Research–Research Analytics, University of Calgary and the Foothills Medical Centre, Calgary, Alberta, Canada 7 Department of Oncology, University of Calgary, Calgary, Alberta, Canada 8 Department of Medicine, University of Calgary, Calgary, Alberta, Canada

M AN U

SC

2

Disclosure Information: Nothing to disclose.

TE D

Disclosures outside the scope of this work: Dr Kirkpatrick has received funding from Kinetic Concepts Incorporated for a randomized controlled trial comparing the ABThera™ Open Abdomen Negative Pressure Therapy system and Barker’s vacuum pack temporary abdominal closure techniques. He has also received travel funding from LifeCell Corp, Syntheses, and Innovative Trauma Care.

EP

Support: This study was supported by an Alberta Innovates–Health Solutions Clinician Fellowship Award, a Knowledge Translation (KT) Canada Strategic Training in Health Research Fellowship, a KT Canada Research Stipend, and funding from the Canadian Institutes of Health Research. These funders had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the manuscript.

AC C

Presented at the Trauma Association of Canada Annual Scientific Conference, Halifax, Nova Scotia, May 2016 Members of the Indications for Trauma Damage Control Surgery International Study Group are listed in the Acknowledgement section. Correspondence Address: Derek J Roberts, MD, PhD Department of Surgery, University of Calgary 10th Floor North Tower, Foothills Medical Centre 1403-29th Street Northwest Calgary, Alberta

ACCEPTED MANUSCRIPT 2

AC C

EP

TE D

M AN U

SC

Brief Title: Indications for Damage Control Surgery

RI PT

Canada T2N 2T9 Telephone: 403-690-3652 Facsimile: 403-283-9994 E-mail: [email protected]

ACCEPTED MANUSCRIPT 3 Structured Abstract Background: Variation in use of damage control (DC) surgery across trauma centers may be partially driven by surgeon uncertainty as to when it is appropriately indicated. We sought to

RI PT

determine opinions of practicing surgeons on the appropriateness of published indications for trauma DC surgery.

Study Design: We asked 384 trauma centers in the United States, Canada, and Australasia to

SC

nominate 1-3 surgeons at their center to participate in a survey about DC surgery. We then asked nominated surgeons their opinions on the appropriateness (benefit-to-harm ratio) of 43 literature-

M AN U

derived indications for use of DC surgery in adult civilian trauma patients.

Results: In total, 232 (64.8%) trauma centers nominated 366 surgeons, of whom 201 (56.0%) responded. Respondents rated 15 (78.9%) preoperative and 23 (95.8%) intraoperative indications to be appropriate. Indications respondents agreed had the greatest expected benefit included a

TE D

temperature <34°C, arterial pH <7.2, and laboratory-confirmed (international normalized ratio/prothrombin time and/or partial thromboplastin time >1.5 times normal) or clinicallyobserved coagulopathy in the pre- or intraoperative setting; administration of >10 units of packed

EP

red blood cells; requirement for a resuscitative thoracotomy in the emergency department; and identification of a juxtahepatic venous injury or devascularized or destroyed pancreas,

AC C

duodenum, or pancreaticoduodenal complex during operation. Ratings were consistent across subgroups of surgeons with different training, experience, and practice settings. Conclusions: We identified 38 indications that practicing surgeons agreed appropriately indicated use of DC surgery. Until further studies become available, these indications constitute a consensus opinion that can be used to guide practice in the current era of changing trauma resuscitation practices.

ACCEPTED MANUSCRIPT 4 Introduction Hemorrhage is the leading cause of preventable death after injury.1 Significant blood loss is often complicated by development of a “vicious cycle” of hypothermia, acidosis, and

RI PT

coagulopathy, which has been linked with a high risk of mortality.2, 3 To prevent the onset of and/or limit the effects of this vicious cycle, surgeons have adopted damage control (DC) surgery to manage severely injured patients.4 As opposed to definitive (i.e., single-stage) surgery, DC

SC

allows the initial operation to be abbreviated after control of exsanguinating hemorrhage and/or gross contamination to allow for restoration of pre-injury physiology in the intensive care unit

M AN U

(ICU) before returning to the operating room (OR) for additional surgery.3, 4

Although widely assumed to improve survival among critically injured patients, survivors of DC surgery may suffer a number of complications (e.g., complicated ventral hernias and enteroatmospheric fistulae), long lengths of hospital and ICU stay, and reduced quality of life.5-7

TE D

Studies have also recently reported data suggesting that a variation in use of DC surgery exists across trauma centers or that the procedure may be overused.3,

4, 8

These observations are

concerning because overuse of DC surgery has been associated with increased morbidity and

EP

mortality.9, 10 Some authors have therefore suggested that clinical outcomes may improve with more selective use of DC surgery, especially with the recent advent of trauma resuscitation

AC C

practices that focus on rapid hemorrhage control, prevention and immediate correction of coagulopathy, and avoidance of over-resuscitation with crystalloid fluids (i.e., DC resuscitation).3

Variation in rates of use of surgical procedures may occur when surgeons are unsure

which treatment is best in varying clinical situations.11 We hypothesized that variation in use of DC surgery across trauma centers may be due to surgeon uncertainty as to when it is

ACCEPTED MANUSCRIPT 5 appropriately indicated.12, 13 The purpose of this study was therefore to determine the opinions of practicing surgeons on the appropriateness of a list of literature-derived candidate indications for use of DC surgery in adult civilian trauma patients. We also sought to determine if surgeons’

RI PT

decisions to perform DC surgery were influenced by whether physiological derangements significantly improve or reverse as a result of rapid surgery and resuscitation (i.e., DC resuscitation). Finally, we examined whether the above opinions/decisions varied across

SC

subgroups of surgeons with different training, experience, and practice settings. Methods

M AN U

Design:

We conducted a self-administered, electronic, cross-sectional survey of trauma centers and surgeons located in four high-income countries with similar emergency medical services.14 The study was approved by the University of Calgary Conjoint Health Research Ethics Board

Study Population:

TE D

and conducted and reported according to recommendations for performing survey research.15, 16

The population of interest included surgeons practicing in level 1, 2, or 3 trauma centers

EP

in the United States, Canada, Australia, and New Zealand who perform emergent thoracic, abdominal, and/or peripheral vascular operations on injured adults. To generate a sampling frame

AC C

of potential respondents, we surveyed the designated trauma program leader (medical director or program manager) of centers in these regions and asked them to provide the names and e-mail addresses of 1-3 surgeons that practice the above types of surgery in their center and would be qualified to participate in a survey about DC surgery. The sampling frame of American, Canadian, and Australian trauma centers was created using lists of those verified by the American College of Surgeons in 2013,17 that contributed data to the Canadian National Trauma

ACCEPTED MANUSCRIPT 6 Registry Comprehensive Data Set in 2010-2011 (with the exception of Quebec),18 or that were part of the Australian Trauma Quality Improvement Program as of August 31, 2014,19 respectively.

