Stephanie A. Savage

A ten-year review of enterocutaneous fistulas after laparotomy for trauma.

BACKGROUND In the era of open abdomen management, the complication of enterocutaneous fistula (ECF) seems to be increasing in frequency. In nontrauma patients, reported mortality rates are 7% to 20%, and spontaneous closure rates are approximately 25%. This study is the largest series of ECFs reported exclusively caused by trauma and examines the characteristics unique to this population. METHODS Trauma patients with an ECF at a single regional trauma center over a 10-year period were reviewed. Parameters studied included fistula output, site, nutritional status, operative history, and fistula resolution (spontaneous vs. operative). RESULTS Approximately 2,224 patients received a trauma laparotomy and survived longer than 4 days. Of these, 43 patients (1.9%) had ECF. The rate of ECF in men was 2.22% and 0.74% in women. Patients with open abdomen had a higher ECF incidence (8% vs. 0.5%) and lower rate of spontaneous closure (37% vs. 45%). Spontaneous closure occurred in 31% with high-output fistulas, 13% with medium output, and 55% with low output. The mortality rate of ECF was 14% after an average stay of 59 days in the intensive care unit. CONCLUSION With damage-control laparotomies, the traumatic ECF rate is increasing and is a different entity than nontraumatic ECF. Although the two populations have similar mortality rates, the trauma cohort demonstrates higher spontaneous closure rates and a curiously higher rate of development in men. Fistula output was not predictive of spontaneous closure.

Blunt cerebrovascular injury screening with 64-channel multidetector computed tomography: more slices finally cut it.

BACKGROUND Aggressive screening to diagnose blunt cerebrovascular injury (BCVI) results in early treatment, leading to improved outcomes and reduced stroke rates. While computed tomographic angiography (CTA) has been widely adopted for BCVI screening, evidence of its diagnostic sensitivity is marginal. Previous work from our institution using 32-channel multidetector CTA in 684 patients demonstrated an inadequate sensitivity of 51% (Ann Surg. 2011,253: 444–450). Digital subtraction angiography (DSA) continues to be the reference standard of diagnosis but has significant drawbacks of invasiveness and resource demands. There have been continued advances in CT technology, and this is the first report of an extensive experience with 64-channel multidetector CTA. METHODS Patients screened for BCVI using CTA and DSA (reference) at a Level 1 trauma center during the 12-month period ending in May 2012 were identified. Results of CTA and DSA, complications, and strokes were retrospectively reviewed and compared. RESULTS A total of 594 patients met criteria for BCVI screening and underwent both CTA and DSA. One hundred twenty-eight patients (22% of those screened) had 163 injured vessels: 99 (61%) carotid artery injuries and 64 (39%) vertebral artery injuries. Sixty-four–channel CTA demonstrated an overall sensitivity per vessel of 68% and specificity of 92%. The 52 false-negative findings on CTA were composed of 34 carotid artery injuries and 18 vertebral artery injuries; 32 (62%) were Grade I injuries. Overall, positive predictive value was 36.2%, and negative predictive value was 97.5%. Six procedure-related complications (1%) occurred with DSA, including two iatrogenic dissections and one stroke. CONCLUSION Sixty-four–channel CTA demonstrated a significantly improved sensitivity of 68% versus the 51% previously reported for the 32-channel CTA (p = 0.0075). Sixty-two percent of the false-negative findings occurred with low-grade injuries. Considering complications, cost, and resource demand associated with DSA, this study suggests that 64-channel CTA may replace DSA as the primary screening tool for BCVI. LEVEL OF EVIDENCE Diagnostic study, level III.

Redefining massive transfusion when every second counts.