RI PT

Questionnaire Development and Testing:

The trauma program leader questionnaire was developed by modifying a previously validated questionnaire administered to trauma program leaders in the above countries.20 In

SC

addition to identifying potential surgeon respondents, the modified questionnaire collected information about the geographic location, accreditation/verification and academic status, and

M AN U

designated level of care of their center as well as trauma program characteristics, including numbers and characteristics of injured patients assessed per year (see Supplementary File 1 for the questionnaire). An assessment of the questionnaire’s clarity, length, and completeness was performed during semistructured interviews with five physicians or surgeons. The instrument

TE D

was pilot tested on five physicians or surgeons and two trauma program managers. We generated and reduced items for the surgeon questionnaire using results of a scoping review and content analysis and expert appropriateness rating study of the indications for DC

EP

surgery and DC interventions reported in peer-reviewed articles between 1983 and 2014.4, 12, 13, 21 The questionnaire focused on three items of interest: 1) respondent training and experience

AC C

(including their estimated operative volume in the last year); 2) opinions on the appropriateness of several published candidate indications for use of DC surgery in adult civilian trauma patients; and 3) whether respondents believed that critically injured patients with physiological derangements (hypothermia, acidosis, and/or coagulopathy) that significantly improve or reverse in the OR as a result of rapid surgery and resuscitation (i.e., DC resuscitation) were candidates for definitive closure of their injured cavity at the index operation.

ACCEPTED MANUSCRIPT 7 After removing similar or redundant indications from those reported in the literature and previously rated by the panel in the expert appropriateness rating study,4, 21 we selected 19 preand 24 intraoperative indications to include in the surgeon questionnaire. For each indication, we

RI PT

asked surgeon respondents to use a 5-point Likert scale to rate the benefit-to-harm ratio (1=significant harm, 2=harm, 3=uncertain, 4=benefit, 5=significant benefit) of conducting DC over definitive surgery in an adult requiring or currently undergoing operation for a thoracic,

SC

abdominal, and/or major peripheral vascular injury. Question stems and response formats were created, phrased, and formatted following recommendations by Dillman and colleagues.22

M AN U

We pre-tested the surgeon questionnaire by having two trauma surgeons, two intensivists/clinical scientists with expertise in survey research, and a statistician review the first and subsequent drafts; assess their length and the appropriateness and clarity of included questions and potential responses and response formats (including whether they appeared to meet

TE D

their intended purpose); and suggest how it could be improved.15 We then pilot-tested the questionnaire by administering it to five trauma or general surgeons who were asked to comment on its flow, arrangement, and administrative ease and identify redundant or poorly worded

EP

questions and responses.15 They were also asked to record the time required for completion (reported to be 7-8 minutes by all). Test-retest reliability was determined by administering the

AC C

survey to three surgeons twice, approximately 8 weeks apart, and demonstrated moderate chance-corrected agreement on Likert-scale ratings (weighted κ-statistic, 0.53). Supplementary File 2 provides the administered version of the surgeon questionnaire. Questionnaire Administration: We first administered questionnaires by e-mail to trauma program leaders in September 2014 and nominated surgeons in December 2014. E-mails explained the study purpose and

ACCEPTED MANUSCRIPT 8 invited potential respondents to participate by accessing a link to a secure, Web-based survey. Consent for study participation was inferred with questionnaire completion. We used several strategies identified in a large meta-analysis to increase response rates to

RI PT

electronic and postal questionnaires.23 These included administering personalized questionnaires, providing assurance of respondent confidentiality, and sending pre-notification and follow-up emails (at approximately 1 week, 2-3 weeks, 4-5 weeks, and 5-6 weeks) and a closing soon e-mail

SC

(at approximately 7-9 weeks) that included a deadline for response and statement that others had responded. We also informed potential respondents that study results would be shared with them

M AN U

if requested and potential surgeon respondents that they would be entered into a random draw for one of six $50.00 coffee cards upon questionnaire completion. Data Analyses:

We summarized categorical and continuous survey responses using counts (percentages)

TE D

and medians [with interquartile ranges (IQRs)] and compared these statistics across surveyed countries using Fisher’s exact and Kruskal-Wallis tests, respectively. Using respondents’ Likert scale ratings, indications were classified as appropriate (median Likert scale score=4-5, without

EP

disagreement), uncertain (median Likert scale score=3 or any median with disagreement), or inappropriate (median Likert scale score=1-2, without disagreement). Disagreement was defined

AC C

a priori by generalizing the RAND/University of California, Los Angeles Appropriateness Rating Method “classic” definition of appropriateness for a panel size of 9 (at least 3 panelists rating the indication inappropriate and at least another 3 rating it appropriate) to a larger sample (at least 33% of respondents rating the indication 1-2 and at least another 33% rating it 4-5).24 We stratified respondents’ assessments of appropriateness by eight prespecified surgeonand practice setting/trauma center-level characteristics. Surgeon characteristics included whether

ACCEPTED MANUSCRIPT 9 they had completed a trauma/acute care surgery or surgical critical care fellowship; performed greater than the median self-reported number of non-elective thoracic, abdominal, and/or peripheral vascular operations on injured patients in the last year; or had been practicing trauma

RI PT

surgery for >10 years. Trauma center characteristics included country, designated level of care, and whether it was reported to be a teaching (regularly had resident physicians on the trauma service) or high volume (assessed >650 patients with an Injury Severity Score >15 in the last

SC

year25) center or one that assessed more than the median percentage of patients with a penetrating injury in the last year.

M AN U

We used a recursive partitioning or classification tree to determine if any of the above surgeon- or practice setting/trauma center-level characteristics were associated with respondents’ decisions to perform cavitary closure in

patients with deranged physiology that

improved/reversed with DC resuscitation.26-28 Classification trees are powerful tools for

TE D

identifying the most “important” predictors among a large set of potential predictors, the best splits within important predictors, and for searching for complex interactions between predictors.27,

28

We then used cross-validation to determine the most appropriate size and

EP

complexity of the classification tree.28

Statistical analyses were conducted using Stata MP version 13.1 (Stata Corp., College

AC C

Station, TX) and R version 3.1.3 (The R Foundation for Statistical Computing, http://www.Rproject.org). All tests were two-sided, and we considered those resulting in p-values <0.05 statistically significant. Results

Response Rates:

ACCEPTED MANUSCRIPT 10 The flow of questionnaires through the study and data used to calculate response rates are presented in Figure 1. Of the 384 trauma centers sent the program leader survey, 232 provided a response (64.8% of 358 potentially eligible trauma centers). These centers nominated 366

RI PT

surgeons to survey about indications for DC surgery, of whom 201 responded (56.0% of 359 potentially eligible surgeons). As the number of centers and surgeons in New Zealand was small, they were analyzed together with Australia (i.e., as Australasia). Response rates from trauma

SC

centers [United States, 192/303 (63.4%); Canada, 24/33 (72.7%); and Australasia, 17/22 (77.3%); p=0.28] and surgeons [United States, 143/268 (53.4%); Canada, 36/56 (64.3%); and

M AN U

Australasia, 22/35 (62.9%); p=0.23] across the surveyed countries were similar. Characteristics of Participating Trauma Centers and Surgeons:

Characteristics of participating trauma centers and surgeons are presented in Table 1. About half (45.5%) of the centers were verified or accredited to provide level 1 trauma care,

TE D

38.3% level 2 care, and 12.6% level 3 care. Trauma centers in the United States reported assessing a higher median number of patients with a penetrating injury in the last year than those in Canada and Australasia (p=0.008). A total of 82.5% of surgeons in the United States, 75.0% in

EP

Canada, and 59.1% in Australasia completed a trauma/acute care surgery or surgical critical care fellowship (p<0.001). Further, 48.0% of surgeons had been practicing trauma surgery for 10 or

AC C

fewer years, 30.0% for 11-20 years, and 22.0% for >20 years. The median reported number of non-elective thoracic, abdominal, and/or peripheral vascular operations performed on injured patients by surgeons practicing in the United States was more than double that of those practicing in both Canada and Australasia (p<0.001).

ACCEPTED MANUSCRIPT 11 Appropriateness Ratings of Published Candidate Indications for Use of DC Surgery in Patients Requiring or Currently Undergoing Operative Exploration/Treatment of a Thoracic, Abdominal, and/or Major Peripheral Vascular Injury:

RI PT

The respondent surgeons assessed 15 (78.9%) of the preoperative (Table 2) and 23 (95.8%) of the intraoperative (Table 3) published candidate indications for use of DC surgery to be appropriate. There were no disagreements between respondents on ratings of any of the

SC

indications.

Preoperative indications assessed to be appropriate by the majority of respondents

M AN U

included a core body temperature <34°C, arterial pH <7.2, or international normalized ratio (INR)/prothrombin time (PT) >1.5 times normal. Those rated to have the greatest expected benefit included the finding of a core body temperature <34°C, arterial pH <7.2, and laboratoryconfirmed [INR/PT and/or partial thromboplastin time (PTT) >1.5 times normal] and/or

TE D

clinically-observed coagulopathy combined in the preoperative setting; the preoperative administration of >10 units of PRBCs; and patients that required a resuscitative thoracotomy in the emergency department (ED) or trauma bay. Preoperative indications assessed by respondents

EP

to have an uncertain appropriateness included patients with high energy blunt torso trauma, a systolic BP that was <90 mmHg once during transport to hospital or upon arrival to the ED or

AC C

trauma bay, and patients with a history of significant, pre-existing medical comorbidities. Intraoperative indications assessed to be appropriate by the majority of respondents

included an expanding or difficult to access pelvic hematoma, abdominal vascular injury and one or more major associated abdominal solid and/or hollow organ injuries, a proximal (i.e., Fullen zone I or II) superior mesenteric artery (SMA) injury, and multiple blunt or penetrating injuries spanning across more than one anatomical region or body cavity that each require surgery with

ACCEPTED MANUSCRIPT 12 or without angioembolization. Others included an estimated blood loss (EBL) >4 L, core body temperature <34°C and arterial pH <7.2 at the beginning of operation, and a systolic BP persistently <90 mmHg or core body temperature persistently <34°C during operation. Those

RI PT

rated to have the greatest expected benefit included the identification of a juxtahepatic venous injury or destroyed or devascularized pancreas, duodenum, or pancreaticoduodenal complex during operation; administration of >10 units of PRBCs across the pre- and intraoperative

SC

settings; an arterial pH persistently <7.2 during operation; the observation of a clinical coagulopathy by an attending surgeon; and the finding of a core body temperature <34°C,

M AN U

arterial pH <7.2, and laboratory-confirmed and/or clinically-observed coagulopathy combined at the beginning of or during the conduct of operation. The intraoperative indication assessed to have an uncertain appropriateness included patients with a systolic BP <90 mmHg at the beginning of operation.

TE D

Stratified Analyses of Appropriateness Ratings of Published Candidate Indications: In stratified analyses, respondents’ assessments of whether indications were appropriate or uncertain remained consistent across strata for 38 of the 43 indications, including seven for

EP

which the median Likert scale score was 5 (indicating significant benefit) among all subgroups of respondents (Figure 2). These included the administration of >10 U PRBCs in the pre- or

AC C

across the pre- and intraoperative settings; requirement for a resuscitative thoracotomy in the ED or trauma bay; identification of a juxtahepatic venous injury or devascularized or destroyed pancreas, duodenum, or pancreaticoduodenal complex during operation; and the finding of a core body temperature <34°C, arterial pH <7.2, and laboratory-confirmed and/or clinically-observed coagulopathy combined, regardless of setting. Assessments of appropriateness ranged from appropriate to unclear across strata for five indications (the occurrence of a mass casualty

ACCEPTED MANUSCRIPT 13 incident; patients with a history of significant, pre-existing medical comorbidities; a systolic BP <90 mmHg upon arrival to the ED or trauma bay; and a temperature <34°C or arterial pH <7.2 at the beginning of operation). For example, surgeons in the United States considered a mass

RI PT

casualty event to be an appropriate indication for use of DC surgery while those in Canada and Australasia were uncertain. DC Resuscitation Scenario:

SC

In total, 168 (88.9%) of the 189 surgeons who responded to the effective DC resuscitation scenario believed that critically injured patients with physiologic derangements that significantly

M AN U

improve or reverse in the OR as a result of surgery and resuscitation were candidates for definitive closure of their injured cavity at the index operation. Cross-validation of the developed classification tree suggested that there were no stable surgeon- or practice setting/trauma centerlevel predictors of the decision to close the affected injured cavity in this scenario.

TE D

Discussion

In this study, we identified 15 pre- and 23 intraoperative indications that were recommended or supported by studies reported in the peer-reviewed literature4,

13, 21, 29

and

EP

independently assessed by a large cohort of surgeons practicing in four high-income countries to appropriately indicate use of DC surgery in adult civilian trauma patients. The indications most

AC C

consistently rated to be appropriate by respondents were based on the finding of severely deranged patient physiology in the pre- or intraoperative setting, the intraoperative estimation of a large volume blood loss or requirement for massive transfusion or an ED resuscitative thoracotomy, and the identification of one of six highly lethal injury patterns during operation (see Table 4 for a summary). We also identified a number of indications recommended in the literature that were assessed by respondents to have an uncertain appropriateness.