BACKGROUND The massive transfusion (MT) concept (>10-U packed red blood cells per 24 hours) is retrospective, arbitrary, and prone to survivor bias. Accounting for rate and timing is a more accurate conceptual framework. We redefined MT as a critical administration threshold (CAT) of 3 U/h, which is clinically pertinent and reflects hemorrhagic shock. The purpose of this study was to compare the traditional form of MT to a CAT definition in predicting mortality. METHODS Patients receiving transfusion in the first 24 hours were included. Precise transfusion times for each unit, in minutes, were calculated from time of injury. MT and CAT were compared to determine risk of death. Univariate and multivariate analyses were used to examine inpatient mortality. RESULTS A total of 169 patients(70%, >10 New Injury Severity Score [NISS]) were studied; 46% were CAT+; 22% met the MT criteria. With logistic regression, a CAT of 3 U/h (CAT+) was more predictive of death compared with 2, 4, 5, or 6 U/h. CAT was met once (CAT 1), twice (CAT 2) or more than 3 times (CAT 3) in 21%, 14%, and 11%, respectively. Increasing CAT was associated with increased mortality. CAT identified 75% of all deaths; MT only identified 33% and failed to identify 42% of CAT+ deaths. CAT (relative risk [RR] 3.58; 95% confidence interval [CI] 1.80–7.15) had a stronger association with mortality compared with MT(RR, 1.82; 95% confidence interval, 1.02–3.26). CONCLUSION The traditional definition of MT is inadequate to reflect illness severity. Using CATs allows prospective identification of critically ill trauma patients and eliminates survivor bias. CAT may serve as an activation trigger for transfusion protocols, allowing early identification of patients with critical transfusion requirements. Clinical trials involving transfusion strategies should consider CAT as an instrument for evaluating outcomes. LEVEL OF EVIDENCE Diagnostic/prognostic study, level II.

The evolution of blunt splenic injury: resolution and progression.

BACKGROUND Nonoperative management of blunt splenic injury (BSI) has become the standard of care for hemodynamically stable patients. Successful nonoperative management raises two related questions: (1) what is the time course for splenic healing and (2) when may patients safely return to usual activities? There is little evidence to guide surgeon recommendations regarding return to full activities. Our hypothesis was that time to healing is related to severity of BSI. METHODS The trauma registry at a level I trauma center was queried for patients diagnosed with a BSI managed nonoperatively between 2002 and 2007. Follow-up abdominal computed tomography scans were reviewed with attention to progression to healing of BSI. Kaplan-Meier curves were compared for mild (American Association for the Surgery of Trauma grades I-II) and severe (grades III-V) BSI. RESULTS Six hundred thirty-seven patients (63.9% mild spleen injury and 36.1% severe injury) with a BSI were eligible for analysis. Fifty-one patients had documented healing as inpatients. Ninety-seven patients discharged with BSI had outpatient computed tomography scans. Nine had worsening of BSI as outpatients and two (1 mild and 1 severe) required intervention (2 splenectomies). Thirty-three outpatients were followed to complete healing. Mild injuries had faster mean time to healing compared with severe (12.5 vs. 37.2 days, p < 0.001). Most healing occurred within 2 months but approximately 20% of each group had not healed after 3 months. CONCLUSION Although mild BSIs heal faster than severe BSIs, nearly 10% of all the BSIs followed as outpatients worsened. Close observation of patients with BSI should continue until healing can be confirmed.

Diagnostic laparoscopy for the evaluation of occult diaphragmatic injury following penetrating thoracoabdominal trauma

BACKGROUND Occult diaphragmatic injury following penetrating thoracoabdominal trauma can be difficult to diagnose. Radiographic findings are often non-specific or absent. Undetected injuries may remain clinically silent, only to present later with life-threatening complications associated with diaphragmatic herniation. Diagnostic laparoscopy allows for the evaluation of trauma patients lacking clinical indications for a formal laparotomy. The purpose of this study was to evaluate the incidence of occult diaphragmatic injury and investigate the role of laparoscopy in patients with penetrating thoracoabdominal trauma who lack indications for exploratory laparotomy except the potential for a diaphragmatic injury. METHODS Haemodynamically stable patients with penetrating thoracoabdominal trauma without indications for laparotomy (haemodynamic instability, evisceration, or peritonitis on exam) and evaluated with diagnostic laparoscopy to determine the presence of a diaphragmatic injury were retrospectively reviewed. Thoracoabdominal wounds were defined as wounds bounded by the nipple line over the anterior and posterior chest superiorly and the costal margin inferiorly. RESULTS One hundred and eight patients were evaluated for penetrating thoracoabdominal injuries (80 stabs and 28 gunshots) over the study period. 22 (20%) diaphragmatic injuries were identified. These were associated with injuries to the spleen (5), stomach (3) and liver (2). There was a greater incidence of haemopneumothorax (HPTX) in patients with diaphragmatic injury (32%) compared to those without injury (20%). 29% of patients with a HPTX had a diaphragmatic injury. However, 18% of patients with a normal chest radiograph were also found to have a diaphragmatic injury. CONCLUSIONS The incidence of diaphragmatic injury associated with penetrating thoracoabdominal trauma is high. Clinical and radiographic findings can be unreliable for detecting occult diaphragmatic injury. Diagnostic laparoscopy provides a vital tool for detecting occult diaphragmatic injury among patients who have no other indications for formal laparotomy.