ACCEPTED MANUSCRIPT 14 Nearly all respondents agreed that DC surgery was appropriate when an adult requiring emergent operation was found to have hypothermia, acidosis, and coagulopathy in the pre- or intraoperative setting. They also all agreed that the procedure was indicated when patients

RI PT

demonstrated persistent hypothermia or acidosis alone during operation. Finally, most indicated that patients with physiologic derangements that significantly improve or reverse during operation were candidates for definitive closure of their injured cavity at the index operation.

SC

These findings collectively suggest that practicing surgeons may believe that unless physiologic derangements are persistent during operation, or have already progressed to a point at which they

M AN U

are unlikely to improve (e.g., severe physiologic derangement or presence of the entire lethal triad upon presentation to the trauma center), that it is likely safe to attempt a definitive operation. In other words, as long as the other clinical scenarios listed in Table 4 have not occurred, if the core body temperature of patients undergoing operation is measured to be >34°C,

TE D

arterial pH >7.2, and INR/PT or PTT <1.5 times normal after hemorrhage and contamination have been definitively controlled, many surgeons likely believe it is appropriate to avoid use of DC surgery and perform definitive surgery.

EP

Many authors have suggested that patient survival may be improved if the decision to perform DC surgery is made early and before the patient develops hypothermia, acidosis, and

AC C

coagulopathy.4 In this study, the indications that were independent of these signs and assessed to be appropriate by the majority of respondents included an EBL >4 L, the administration of >10 U of PRBCs, and the identification of six different injury patterns during operation. These injury patterns include those that often result in exsanguination during attempts at definitive repair (an expanding and difficult to access pelvic hematoma, juxtahepatic venous injury, or Fullen zone I or

II

SMA

injury),

are

complicated

by

competing

management

priorities

or

ACCEPTED MANUSCRIPT 15 multicavitary/multiregional hemorrhage (an abdominal vascular injury and at least one major associated abdominal organ injury or multiple injuries spanning across more than anatomical region or body cavity that each require surgery with or without angioembolization), or that result

pancreaticoduodenectomy

(a

devascularized

pancreaticoduodenal complex).30-32

or

destroyed

RI PT

from high velocity gunshot wounds or extensive blunt abdominal trauma and require pancreas,

duodenum,

or

SC

There was consistent agreement regarding the appropriateness of the published indications across subgroups of surgeons with different training, experience, and practice

M AN U

settings. This included seven indications (administration of >10 U PRBCs; requirement for a resuscitative thoracotomy in the ED or trauma bay; identification of a juxtahepatic venous injury or devascularized or destroyed pancreas, duodenum, or pancreaticoduodenal complex during operation; and the finding of hypothermia, acidosis, and a laboratory-confirmed and/or

TE D

clinically-observed coagulopathy) that were assessed to have a significant benefit by all subgroups of surveyed surgeons. These findings suggest that practicing surgeons have relatively consistent impressions of the appropriateness of using DC surgery in certain clinical scenarios.

EP

Therefore, other factors need to be considered to explain regional variation in rates of use of the procedure, including differences in patient mix between institutions, surgeons’ interpretations of

AC C

or use of indications at the bedside or in the OR, and/or institutional characteristics or culture. This study has several potential limitations. First, although we used several techniques

that have been shown to increase response rates to surveys of clinicians (and therefore survey study generalizability), it is possible that respondents’ opinions may have differed systematically from those who did not respond (i.e., nonresponse bias). Second, while we described surgeons’

ACCEPTED MANUSCRIPT 16 self-reported indications for DC surgery, we did not measure how they practice. Thus, their assessments may reflect idealized rather than actual practices.20 Conclusions:

RI PT

This study identified 38 indications that a large cohort of practicing surgeons with different training, experience, and practice settings agreed appropriately indicated use of DC surgery in adult civilian trauma patients. The consistent agreement between respondents suggests

SC

that uncertainty regarding when the procedure is indicated may not be the cause of regional variation in its use. The indications assessed to be appropriate by the majority of respondents

M AN U

constitute a consensus opinion that can be used to guide practice in the current era of changing

AC C

EP

TE D

resuscitation practices while studies are conducted to evaluate their impact on patient outcomes.

ACCEPTED MANUSCRIPT 17 Acknowledgement: The authors thank all of the trauma program leaders and surgeons who participated in the study. They also thank Nancy Clayden and Jamie M Boyd, BSc for assistance with the creation

RI PT

and conduct of the electronic surveys as well as Rohan N Lall, MD; Bryan J Wells, MD, MSc; Stephen A Smith, MD; Sarah Lai, MD, MSc; Dianne Dyer, BN, MN, RN; and Christine Vis, BScN, RN for their feedback during testing of the survey instrument.

SC

Dr Roberts had full access to all data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Members of the Indications for Trauma

M AN U

Damage Control Surgery International Study Group include Karim Brohi, MD (Centre for Trauma Sciences, Blizard Institute, Queen Mary University of London, United Kingdom); Scott D’Amours, MD (Department of Surgery, University of New South Wales, Liverpool Hospital, Australia); Timothy C Fabian, MD (Department of Surgery, University of Tennessee Health

TE D

Science Center, Memphis, Tennessee, United States of America); Kenji Inaba, MD (Department of Surgery, University of Southern California, Los Angeles, California, United States of America); Ari K Leppäniemi, MD, PhD (Department of Surgery, Helsinki University Central

EP

Hospital, Helsinki, Finland); Ernest E Moore, MD (Department of Surgery, University of Colorado, Denver, Colorado, United States of America); Pradeep H Navsaria, MBChB, MMed

AC C

(Surg) (Trauma Centre, Department of Surgery, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa); Andrew J Nicol, MBChB, PhD (Trauma Centre, Department of Surgery, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa); and Neil Parry, MD (Divisions of General Surgery and Critical Care, Departments of Surgery and Medicine, Schulich School of Medicine and Dentistry, Western University and the Trauma Program, London Health Sciences Centre, London, Ontario, Canada).

ACCEPTED MANUSCRIPT 18 References 1.

Kauvar DS, Lefering R, Wade CE. Impact of hemorrhage on trauma outcome: an overview of epidemiology, clinical presentations, and therapeutic considerations. J

2.

RI PT

Trauma 2006;60(6 Suppl):S3-11.

Elerding SC, Aragon GE, Moore EE. Fatal hepatic hemorrhage after trauma. Am J Surg 1979;138:883-8.

Roberts DJ, Ball CG, Feliciano DV, et al. History of the innovation of damage control for

SC

3.

management of trauma patients: 1902-2016. Ann Surg 2016. In press. Roberts DJ, Bobrovitz N, Zygun DA, et al. Indications for use of damage control surgery

M AN U

4.

in civilian trauma patients: a content analysis and expert appropriateness rating study. Ann Surg 2016;263:1018-27. 5.

Miller RS, Morris JA, Jr, Diaz JJ, Jr, et al. Complications after 344 damage-control open

6.

TE D

celiotomies. J Trauma 2005;59:1365-71; discussion 1371-4.

Sutton E, Bochicchio GV, Bochicchio K, et al. Long term impact of damage control surgery: a preliminary prospective study. J Trauma 2006;61:831-4; discussion 835-6. Codner PA, Brasel KJ, Deroon-Cassini TA. Staged abdominal repairs reduce long-term

EP

7.

quality of life. Injury 2012;43:1513-6. Higa G, Friese R, O'Keeffe T, et al. Damage control laparotomy: a vital tool once

AC C

8.

overused. J Trauma 2010;69:53-9.

9.

Martin MJ, Hatch Q, Cotton B, et al. The use of temporary abdominal closure in low-risk

trauma patients: helpful or harmful? J Trauma Acute Care Surg 2012;72:601-6; discussion 606-8.

ACCEPTED MANUSCRIPT 19 10.

Harvin JA, Wray CJ, Steward J, et al. Control the damage: morbidity and mortality after emergent trauma laparotomy. Am J Surg. In press.

11.

Birkmeyer JD, Reames BN, McCulloch P, et al. Understanding of regional variation in

12.

RI PT

the use of surgery. Lancet 2013;382:1121-9.

Roberts DJ, Zygun DA, Kirkpatrick AW, et al. A protocol for a scoping and qualitative study to identify and evaluate indications for damage control surgery and damage control

13.

SC

interventions in civilian trauma patients. BMJ Open 2014;4:e005634.

Roberts DJ, Bobrovitz N, Zygun DA, et al. Indications for use of damage control surgery

M AN U

and damage control interventions in civilian trauma patients: A scoping review. J Trauma Acute Care Surg 2015;78:1187-96. 14.

Arnold JL. International emergency medicine and the recent development of emergency medicine worldwide. Ann Emerg Med 1999;33:97-103.

Burns KE, Duffett M, Kho ME, et al. A guide for the design and conduct of self-

TE D

15.

administered surveys of clinicians. CMAJ 2008;179:245-52. 16.

Bennett C, Khangura S, Brehaut JC, et al. Reporting guidelines for survey research: an

17.

EP

analysis of published guidance and reporting practices. PLoS Med 2010;8:e1001069. American College of Surgeons Committee on Trauma. Verified Trauma Centers 2013. IL,

U.S.A.:

AC C

Chicago,

American

College

of

Surgeons,

2013.

Available

at:

https://http://www.facs.org/search/trauma-centers?country=United States&distance=any.

Accessed August 29, 2013.

18.

Canadian Institute for Health Information. National Trauma Registry Report 2013: Hospitalizations for Major Injury in Canada, 2010-2011 Data 2013. Ottawa, ON, Canada: Canadian

Institute

for

Health

Information,

2013.

Available

at:

ACCEPTED MANUSCRIPT 20 https://secure.cihi.ca/free_products/NTR_Annual_Report_2013_EN.xls.

Accessed

August 22, 2013, 2013.. 19.

Health A. Caring for the Severely Injured in Australia: Inaugural Report of the Australian

20.

RI PT

Trauma Registry 2010 to 2012. Alfred Health, 2014.

Stelfox HT, Straus SE, Nathens A, et al. Trauma center quality improvement programs in the United States, Canada, and Australasia. Ann Surg 2012;256:163-9.

Roberts DJ, Bobrovitz N, Zygun DA, et al. Indications for use of thoracic, abdominal,

SC

21.

pelvic, and vascular damage control interventions in trauma patients: A content analysis

22.

M AN U

and expert appropriateness rating study. J Trauma Acute Care Surg 2015;79:568-79. Dillman DA, Smyth JD, Christian LM. Internet, Mail, and Mixed-Mode Surveys: The Tailored Design Method. 3 ed. Hoboken, N.J., U.S.A.: John Wiley & Sons, Inc., 2009. 23.

Edwards PJ, Roberts I, Clarke MJ, et al. Methods to increase response to postal and

24.

The

TE D

electronic questionnaires. Cochrane Database Syst Rev 2009:MR000008. RAND/UCLA

appropriateness

method

user's

manual.

Available

at:

http://www.rand.org/pubs/monograph_reports/MR1269.html. Accessed 1 Jan 2014. Nathens AB, Jurkovich GJ, Maier RV, et al. Relationship between trauma center volume

EP

25.

and outcomes. JAMA 2001;285:1164-71. Breiman L, Friedman JH, Olshen RA, et al. Classification and Regression Trees.

AC C

26.

Belmont, CA: Wadsworth International, 1984.

27.

Morgan J. Technical Report No. 1. Classification and regression tree analysis. Boston University School of Public Health, Department of Health Policy & Management, 2014. Available at: http://www.bu.edu/sph/files/2014/05/MorganCART.pdf. Accessed March

17, 2015.

ACCEPTED MANUSCRIPT 21 28.

Therneau TM, Atkinson EJ, Mayo Foundation. An introduction to recursive partitioning using the RPART routines. Mayo Foundation, 2015. Available at: http://cran.rproject.org/web/packages/rpart/vignettes/longintro.pdf. Accessed March 15, 2015. Asensio JA, McDuffie L, Petrone P, et al. Reliable variables in the exsanguinated patient

RI PT

29.

which indicate damage control and predict outcome. Am J Surg 2001; 182:743-51. 30.

Buckman RF, Jr., Miraliakbari R, Badellino MM. Juxtahepatic venous injuries: a critical

31.

SC

review of reported management strategies. J Trauma 2000;48:978-84.

Asensio JA, Britt LD, Borzotta A, et al. Multiinstitutional experience with the

M AN U

management of superior mesenteric artery injuries. J Am Coll Surg 2001;193:354-65; discussion 365-6.

Asensio JA, Petrone P, Roldan G, et al. Pancreaticoduodenectomy: a rare procedure for the management of complex pancreaticoduodenal injuries. J Am Coll Surg 2003;197:937-

EP

TE D

42.

AC C

32.

ACCEPTED MANUSCRIPT

22

Royal Australasian College of Surgeons (N=17 trauma centers and N=22 surgeons)*

11 (47.8) 5 (21.7) 1 (4.4) 6 (29.1)

8 (46.1) 5 (29.4) 4 (23.5) 0 (0)

10 (43.5) 5 (21.7) 3 (13.0) 4 (17.4) 1 (4.4)

12/16 (75.0) 2/16 (12.5) 1/16 (6.3) 1/16 (6.3) 0/16 (0)

16 (69.6) 6 (26.1) 1 (4.4) 15/22 (68.2) 20/22 (90.9) 16/22 (72.7) 16/22 (72.7) 6/22 (27.3) 19 (82.6) 13 (56.5) 22/22 (100)

10 (58.8) 7 (41.2) 0 (0) 11 (64.7) 17 (100) 13 (76.5) 11 (64.7) 5 (29.4) 17 (100) 14 (82.4) 17 (100)

0.001 <0.001 NA

850 (650-1349) 376.5 (129.5-533.5) 30.5 (0-105) 6 (0-43.5) 4/18 (22.2) 5 (3.1-9.1)