The American association for the surgery of trauma prospective observational vascular injury treatment (PROOVIT) registry: Multicenter data on modern vascular injury diagnosis, management, and outcomes

BACKGROUND There is a need for a prospective registry designed to capture trauma-specific, in-hospital, and long-term outcomes related to vascular injury. METHODS The American Association for the Surgery of Trauma PROspective Vascular Injury Treatment (PROOVIT) registry was used to collect demographic, diagnostic, treatment, and outcome data on vascular injuries. RESULTS A total of 542 injuries from 14 centers (13 American College of Surgeons–verified Level I and 1 American College of Surgeons–verified Level II) have been captured since February 2013. The majority of patients are male (70.5%), with an Injury Severity Score (ISS) of 15 or greater among 32.1%. Penetrating mechanisms account for 36.5%. Arterial injuries to the head/neck (26.7%), thorax (10.4%), abdomen/pelvis (7.8%), upper extremity (18.4%), and lower extremity (26.0%) were identified, along with 98 major venous injuries. Hard signs of vascular injury, including hypotension (systolic blood pressure < 90 mm Hg, 11.8%), were noted in 28.6%. Prehospital tourniquet use for extremity injuries occurred in 20.2% (47 of 233). Diagnostic modalities included exploration (28.8%), computed tomographic angiography (38.9%), duplex ultrasound (3.1%), and angiography (10.7%). Arterial injuries included transection (24.3%), occlusion (17.3%), partial transection/flow limiting defect (24.5%), pseudoaneurysm (9.0%), and other injuries including intimal defects (22.7%). Nonoperative management was undertaken in 276 (50.9%), with failure in 4.0%. Definitive endovascular and open repair were used in 40 (7.4%) and 126 (23.2%) patients, respectively. Damage-control maneuvers were used in 57 (10.5%), including ligation (31, 5.7%) and shunting (14, 2.6%). Reintervention of initial repair was required in 42 (7.7%). Amputation was performed in 7.7% of extremity vascular injuries, and overall hospital mortality was 12.7%. Follow-up ranging from 1 month to 7 months is available for 48 patients via a variety of modalities, with reintervention required in 1 patient. CONCLUSION The PROOVIT registry provides a contemporary picture of the management of vascular injury. This resource promises to provide needed information required to answer questions about optimal diagnosis and management of these patients—including much needed long-term outcome data. LEVEL OF EVIDENCE Epidemiologic study, level V.

The American Association for the Surgery of Trauma grading scale for 16 emergency general surgery conditions: Disease-specific criteria characterizing anatomic severity grading.

Abstract The American Association for the Surgery of Trauma (AAST) Committee on Patient Assessment has previously published a uniform system for describing anatomic severity for Emergency General Surgery (EGS) diseases and applied these to 16 specific EGS disease processes: appendicitis, breast infections, acute cholecystitis, acute diverticulitis of the colon, esophageal perforation, hernias (internal or abdominal wall), infectious colitis, intestinal obstruction, intestinal arterial ischemia of the bowel, acute pancreatitis, pelvic inflammatory disease, perforated peptic ulcer disease, perirectal abscess, pleural space infection, skin and soft tissue infections, and surgical site infections. Standardized definitions of categorizing these diseases will be essential for risk adjustment and comparing patient outcomes among different centers. We now report on the final construct for the data dictionaries including clinical, imaging, operative, and pathologic criteria to correspond with each grade of each EGS disease. The data dictionaries are based on review of the literature, examination of existing grading systems, and discussion with expert consensus. Level V - Expert opinion STUDY TYPE: Current Opinion or Special Report.

Application of a uniform anatomic grading system to measure disease severity in eight emergency general surgical illnesses.