1998.5 (1300-3500) 310 (220-500) 68 (20.5-400) 10 (5-30) 1/15 (6.7) 5 (3-8)

0.01 0.76 0.08 0.97 0.51 0.008

RI PT

Trauma Association of Canada (N=23 trauma centers and N=36 surgeons)*

AC C

EP

TE D

M AN U

SC

Table 1. Characteristics of Trauma Centers and Surgeons Participating in the Study Characteristic American College of Surgeons (N=184 trauma centers and N=143 surgeons)* Trauma center (N=224) Accredited/verified for treatment of, n (%) Adult patients 144 (78.3) Adult and pediatric patients 37 (20.1) Not accredited/verified – treat adult patients 0 (0) Not accredited/verified – treat adult and pediatric patients 3 (1.6) ACS designated level of adult care, n (%) Level 1 79/183 (43.2) Level 2 78/183 (42.6) Level 3 24/183 (13.1) Not accredited/verified 1/183 (0.6) Other† 1/185 (0.6) Geographic location, n (%) Urban (within a city) 105/177 (59.3) Suburban (residential area on outskirts of a city) 49/177 (27.7) Rural (outside a city) 23/177 (13.0) Teaching center‡, n (%) 94/174 (54.0) Participate in research, n (%) 128/179 (71.5) Local investigator-initiated research 115/179 (64.3) Multicenter research 88/179 (49.2) Industry-sponsored research 50/179 (27.9) Designated trauma team, n (%) 178/179 (99.4) Designated trauma service, n (%) 176/180 (97.8) Intensive care unit that admits and cares for injured patients, n 180/180 (100) (%) No. trauma patients assessed in last year§, median (IQR) Adult, any ISS score 1500 (1000-2560.5) Adult, ISS score >15 273 (143-480) Pediatric, any ISS score 90 (38-200) Pediatric, ISS score >15 9 (2-27) High volume trauma center║, n (%) 19/129 (14.7) Percentage of trauma patients assessed in last year with a 8 (5-15) penetrating injury, median (IQR)¶ Surgeons (N=201)

p Value

<0.001

<0.001

0.37 0.37 0.10

ACCEPTED MANUSCRIPT

23

AC C

EP

TE D

M AN U

SC

RI PT

Clinical training completed, n (%) General surgery residency 142/142 (100) 36 (100) 19 (86.4) 0.001 Other surgery residency# 2 (1.4) 0 (0) 2 (9.1) 0.11 Trauma/acute care surgery or surgical critical care fellowship 118 (82.5) 27 (75.0) 13 (59.1) 0.04 Other clinical fellowship** 5 (3.5) 8 (22.2) 11 (50.0) <0.001 No. years performing trauma surgery as part of practice, n (%) <5 39/142 (27.5) 9 (25.0) 7 (31.8) 6-10 27/142 (19.0) 9 (25.0) 5 (22.7) 0.41 11-15 29/142 (20.4) 2 (5.6) 2 (9.1) 16-20 16/142 (11.3) 8 (22.2) 3 (16.4) >20 31 (21.8) 8 (22.2) 5 (22.7) Estimated no. non-elective thoracic, abdominal, and/or peripheral 45 (20-80) 20 (8-32.5) 12.5 (6-25) <0.001 vascular operations performed on injured patients in last year, median (IQR) Surgeon practice setting/trauma center characteristics ACS designated level of adult care, n (%) Level 1 101 (70.6) 23 (63.9) 18/21 (85.7) <0.001 Level 2 40 (28.0) 5 (13.9) 1/21 (4.8) Level 3 1 (0.7) 3 (8.3) 0 (0) Not accredited/verified 1 (0.7) 5 (13.9) 2/21 (9.5) Teaching center (regularly has resident physicians on the 110/141 (78.0) 20 (55.6) 15 (68.2) 0.02 trauma service), n (%) Center participates in research, n (%) 122 (85.3) 36 (100) 21 (95.5) 0.01 High volume center║, n (%) 25/118 (21.2) 11/30 (36.7) 2 (9.1) 0.06 Percentage of trauma patients assessed at center in last year with a penetrating injury, n (%) <5 19/138 (13.8) 7/35 (20.0) 13 (59.1) <0.001 5-7.9 28/138 (20.3) 15/35 (42.9) 5 (22.7) 8-14.9 38/138 (27.5) 9/35 (25.7) 4 (18.2) >15 53/138 (38.4) 4/35 (11.4) 0 (0) *Denominator of responses is given if different than stated in the column heading. The number of responses in a category may be greater than the column or category total if responses are not mutually exclusive. †Others included an area trauma facility (1 United States) or NR (1 Canada). ‡Defined as a center that regularly has resident physicians on the trauma service. §The number of adult patients, any ISS score was provided by 152 American, 21 Canadian, and 14 Australasian; adult patients, ISS score >15 trauma patients by 135 American, 20 Canadian, and 15 Australasian; pediatric patients, any ISS score by 125 American, 16 Canadian, and 12 Australasian; pediatric patients, ISS score >15 by 115 American, 16 Canadian, and 13 Australasian trauma program leaders. ║ Defined as a center that assessed >650 major trauma (ISS >15) patients in the last year.25 ¶ The percentage of trauma patients with a penetrating injury was provided by 154 American, 22 Canadian, and 15 Australasian trauma program leaders.

ACCEPTED MANUSCRIPT

24 #

AC C

EP

TE D

M AN U

SC

RI PT

Other residency programs completed included cardiac surgery and family medicine by 2 surgeons practicing in the United States and vascular surgery and orthopedic trauma, general medicine by 2 surgeons practicing in Australasia. ** Other fellowship programs completed included colorectal (n=1) or vascular (n=3) surgery or hospice palliative medicine (n=1) by surgeons practicing in the United States; hepatobiliary/pancreatic (n=1) and minimally invasive surgery (n=1) or critical care medicine (n=6) by surgeons practicing in Canada; and colorectal (n=1), FRACS/general (n=5) (which for 1 surgeon included 2 years as principle trauma fellow at a level 1 trauma center), hepatobiliary/pancreatic (n=1), upper gastrointestinal/trauma (n=1), upper gastrointestinal/bariatric (n=1), and vascular (n=1) surgery and breast surgical oncology (n=1). ACS, American College of Surgeons; RACS, Royal Australasian College of Surgeons; TAC, Trauma Association of Canada; ISS, Injury Severity Scale; IQR, interquartile range; NA, not applicable;

ACCEPTED MANUSCRIPT

25

RI PT

Table 2. Respondents’ Appropriateness Ratings of Published Candidate Indications for Use of Damage Control Over Definitive Surgery in Civilian Trauma Patients Requiring Operative Exploration/Treatment of a Thoracic, Abdominal, and/or Major Peripheral Vascular Injury Respondents’ assessment of appropriateness Consistency of agreement (N=201) Median Likert scale rating (IQR)†

Candidate indication category (no. respondents providing rating)*

Inappropriate (rating=1/2) n

%

Uncertain (rating=3) n

AC C

EP

TE D

M AN U

SC

Information relayed about prehospital trauma patient findings or events The patient suffered high energy blunt torso trauma (eg secondary to a high speed MVC) 3 (3-4) 16 8.4 104 (n=191) The patient was injured by multiple high velocity GSWs involving a single body cavity 4 (3-5) 13 6.8 61 (eg the abdomen) (n=191) The patient’s systolic BP was reported to be <90 mmHg once during transport (n=191) 3 (3-3) 28 14.7 116 The patient’s systolic BP was reported to be persistently <90 mmHg during transport to 4 (4-5) 4 2.1 43 hospital (n=191) The patient was reported to have suffered a successfully-resuscitated cardiac arrest 4 (4-5) 3 1.6 37 during transport to hospital (n=191) Trauma patient primary or secondary survey findings The patient presented during a mass casualty incident (n=190) 4 (3-4) 11 5.8 77 The patient presented with a concomitant severe TBI (n=189) 4 (3-4) 9 4.8 58 The calculated ISS score of the patient was very high (n=191) 4 (3-4) 6 3.1 70 The patient had a history of significant, pre-existing medical comorbidities (eg severe 3 (3-4) 25 13.1 98 heart and/or lung disease) (n=191) The patient’s systolic BP was <90 mmHg upon arrival to the ED or trauma bay (n=191) 3 (3-4) 11 5.8 108 The patient’s preoperative systolic BP was persistently <90 mmHg (n=191) 4 (4-5) 4 2.1 27 The patient’s preoperative core body temperature was <34°C (n=191) 5 (4-5) 7 3.7 16 The patient’s preoperative arterial pH was <7.2 4 (4-5) 5 2.6 17 The patient’s preoperative INR or PT was >1.5 times normal (n=190) 4 (4-5) 8 4.2 33 The patient’s preoperative PTT was >1.5 times normal (n=189) 4 (3-5) 8 4.2 50 The patient’s preoperative INR/PT and PTT was >1.5 times normal (n=187) 4 (4-5) 7 3.7 31 The patient’s preoperative core body temperature was <34°C, arterial pH <7.2, and they 5 (5-5) 5 2.6 5 had a laboratory-confirmed (INR/PT and/or PTT >1.5 times normal) and/or clinicallyobserved coagulopathy (n=190) >10 U of PRBCs were given to the patient preoperatively 5 (4-5) 3 1.6 13 The patient received a resuscitative thoracotomy in the ED or trauma bay (n=191) 5 (4-5) 3 1.6 13 *Categories, subcategories, and candidate indications and bracketed examples are presented exactly as in the surgeon survey instrument.

Appropriate (rating=4/5)

%

n

%

54.5

71

37.2

31.9

117

61.3

60.7 22.5

47 144

24.6 75.4

19.4

151

79.1

40.5 30.7 36.7 51.3

102 122 115 68

53.7 64.6 60.2 35.6

56.5 8.5 8.4 8.9 17.4 26.5 16.6 2.6

72 160 168 169 149 131 149 180

37.7 83.8 88.0 88.5 78.4 69.3 79.7 94.7

6.8 6.8

175 175

91.6 91.6

ACCEPTED MANUSCRIPT

26 †

AC C

EP

TE D

M AN U

SC

RI PT

On the utilized Likert scale, 1 means that the expected harms of conducting damage control over definitive surgery greatly outweigh the benefits, 3 means that the expected benefits and harms are about equal, and 5 means that the expected benefits greatly outweigh the harms. Using these ratings, the indications were classified as appropriate (panel median of 4-5, without disagreement), uncertain (panel median of 3 or any median with disagreement), or inappropriate (panel median of 1-2, without disagreement), with disagreement being defined as 33% of panelists rating the indication in the 1-2 region and at least another 33% rating it in the 4-5 region. BP, blood pressure; ED, emergency department; GSW, gunshot wound; INR, international normalized ratio; ISS, Injury Severity Scale; MVC, motor vehicle collision; PRBCs, packed red blood cells; PT, prothrombin time; PTT, partial thromboplastin time; TBI, traumatic brain injury.

ACCEPTED MANUSCRIPT

27

AC C

EP

TE D

M AN U

SC

RI PT

Table 3. Respondents’ Appropriateness Ratings of Published Candidate Indications for Use of Damage Control Over Definitive Surgery in Civilian Trauma Patients Currently Undergoing Operative Exploration/Treatment of a Thoracic, Abdominal, and/or Major Peripheral Vascular Injury Respondents’ assessment of appropriateness Consistency of agreement (N=201) Median Likert Candidate indication category (no. respondents providing rating)* Inappropriate Uncertain Appropriate scale rating (rating=1/2) (rating=3) (rating=4/5) (IQR)† n % n % n % Injury pattern identified during operation The patient is found to have an expanding and difficult to access pelvic hematoma 4 (4-5) 17 9.0 21 11.2 150 79.8 (n=188) The patient is found to have a juxtahepatic venous injury (ie AAST grade V liver injury) 5 (4-5) 7 3.7 12 6.4 168 89.8 (n=187) The patient is found to have an abdominal vascular injury and a major associated 4 (4-5) 6 3.2 31 16.6 150 80.2 abdominal solid or hollow organ injury (n=187) The patient is found to have an abdominal vascular injury and two or more major 5 (4-5) 3 1.6 24 12.8 161 85.6 associated abdominal solid or hollow organ injuries (n=188) The patient is found to have a proximal (ie Fullen zone I or II) injury to the superior 4 (4-5) 14 7.5 31 16.5 143 76.1 mesenteric artery (n=188) The patient is found to have devascularization or disruption of the pancreas, duodenum, 5 (4-5) 6 3.2 9 4.8 172 92.0 or pancreaticoduodenal complex with involvement of the ampulla/proximal pancreatic duct and/or distal CBD (n=187) The patient is found to have multiple blunt or penetrating injuries spanning across more 4 (4-5) 4 2.1 31 16.5 153 81.4 than anatomical region or body cavity that each require surgery with or without angioembolization (n=188) Time required for definitive surgery An anticipated prolonged time will be required to complete definitive repair of the 4 (3-5) 5 2.7 47 25.1 135 71.8 patient’s injuries during the index operation (n=187) >90 min has already elapsed during attempts at definitive repairs of the patient’s injuries 4 (3-5) 6 3.2 48 25.8 132 71.0 during the index operation (n=186) Estimated blood loss and volume of blood products administered across the pre- and intraoperative settings The patient’s estimated blood loss is >4 L (n=187) 4 (4-5) 4 2.1 14 7.5 169 90.4 >10 U of PRBCs have been given to the patient across the pre- and intraoperative 5 (4-5) 3 1.6 9 4.8 175 93.6 settings (n=187) Degree of physiologic insult in the operating room The patient’s systolic BP was <90 mmHg at the beginning of the operation (n=187) 3 (3-4) 11 5.9 108 57.8 68 36.4 The patient’s systolic BP has been persistently <90 mmHg during the operation (n=187) 4 (4-5) 5 2.7 14 7.5 168 89.8 The patient’s core body temperature was <34°C at the beginning of the operation 4 (3-4) 9 4.8 75 40.3 102 54.8 (n=186) The patient’s core body temperature has been persistently <34°C during the operation 4 (4-5) 5 2.2 15 8.1 166 89.2