BACKGROUND Emergent general surgical diseases encompass a broad spectrum of anatomy and pathophysiology, creating challenges for outcomes assessment, research, and surgical training. The goal of this study was to measure anatomic disease severity for eight emergent general surgical diseases using the uniform grading system of the American Association for the Surgery of Trauma (AAST). METHODS The Committee on Patient Assessment and Outcomes of AAST applied the previously developed uniform grading system to eight emergent general surgical diseases using a consensus of experts. It was then reviewed and approved by the Board of Managers of AAST. RESULTS Severity grades for eight commonly encountered emergent general surgical conditions were created: breast abscess, esophageal perforation, infectious colitis, pelvic inflammatory disease, perirectal abscess, pleural space infections, soft tissue infections, and surgical site infections. The range of grades from I through V, reflect progression from mild disease, limited to within the organ itself, to widespread severe disease. CONCLUSION This article provides a uniform grading system for measuring anatomic severity of eight emergent general surgical diseases. Consistent adoption of these grades could improve standardization for quality assurance, outcomes research, and surgical training.

The new metric to define large-volume hemorrhage: Results of a prospective study of the critical administration threshold

BACKGROUND Definitions of massive transfusion (MT), 10 or more units of packed red blood cells (PRBCs) in 24 hours, focus on static volumes over fixed times. This arbitrary volume definition promotes survivor bias and fails to identify the “massively” transfused patient. In previous work, the critical administration threshold (CAT) was created to incorporate both rate and volume of transfusion. CAT proved a superior predictor of mortality compared with traditional MT. The purpose of this study was to prospectively validate CAT in a larger trauma population. METHODS Patients receiving at least 1 U of PRBCs within the first day of admission were identified prospectively. Administration time of each unit of PRBCs was recorded in minutes. CAT status, defined as receipt of at least 3 U of blood in a 60-minute period, was identified for the first 24 hours. CAT+ patients were quantified by the number of times CAT+ was reached, that is, once (CAT1), twice (CAT2), three times (CAT3), or 4 or more times (CAT4). A multivariable Cox proportional hazard model with a time-varying covariate was used to quantify a patient’s risk of death with increasing CAT status. RESULTS A total of 316 met inclusion criteria, 161 of whom were CAT+. Seventy-six percent were male, mean age was 38 years, and median Injury Severity Score (ISS) was 15. CAT+ was associated with a twofold increased risk of death (hazard ratio, 1.809; 95% confidence interval, 1.020–3.209). Ninety-one patients were CAT+ and received less than 10 U of blood, thereby MT− (CAT+/MT−). CAT+/MT− had significant injury patterns, with a median ISS of 14, 43% penetrating injury, and 10% mortality. CONCLUSION CAT allows early identification of injured patients at greatest risk of death. Encompassing both rate and volume of transfusion, CAT is a tool more sensitive than common MT definitions. Studies examining large-volume blood transfusions should use CAT to identify patients, to accurately identify cohorts of interest. LEVEL OF EVIDENCE Diagnostic tests, level II.

Early tracheostomy in trauma patients saves time and money

INTRODUCTION Patients suffering traumatic brain and chest wall injuries are often difficult to liberate from the ventilator yet best timing of tracheostomy remains ill-defined. While prior studies have addressed early versus late tracheostomy, they generally suffer from the use of historical controls, which cannot account for variations in management over time. Propensity scoring can be utilized to identify controls from the same patient population, minimizing impact of confounding variables. The purpose of this study was to determine outcomes associated with early versus late tracheostomy by application of propensity scoring. METHODS Patients requiring intubation within 48h and receiving tracheostomy from January 2010 to June 2012 were identified. Early tracheostomy (ET) was a tracheostomy performed by the fifth hospital day. ET patients were matched to late tracheostomy patients (LT, tracheostomy after day 5) using propensity scoring and compared for multiple outcomes. Cost for services was calculated using average daily billing rates at our institution. RESULTS One hundred and six patients were included, 53 each in the ET (mean day tracheostomy=4) and the LT (mean day tracheostomy=10) cohorts. The average age was 47 years and 94% suffered blunt injury, with an average NISS of 23.7. Patients in the ET group had significantly shorter TICU LOS (21.4 days vs. 28.6 days, p<0.0001) and significantly fewer ventilator days (16.7 days vs. 21.9, p<0.0001) compared to the LT group. ET patients also had significantly less VAP (34% vs. 64.2%, p=0.0019). CONCLUSION In the current era of increased health-care costs, early tracheostomy significantly decreased both pulmonary morbidity and critical care resource utilization. This translates to an appreciable cost savings, at minimum