ACCEPTED MANUSCRIPT

28

AC C

EP

TE D

M AN U

SC

RI PT

(n=186) The patient’s arterial pH was <7.2 at the beginning of the operation (n=185) 4 (3-4) 8 4.3 74 40.0 103 55.7 The patient’s arterial pH has been persistently <7.2 during the operation (n=187) 5 (4-5) 3 1.6 9 4.8 175 93.6 The patient’s intraoperative INR or PT is >1.5 times normal (n=187) 4 (3-5) 2 1.1 51 27.3 134 71.7 The patient’s intraoperative PTT is >1.5 times normal (n=186) 4 (3-4) 3 1.6 59 31.7 124 66.7 The patient’s intraoperative INR/PT and PTT is >1.5 times normal (n=185) 4 (4-5) 4 2.2 36 19.5 145 78.4 The patient is noted by the attending surgeon to have a clinically-observed coagulopathy 5 (4-5) 4 2.1 5 2.7 178 95.2 during the operation (n=187) The patient’s core body temperature was <34°C and arterial pH <7.2 at the beginning of 4 (4-5) 4 2.1 336 19.3 147 78.6 the operation (n=187) The patient’s temperature was <34°C, arterial pH <7.2, and they had a laboratory5 (4-5) 6 3.2 16 8.6 165 88.2 confirmed (INR/PT and/or PTT >1.5 times normal) or clinically-observed coagulopathy at the beginning of the operation (n=187) The patient’s temperature was <34°C, arterial pH <7.2, and they had a laboratory5 (4-5) 5 2.7 7 3.7 176 93.6 confirmed (INR/PT and/or PTT >1.5 times normal) or clinically-observed coagulopathy during the conduct of the operation (n=188) * Categories, subcategories, and candidate indications and bracketed examples are presented exactly as in the surgeon survey instrument. † On the utilized Likert scale, 1 means that the expected harms of conducting damage control over definitive surgery greatly outweigh the benefits, 3 means that the expected benefits and harms are about equal, and 5 means that the expected benefits greatly outweigh the harms. Using these ratings, the indications were classified as appropriate (panel median of 4-5, without disagreement), uncertain (panel median of 3 or any median with disagreement), or inappropriate (panel median of 1-2, without disagreement), with disagreement being defined as 33% of panelists rating the indication in the 1-2 region and at least another 33% rating it in the 4-5 region. AAST, American Association for the Surgery of Trauma; BP, blood pressure; CBD, common bile duct; ED, emergency department; GSW, gunshot wound; INR, international normalized ratio; ISS, Injury Severity Scale; MVC, motor vehicle collision; PRBCs, packed red blood cells; PT, prothrombin time; PTT, partial thromboplastin time; TBI, traumatic brain injury; U, units.

ACCEPTED MANUSCRIPT 29

AC C

EP

TE D

M AN U

SC

RI PT

Table 4. Published Candidate Indications for Use of Damage Control Surgery in Adult Civilian Trauma Patients That Were Rated to be Appropriate by the Majority of Respondents Indication Degree of physiologic insult in the pre- or intraoperative settings Persistent systolic BP <90 mmHg or a successfully-resuscitated cardiac arrest during transport to hospital Persistent systolic BP <90 mmHg in the preoperative setting or during operation Preoperative core body temperature <34°C, arterial pH <7.2, or INR/PT >1.5 times normal (with or without a concomitant PTT >1.5 times normal) Core body temperature <34°C and arterial pH <7.2 at the beginning of operation Persistent core body temperature <34°C or persistent arterial pH <7.2 during operation INR/PT and PTT >1.5 times normal during operation Clinically-observed coagulopathy during operation Core body temperature <34°, arterial pH <7.2, and laboratory-confirmed (INR/PT and/or PTT >1.5 times normal) or clinically-observed coagulopathy in the preoperative setting, at the beginning of operation, or during the conduct of operation Estimated blood loss and amount or type of resuscitation provided Estimated blood loss >4 L >10 U of PRBCs were administered to the patient in the pre- or pre- and intraoperative settings Injury pattern identified during operation An expanding and difficult to access pelvic hematoma A juxtahepatic venous injury An abdominal vascular injury and at least one major associated abdominal solid or hollow organ injury A proximal (ie Fullen zone I or II) superior mesenteric artery injury Devascularization or destruction of the pancreas, duodenum, or pancreaticoduodenal complex with involvement of the ampulla/proximal pancreatic duct and/or distal CBD Multiple blunt or penetrating injuries spanning across more than one anatomical region or body cavity that each require surgery with or without angioembolization BP, blood pressure; CBD, common bile duct; INR, international normalized ratio; PT, prothrombin time; and PTT, partial thromboplastin time.

ACCEPTED MANUSCRIPT 30 Figure Legend: Figure 1. Flow of Questionnaire Responses Through the Study. Figure 2. Color Map of Respondents’ Appropriateness Ratings of Published Candidate Pre- and

RI PT

Intraoperative Indications for Use of Damage Control Surgery Stratified by Surgeon- and Trauma Center-Level Characteristics. Interpolated median values that lied halfway between 2 integers were rounded upward. Disagreement was defined as at least 33% of respondents rating

SC

the indication as 1-2 (significant harm-harm) on the Likert Scale and at least another 33% rating it 4-5 (benefit-significant benefit). ANZ, Australia and New Zealand (ie Australasia); BP, blood

M AN U

pressure; ED, emergency department; GSWs, gunshot wounds; INR, international normalized ratio; ISS, Injury Severity Scale; PRBCs, packed red blood cells; PT, prothrombin time; and PTT, partial thromboplastin time; TBI, traumatic brain injury.

TE D

List of Supplementary Files

Supplementary File 1. Administered American Version of Trauma Program Leader Questionnaire.

AC C

EP

Supplementary File 2. Administered American Version of Surgeon Questionnaire.

ACCEPTED MANUSCRIPT 31 Precis This study identified a list of published indications that practicing surgeons agreed indicated use of damage control surgery in civilian trauma patients. Until further studies become available,

AC C

EP

TE D

M AN U

SC

era of changing trauma resuscitation practices.

RI PT

these indications constitute a consensus opinion that can be used to guide practice in the current

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT