Search

Nicolas J Mouawad, Stefan W Leichtle, Jeffrey V Manchio, Richard M Lampman, Brian G Halloran, Walter M Whitehouse JrMichigan Heart and Vascular Institute, Saint Joseph Mercy Health System, Ann Arbor, MI, USAPurpose: Many clinical trials comparing the outcomes of open surgical repair (OSR) versus endovascular aneurysm repair (EVAR) for abdominal aortic aneurysms (AAAs) have been conducted, with varying results. Surprisingly, few outcomes studies have closely examined perceived physical and mental health-related quality of life (HRQOL) factors through a validated survey tool. The purpose of this prospective observational study was to describe the trajectory of HRQOL measures, from baseline to 1 year after surgery, in patients undergoing OSR or EVAR for AAA, and to explore for differences in physical and mental composite scores and their construct domains (subscales) using the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36®) tool.Patients and methods: Over an 18-month period, a small sample of patients undergoing elective AAA repair in a community hospital setting were prospectively enrolled. Fifteen patients undergoing OSR and twenty patients undergoing EVAR were studied. Physical and mental HRQOL parameters were assessed using the SF-36.Results: No significant differences in demographic and clinical variables were found between the OSR and EVAR groups. In the multivariable linear models with repeated measures, both groups showed a significant decline in physical health composite scores 30 days after the surgical procedure (P < 0.01). However, although the OSR group showed a statistically significant decline in three of the four physical health domains, the EVAR group declined in only one physical health domain. Only the OSR group showed a significant decline in three of the four mental health domains at 30 days; however, the decline of these domains was not reflected in the group’s mental health composite scores. By 90 days after surgery, both groups were not significantly different from their baseline in physical or mental health composite scores, or in any of their respective physical health domains.Conclusion: In this small sample of patients undergoing AAA repair, EVAR resulted in less physical and emotional decline than OSR in the early postoperative period. However, patients in both groups may return to near baseline status at 90 days.Keywords: AAA, endovascular aneurysm repair (EVAR), open surgical repair (OSR), Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36®), health related quality of life (HRQ)

Background: Rib fractures are a common in thoracic trauma. Increasingly, patients with flail chest are being treated with surgical stabilization of rib fractures (SSRF). We performed a retrospective review of the Trauma Quality Improvement Program database to determine if there was a difference in outcomes between patients undergoing early SSRF (≤3 days) versus late SSRF (>3 days)., Methods: Patients with flail chest in Trauma Quality Improvement Program were identified by CPT code, assessing those who underwent SSRF between 2017 and 2019. We excluded those younger than 18 years and Abbreviated Injury Scale head severity scores greater than 3. Patients were grouped based on SSRF before and after hospital Day 3. These patients were case matched based on age, Injury Severity Score, Abbreviated Injury Scale head and chest, body mass index, Glasgow Coma Scale, and five modified frailty index. All data were examined using χ2, one-way analysis of variance, and Fisher's exact test within SPSS version 28.0., Results: For 3 years, 20,324 patients were noted to have flail chest, and 3,345 (16.46%) of these patients underwent SSRF. After case matching, 209 patients were found in each group. There were no significant differences between reported major comorbidities. Patients with early SSRF had fewer unplanned intubations (6.2% vs. 12.0%; p = 0.04), fewer median ventilator days (6 days Q1: 3 to Q3: 10.5 vs. 9 Q1: 4.25 to Q3: 14; p = 0.01), shorter intensive care unit length of stay (6 days Q1: 4 to Q3: 11 vs. 11 Q1: 6 to Q3: 17; p < 0.01), and hospital length of stay (15 days Q1: 11.75 to Q3: 22.25 vs. 20 Q1: 15.25 - Q3: 27, p < 0.01. Early plating was associated with lower rates of deep vein thrombosis and ventilator-acquired pneumonia., Conclusion: In trauma-accredited centers, patients with flail chest who underwent early SSRF (<3 days) had better outcomes, including fewer unplanned intubations, decreased ventilator days, shorter intensive care unit LOS and HLOS, and fewer DVTs, and ventilator-associated pneumonia., Level of Evidence: Therapeutic/Care Management; Level IV., (Copyright © 2023 American Association for the Surgery of Trauma.)

(1) Background: The disease-modifying mechanisms of high-dose intravenous vitamin C (HDIVC) in sepsis induced acute respiratory distress syndrome (ARDS) is unclear. (2) Methods: We performed a post hoc study of plasma biomarkers from subjects enrolled in the randomized placebo-controlled trial CITRIS-ALI. We explored the effects of HDIVC on cell-free DNA (cfDNA) and syndecan-1, surrogates for neutrophil extracellular trap (NET) formation and degradation of the endothelial glycocalyx, respectively. (3) Results: In 167 study subjects, baseline cfDNA levels in HDIVC (84 subjects) and placebo (83 subjects) were 2.18 ng/µL (SD 4.20 ng/µL) and 2.65 ng/µL (SD 3.87 ng/µL), respectively, p = 0.45. At 48-h, the cfDNA reduction was 1.02 ng/µL greater in HDIVC than placebo, p = 0.05. Mean baseline syndecan-1 levels in HDIVC and placebo were 9.49 ng/mL (SD 5.57 ng/mL) and 10.83 ng/mL (SD 5.95 ng/mL), respectively, p = 0.14. At 48 h, placebo subjects exhibited a 1.53 ng/mL (95% CI, 0.96 to 2.11) increase in syndecan-1 vs. 0.75 ng/mL (95% CI, 0.21 to 1.29, p = 0.05), in HDIVC subjects. (4) Conclusions: HDIVC infusion attenuated cell-free DNA and syndecan-1, biomarkers associated with sepsis-induced ARDS. Improvement of these biomarkers suggests amelioration of NETosis and shedding of the vascular endothelial glycocalyx, respectively.

Purpose: Thoracic injury is a major contributor to morbidity in trauma patients. There is limited data regarding practice patterns of video-assisted thoracoscopic surgery (VATS) across trauma-accredited hospitals in the United States. We hypothesized that early VATS remains underutilized affecting patient outcomes., Method: We evaluated a cohort of patients who underwent non-urgent thoracic surgical intervention for trauma from the ACS-TQIP database in 2017 excluding patients who were discharged within 48-h or died within 72-h. We selected patients who underwent partial lung resection and decortication to assess the effect of early (day 2-5) versus late VATS. Univariate followed by multivariate regression analyses were utilized to evaluate the independent impact of timing., Results: Over 12 months, 997,970 patients were admitted to 850 trauma-accredited centers. Thoracic injury occurred in 23.5% of patients, 1% of whom had non-urgent thoracic procedures. A total of 406 patients underwent VATS for pulmonary decortication with/out partial resection, 39% were Early VATS (N = 159) compared to 61% late VATS (N = 247). Both groups had comparable demographics and comorbidities with exception of a higher ISS score in the late surgical group (17.9 ± 9.8 vs 14.9 ± 7.6, p < 0.01). The late VATS patients' group had higher rates of superficial site infection, unplanned intubation, and pneumonia. Early VATS was associated with shorter ICU stay and HLOS. Multivariate analysis confirmed the independent effect of surgical timing on postoperative complications and LOS. The conversion rate from VATS to thoracotomy was 1.9% in early group compared to 6.5%, p = 0.03. There was no difference in surgical pattern among participating facilities., Conclusion: Despite established practice guidelines supporting early VATS for thoracic trauma management, there is underutilization with less than half of patients undergoing early VATS. Early VATS is associated with improved patient outcomes., (© 2022. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany.)

Introduction: The Bowel Injury Prediction Score (BIPS) is a tool for identifying patients at risk for blunt bowel and mesenteric injury (BBMI) requiring surgery. BIPS is calculated by assigning one point for each of the following: (1) WBC ≥ 17,000, (2) abdominal tenderness, and (3) injury grade ≥ 4 (mesenteric contusion or hematoma with bowel wall thickening or adjacent interloop fluid collection) on CT scan. A total score ≥ 2 is associated with BBMI requiring surgery. We aimed to validate the BIPS as a predictor for patients with BBMIs requiring operative intervention in a multi-center prospective study., Materials and Methods: Patients were prospectively enrolled at 15 U.S. trauma centers following blunt trauma with suspicion of BBMI on CT scan between July 1, 2018 and July 31, 2019. The BIPS was calculated for each patient enrolled in the study., Results: Of 313 patients, 38% had BBMI requiring operative intervention. Patients were significantly more likely to require surgery in the presence of abdominal tenderness (OR, 3.6; 95% CI, 1.6-8.0) and CT grade ≥ 4 (OR, 11.7; 95% CI, 5.7-23.7). Patients with a BIPS ≥ 2 were more than ten times more likely to require laparotomy than those with a BIPS < 2 (OR, 10.1; 95% CI, 5.0-20.4). The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of a BIPS ≥ 2 for BBMI requiring surgery was 72% (CI 0.6-0.8), 78% (CI 0.7-0.8), 67% (CI 0.6-0.8), and 82% (CI 0.8-0.9), respectively. The AUROC curve for BIPS ≥ 2 was 0.75. The sensitivity, specificity, PPV, and NPV of a BIPS ≥ 2 for BBMI requiring surgery in patients with severe alteration in mental status (GCS 3-8) was 70% (CI 0.5-0.9), 92% (CI 0.8-1.0), 82% (CI 0.6-1.0), and 86% (CI 0.7-1.0), respectively., Conclusion: This prospective multi-center trial validates BIPS as a predictor of BBMI requiring surgery. Calculation of BIPS during the initial evaluation of trauma patients is a useful adjunct to help general surgeons taking trauma call determine operative versus non-operative management of patients with BBMI including those with severe alteration in mental status., Competing Interests: Declarations of Competing Interest None., (Copyright © 2021. Published by Elsevier Ltd.)

Background: Trauma care providers often face a dilemma regarding anticoagulation therapy initiation in patients with traumatic brain injury owing to the associated risks of traumatic brain injury progression. The aims of this study were the following: (1) to describe the current practice of anticoagulation therapy in traumatic brain injury patients and their outcomes and (2) to identify factors associated with the progression of traumatic brain injury after anticoagulation therapy., Methods: In this multicenter prospective observational study, we included computed tomography-proven traumatic brain injury patients who received anticoagulation therapy within 30 days of hospital admission. Our primary outcome was the incidence of clinically significant progression of traumatic brain injury after anticoagulation therapy initiation., Results: A total of 168 patients were enrolled more than 22 months. Atrial fibrillation and venous thromboembolism were the most common pre-injury and postinjury anticoagulation therapy indications, respectively. Overall, 16 patients (9.6%) experienced clinically significant traumatic brain injury progression after anticoagulation therapy, out of which 9 (5.4%) patients subsequently required neurosurgical interventions. Between patients with clinical progression of traumatic brain injury and patients who showed no such progression, there were no significant differences in the baseline demographics and severity of traumatic brain injury. However, anticoagulation therapy was initiated significantly earlier in patients of the deterioration group than those of the no-deterioration group (4.5 days vs 11 days, P = .015). In a multiple logistic regression model, patients who received anticoagulation therapy later after injury had significantly lower risk of clinically significant traumatic brain injury progression (odds ratio: 0.915 for each day, 95% confidence interval: 0.841-0.995, P = .037)., Conclusion: Our results suggest that early anticoagulation therapy is associated with higher risk of traumatic brain injury progression, thus a balance between bleeding and thromboembolic risks should be carefully evaluated in each case before initiating anticoagulation therapy., (Copyright © 2020 Elsevier Inc. All rights reserved.)

The COVID pandemic of 2020 resulted in unprecedented restrictions of public life in most countries around the world, and many hospital systems experienced dramatic decreases in non-COVID related patient admissions. We aimed to compare trauma volumes, patient characteristics, and trauma mechanisms at a large, urban Level 1 trauma center in the United States during a state-wide "State of Emergency" and "stay-at-home" order to corresponding historic dates. All adult trauma activations from March 1 through April 30, 2020 and a historic control from March 1 through April 30, 2018 and 2019 were reviewed in the institution's trauma registry. Trauma volumes, patient characteristics, and trauma mechanisms were compared over time as increasingly stricter COVID-related restrictions were enacted in the Commonwealth of Virginia. After declaration of a state-wide "Public Health Emergency" on March 17, 2020, the daily number of trauma activations significantly declined to a mean of 4.7 (standard deviation, SD = 2.6), a decrease by 43% from a mean of 8.2 (SD = 0.3) for the same dates in 2018 and 2019. Trauma activations during COVID restrictions vs. historic control were characterized by significantly higher prevalence of chronic alcohol use (15.5% vs. 6.8%, p < 0.01), higher median (25th - 75th percentile) Injury Severity Score of 9 (5 - 16) vs. 6 (4 - 14), p = 0.01, and shorter median (25th - 75th percentile) length of hospital stay of 2 (1 - 6) days vs. 3 (1 - 7) days, p = 0.03. The COVID-related Public Health Emergency and "stay-at-home" order in the Commonwealth of Virginia dramatically reduced overall trauma volumes with minor but interesting changes in trauma patterns., Competing Interests: Declaration of Competing Interest This research did not receive any specific grant from funding agencies. Declarations of interest: none., (Copyright © 2020 Elsevier Ltd. All rights reserved.)

Background: Current evidence-based screening algorithms for blunt cerebrovascular injury (BCVI) may miss more than 30% of carotid or vertebral artery injuries. We implemented universal screening for BCVI with computed tomography angiography of the neck at our level 1 trauma center, hypothesizing that only universal screening would identify all clinically relevant BCVIs., Methods: Adult blunt trauma activations from July 2017 to August 2019 underwent full-body computed tomography scan including computed tomography angiography neck with a 128-slice computed tomography scanner. We calculated sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of common screening criteria. We determined independent risk factors for BCVI using multivariate analyses., Results: A total of 4,659 patients fulfilled the inclusion criteria, 2.7% (n = 126) of which had 158 BCVIs. For the criteria outlined in the American College of Surgeons Trauma Quality Improvement Program Best Practices Guidelines, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 72.2%, 64.9%, 6.8%, 98.5%, and 65.2%, respectively; for the risk factors suggested in the more extensive expanded Denver criteria, they were 82.5%, 50.4%, 5.3%, 98.9%, and 51.4%, respectively. Twenty-three percent (n = 14) of patients with BCVI grade 3 or higher would not have been captured by any screening criteria. Cervical spine, facial, and skull base fractures were the strongest predictors of BCVI with odds ratios and 95% confidence intervals of 8.1 (5.4-12.1), 5.7 (2.2-15.1), and 2.7 (1.5-4.7), respectively. Eighty-three percent (n = 105) of patients with BCVI received antiplatelet agents or therapeutic anticoagulation, with 4% (n = 5) experiencing a bleeding complication, 3% (n = 4) a BCVI progression, and 8% (n = 10) a stroke., Conclusion: Almost 20% of patients with BCVI, including a quarter of those with BCVI grade 3 or higher, would have gone undiagnosed by even the most extensive and sensitive BCVI screening criteria. Implementation of universal screening should strongly be considered to ensure the detection of all clinically relevant BCVIs., Level of Evidence: Diagnostic study, level III.

Background: Most triage guidelines for blunt chest wall trauma focus on advanced age and multiple fractured ribs to indicate a high-risk patient population that should be admitted to an intensive care unit (ICU). Overly sensitive ICU admission criteria, however, may result in overutilization of resources. We revised our rib fracture triage guideline to de-emphasize age and number of rib fractures, hypothesizing that we could lower ICU admission rates without compromising outcomes., Methods: Patients admitted to our level 1 trauma center over 9 months after the institution of the revised guideline (N = 248) were compared with those admitted over 6 months following the original guideline (N = 207) using Fisher's exact and Wilcoxon-Mann-Whitney tests, as appropriate. Univariate followed by multivariate analyses were performed to determine risk factors for complications., Results: The ICU admission rate significantly decreased from 73% to 63% ( P = .02) after the institution of the revised guideline, despite an increase in the patient's age and injury acuity of the cohort. There was no significant difference in respiratory complications, unplanned ICU admission rates, and overall mortality. Poor incentive spirometer effort (750 mL or less) and dyspnea in the trauma bay were the strongest predictors of an adverse composite outcome and prolonged hospital length of stay., Discussion: A revised rib fracture triage guideline with less emphasis on the patient's age and the number of fractured ribs safely lowered ICU admission rates. Poor functional status rather than age and anatomy was the strongest predictor of complications and prolonged hospital stay.

Background: Accurate medication reconciliation in trauma patients is essential but difficult. Currently, there is no established clinical method of detecting direct oral anticoagulants (DOACs) in trauma patients. We hypothesized that a liquid chromatography-mass spectrometry (LCMS)-based assay can be used to accurately detect DOACs in trauma patients upon hospital arrival., Methods: Plasma samples were collected from 356 patients who provided informed consent including 10 healthy controls, 19 known positive or negative controls, and 327 trauma patients older than 65 years who were evaluated at our large, urban level 1 trauma center. The assay methodology was developed in healthy and known controls to detect apixaban, rivaroxaban, and dabigatran using LCMS and then applied to 327 samples from trauma patients. Standard medication reconciliation processes in the electronic medical record documenting DOAC usage were compared with LCMS results to determine overall accuracy, sensitivity, specificity, and positive and negative predictive values (PPV, NPV) of the assay., Results: Of 356 patients, 39 (10.96%) were on DOACs: 21 were on apixaban, 14 on rivaroxaban, and 4 on dabigatran. The overall accuracy of the assay for detecting any DOAC was 98.60%, with a sensitivity of 94.87% and specificity of 99.05% (PPV, 92.50%; NPV, 99.37%). The assay detected apixaban with a sensitivity of 90.48% and specificity of 99.10% (PPV, 86.36%; NPV 99.40%). There were three false-positive results and two false-negative LCMS results for apixaban. Dabigatran and rivaroxaban were detected with 100% sensitivity and specificity., Conclusion: This LCMS-based assay was highly accurate in detecting DOACs in trauma patients. Further studies need to confirm the clinical efficacy of this LCMS assay and its value for medication reconciliation in trauma patients., Level of Evidence: Diagnostic Test, level III.

The Joint Commission has established medication reconciliation as a National Patient Safety Goal, but it has not been studied much in trauma even though it is integral to safe patient care. This article reviews the existing medication reconciliation strategies and their applicability to the trauma setting. To perform medication reconciliation, hospitals use a variety of strategies including pharmacists or pharmacy technicians, electronic medical record tools, and patient-centered strategies. All of these strategies are limited in trauma. Subpopulations such as injured children, the elderly, and those with brain trauma are particularly challenging and are at risk for suboptimal care from inaccurate medication reconciliation. Further research is necessary to create a safe and efficient system for trauma patients., (Copyright © 2019 Elsevier Inc. All rights reserved.)

Sepsis, a life-threatening organ dysfunction due to a dysregulated host response to infection, is a leading cause of morbidity and mortality worldwide. Decades of research have failed to identify any specific therapeutic targets outside of antibiotics, infectious source elimination, and supportive care. More recently, vitamin C has emerged as a potential therapeutic agent to treat sepsis. Vitamin C has been shown to be deficient in septic patients and the administration of high dose intravenous as opposed to oral vitamin C leads to markedly improved and elevated serum levels. Its physiologic role in sepsis includes attenuating oxidative stress and inflammation, improving vasopressor synthesis, enhancing immune cell function, improving endovascular function, and epigenetic immunologic modifications. Multiple clinical trials have demonstrated the safety of vitamin C and two recent studies have shown promising data on mortality improvement. Currently, larger randomized controlled studies are underway to validate these findings. With further study, vitamin C may become standard of care for the treatment of sepsis, but given its safety profile, current treatment can be justified with compassionate use., Competing Interests: The primary authors (M.G.K., B.J.F., A.A.F.III) were investigators on the CITRUS-ALI study.

Patient physiology and crash characteristics are essential components of field triage for motor vehicle crashes. We aimed to identify prehospital information that predicted high injury severity or critical patient condition on hospital arrival. The association of demographics, shock index (SI), Glasgow Coma Scale, and 10 crash characteristics of trauma activations for motor vehicle crashes with injury severity score (ISS) ≥ 16 and a composite of hypotension, need for blood transfusions, or immediate operation was determined using univariate and multivariate analyses. A total of 133 of 498 patients (27%) had ISS ≥ 16; SI ≥ 0.9, Glasgow Coma Scale ≤ 8, speed ≥ 55 mph, seatbelt use, airbag deployment, ambulatory patient, severe vehicle damage, ejection, and extrication were associated with ISS ≥ 16. Only abnormal SI and high speed remained independent predictors for ISS ≥ 16 with Odds Ratio (OR) = 10.76 (95% confidence interval (CI), 1.14-101, P = 0.04) and OR = 10.37 (95% CI, 1.48-72.93, P = 0.02), respectively. SI ≥ 0.9 predicted the composite outcome with OR = 5.92 (95% CI, 2.32-15.08, P < 0.01). Many commonly reported crash characteristics did not predict clinically important outcomes. Improvements in road and vehicle safety may be resulting in lower injury severity despite major crash mechanisms.

Background: The shortage in intensivist workforce has been long recognized but no solution has been identified. Meanwhile, fellowships in pulmonary and critical care medicine (PCCM) are expanding, other critical care medicine (CCM) programs are contracting. No explanation exists for this contradictory trend, although understanding contributory factors may lead to a solution for the shortage. The fundamental difference between PCCM and other CCM programs lies in the residency training of trainees. We tested the hypothesis that the nature of CCM practice determines its attractiveness to potential candidates., Methods: A questionnaire-based survey was administered recording all daily activities in four different kinds of ICUs at two teaching hospitals one was public, and one was private. Activities were categorized into conventional CCM, respiratory, medical, and surgical interventions., Results: The average daily census was 17.6 ± 6.6. Across two MICU, one trauma/surgical and one cardiothoracic ICU the average daily activity ranged from 152 to 203 of these CCM formed 27%-36%, respiratory 10%-13%, medical 43%-59%, and surgical 1%-15%. The combination of medical and respiratory interventions represented >50% of daily activities among all the ICUs., Conclusions: Quantitative description of ICU activities indicates that the majority of the ICU daily practice relies on medical and respiratory interventions, which may explain why PCCM remains popular., (© 2019 John Wiley & Sons Ltd.)

Background: Medication errors account for the most common adverse events and a significant cause of mortality in the USA. The Joint Commission has required medication reconciliation since 2006. We aimed to survey the literature and determine the challenges and effectiveness of medication reconciliation in the trauma patient population., Materials and Methods: We conducted a systematic review of the literature to determine the effectiveness of medication reconciliation in trauma patients. English language articles were retrieved from PubMed/Medline, CINAHL, and Cochrane Review databases with search terms "trauma OR injury, AND medication reconciliation OR med rec OR med rek, AND effectiveness OR errors OR intervention OR improvements.", Results: The search resulted in 82 articles. After screening for relevance and duplicates, the 43 remaining were further reviewed, and only four articles, which presented results on medication reconciliation in 3041 trauma patients, were included. Two were retrospective and two were prospective. Two showed only 4% accuracy at time of admission with 48% of medication reconciliations having at least one medication discrepancy. There were major differences across the studies prohibiting comparative statistical analysis., Conclusions: Trauma medication reconciliation is important because of the potential for adverse outcomes given the emergent nature of the illness. The few articles published at this time on medication reconciliation in trauma suggest poor accuracy. Numerous strategies have been implemented in general medicine to improve its accuracy, but these have not yet been studied in trauma. This topic is an important but unrecognized area of research in this field., Competing Interests: N/A: systematic reviewN/A: systematic reviewThe authors declare that they have no competing interests.Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Objective: To compare continuous infusion preperitoneal wound catheters (CPA) versus continuous epidural analgesia (CEA) after elective colorectal surgery., Methods: An open-label equivalence trial randomizing patients to CPA or CEA. Primary outcomes were postoperative pain as determined by numeric pain scores and supplemental narcotic analgesia requirements. Secondary outcomes included incidence of complications and patient health status measured with the SF-36 Health Survey (Acute Form)., Results: 98 patients were randomized [CPA (N = 50, 51.0%); CEA (N = 48, 49.0%)]. 90 patients were included [ CPA 46 (51.1%); CEA 44 (48.9%)]. Pain scores were significantly higher in the CPA group in the PACU (p = 0.04) and on the day of surgery (p < 0.01) as well as supplemental narcotic requirements on POD 0 (p = 0.02). No significant differences were noted in postoperative complications between groups, aggregate SF-36 scores and SF-36 subscale scores., Conclusions: Continuous epidural analgesia provided superior pain control following colorectal surgery in the PACU and on the day of surgery. The secondary endpoints of return of bowel function, length of stay, and adjusted SF-36 were not affected by choice of peri-operative pain control., (Copyright © 2017 Elsevier Inc. All rights reserved.)

The point-of-care ultrasound exam has become an essential tool for hemodynamic monitoring and resuscitation in the trauma bay as well as the intensive care unit. Transthoracic ultrasound provides a dynamic assessment of cardiac function, volume status, and fluid responsiveness that offers potential advantage over traditional methods of hemodynamic monitoring. More recently, a focused transthoracic echocardiography exam was described to improve immediate resuscitation of severely injured patients in the trauma bay. Transesophageal echocardiography (TEE) for trauma could expand upon the role of focused echocardiography. TEE offers improved visualization of cardiac anatomy and physiology, improved diagnostic accuracy, and real-time assessment of intraoperative resuscitation progress, particularly in the operating room. This review discusses the fundamental principles of echocardiography as well as different ultrasound modes with their respective strengths and limitations. It reviews the current literature on the use of TEE in trauma, and suggests views for a trauma resuscitation transesophageal echocardiography exam (TREE), including sample images and videos., (Copyright © 2017 Elsevier Inc. All rights reserved.)

Background: The role of diagnostic imaging in suspected necrotizing soft tissue infections (NSTIs) is not clear owing to concerns about its value and possible delays in definitive surgical care., Methods: Plain radiograph (XR) and computed tomography (CT) results of all patients who underwent operative debridement for a presumed NSTI from 2007 through 2014 at LAC + USC Medical Center were reviewed. Preoperative imaging was classified as being negative, suspicious (inflammatory changes), or diagnostic (soft tissue gas) for NSTI., Results: Of 226 patients undergoing operative exploration for a suspected NSTI, 172 (76.1%) were found to have a true NSTI based on intraoperative or pathology findings. In patients with true NSTI, preoperative XR and CT demonstrated soft tissue gas in 47.9% and 70.3% of cases, respectively. CT diagnosed or highly suspected NSTI in 97.3% of cases with true NSTI compared to 83.6% with XR; p < 0.001)., Conclusion: CT was superior to XR in the radiologic evaluation of patients with suspected NSTIs., Level of Evidence: Diagnostic test, level IV.

Background: Colonic obstruction is a surgical emergency, and delay in decompression results in added morbidity and mortality. Advances have led to less invasive procedures such as stenting as a bridge for definitive surgery. The aim of this article was to perform a systematic review regarding colon obstruction (malignant or benign) and to provide recommendations following the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework., Methods: A systematic literature review was conducted using the PubMed, EMBASE, and the Cochrane Library databases of published studies. The search was last performed on January 2, 2015. Two independent reviewers extracted the desired variables from the studies. For our meta-analysis, we used Review Manager X.6 (RevMan). Recommendations are provided using GRADE methodology. A single POPULATION, Intervention, Comparator, Outcome (PICO) question with two outcomes was addressed as follows:, Population: in adult patients with a colonic obstruction (malignant or benign)., Intervention: should surgery be performed.Comparator: versus endoscopic stenting., Outcomes: decreased mortality and decreased emergency, nonplanned procedures?, Results: The search yielded 210 results. Screening of the titles excluded 102 articles, leaving 108 for review. After abstract review, 71 additional articles were excluded because of failure to address the PICO questions of this guideline. Thirty-seven articles were reviewed in their entirety, of those six randomized control trials that evaluated the use of stents versus emergency surgery in colonic obstruction caused by malignant disease were included in the final qualitative review., Conclusion: We conditionally recommend endoscopic, colonic stenting (if available) as initial therapy for colonic obstruction. In our review, stent use was associated with decreased mortality and rates for emergency, nonplanned procedures to include reoperations. This conditional recommendation is limited to those with malignancy because of the lack of literature supporting this practice in benign colonic disease.

Background: The role of hepato-imino diacetic acid scan (HIDA) in the diagnosis of acute cholecystitis remains controversial when compared with the more commonly used abdominal ultrasound (AUS)., Methods: The diagnostic imaging workup of 1,217 patients who presented to the emergency department at a single hospital with acute abdominal pain and suspicion of acute cholecystitis was reviewed to calculate the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of AUS and HIDA., Results: In patients undergoing both imaging modalities, HIDA had significantly higher sensitivity (90.7% vs 64.0%, P < .001) and specificity (71.4% vs 58.4%, P = .005) than AUS for the diagnosis of acute cholecystitis. Additionally, PPV and NPV of HIDA (56.2% and 95.0%, respectively) were higher than PPV and NPV of AUS (38.4% and 80.0%, respectively) when both imaging modalities were used for the same patient., Conclusion: In adults with acute abdominal pain, HIDA significantly increases the accuracy of the correct diagnosis., (Copyright © 2015 Elsevier Inc. All rights reserved.)

Background: Surgical site infections are a major cause of morbidity and mortality after colorectal operations. Preparation of the surgical site with antiseptic solutions is an essential part of wound infection prevention. To date, there is no universal consensus regarding which preparation is most efficacious., Objective: This study compared 2.0% chlorhexidine with 70.0% isopropyl alcohol versus 0.7% iodine povacrylex with 74.0% isopropyl alcohol and alcohol-based versus nonalcohol-based skin preparations with regard to efficacy in preventing postoperative wound infections., Design: This is a retrospective study from 2 prospectively collected statewide databases combined. A propensity score model was used to adjust for differences between the groups in patient demographics, characteristics, comorbidities, and laboratory values., Settings: The multicenter data set used in this analysis represents a variety of academic and community hospitals within the state of Michigan from January 2010 through June 2012., Patients: Patients over the age of 18 years who underwent clean-contaminated colorectal operations were included., Main Outcome Measures: The incidence of superficial surgical site infections, any surgical site infection, any wound complication, and readmission within 30 days for surgical site infection were measured., Results: When 2.0% chlorhexidine with 70.0% isopropyl alcohol (n = 425) and 0.7% iodine povacrylex with 74.0% isopropyl alcohol (n = 115) were compared, a total of 540 colorectal cases met inclusion criteria. When alcohol-based (n = 610) and nonalcohol-based (n = 177) skin preparations were compared, a total of 787 colorectal cases met inclusion criteria. There was no significant difference in the propensity-adjusted odds for having any of the 4 outcomes of interest when comparing 2.0% chlorhexidine with 70.0% isopropyl alcohol to 0.7% iodine povacrylex with 74.0% isopropyl alcohol and when comparing alcohol-based with nonalcohol-based skin preparations., Limitations: This was a nonrandomized study performed retrospectively based on data collected within the state of Michigan., Conclusions: The use of 2.0% chlorhexidine with 70.0% isopropyl alcohol versus 0.7% iodine povacrylex with 74.0% isopropyl alcohol or alcohol-based versus nonalcohol-based skin preparations does not significantly influence the incidence of surgical site infections or readmission within 30 days for surgical site infection after clean-contaminated colorectal operations.

Purpose: The purpose of this study was to identify predictive factors for postoperative surgical site infections (SSIs), and increased length of hospital stay (LOS) after ventral/incisional hernia repair (VIHR) using multi-center, prospectively collected data., Study Design: Cases of VIHR from 2009 to 2010 were identified in the American College of Surgeons National Surgical Quality Improvement Program database. Using logistic regression, a prediction model utilizing 41 variables was developed to identify risk factors for postoperative SSIs, and increased LOS. Separate analyses were carried out for reducible and incarcerated/strangulated cases., Results: A total of 28,269 cases of VIHR were identified, 25,172 of which met inclusion criteria. 18,263 cases were reducible hernias, and 6,909 cases were incarcerated/strangulated hernias. Our prediction model demonstrated that body mass index ≥30 kg/m(2), smoking, American Society of Anesthesiology (ASA) class 3, open surgical approach, prolonged operative times, and inpatient admission following VIHR were significant predictors of postoperative SSIs. In addition, risk factors associated with prolonged LOS included older age, African American ethnicity, history of alcohol abuse, ASA classes 3 and 4, poor functional status, operation within the last 30 days of the index operation, history of chronic obstructive pulmonary disease, congestive heart failure, and bleeding disorder, as well as open surgical approach, non-involvement of residents, prolonged operative times, recurrent hernia, emergency operation, and low preoperative serum albumin level., Conclusions: Obesity and smoking are modifiable risk factors for SSIs after VIHR, whereas a low serum albumin level is a modifiable risk factor for prolonged LOS. Addressing factors preoperatively might improve patient outcome, and reduce health care expenditures associated with VIHR. In addition, if feasible, the laparoscopic approach should be strongly considered.

Background: Robotic assistance may offer unique advantages over conventional laparoscopy in colorectal operations., Methods: This prospective observational study compared operative measures and postoperative outcomes between laparoscopic and robotic abdominal and pelvic resections for benign and malignant disease., Results: From 2005 through 2012, 200 (58%) laparoscopic and 144 (42%) robotic operations were performed by a single surgeon. After adjustment for differences in demographics and disease processes using propensity score matching, all laparoscopic operations had a significantly shorter operative time (P < .01), laparoscopic left colectomies had a longer length of hospital stay (2009 and 2010: 6.5 vs 3.6 days, P = .01); and laparoscopic right colectomies had a higher risk for overall complications (P = .03) and postoperative ileus (P = .04). There were no significant differences in the outcomes of pelvic operations (P = .15)., Conclusions: Compared with conventional laparoscopy, some types of robotic-assisted colorectal operations may offer advantages regarding postoperative length of stay and perioperative complications., (Copyright © 2014 Elsevier Inc. All rights reserved.)

Background: Acute cholecystitis is one of the most common surgical problems, yet substantial debate remains over the utility of simple examination, abdominal ultrasound (AUS), or advanced imaging such as hepato-imino diacetic acid (HIDA) scan to support the diagnosis., Materials and Methods: The preoperative diagnostic workup of patients who underwent cholecystectomy with histologically confirmed acute cholecystitis was reviewed to calculate the sensitivity of AUS, HIDA scan, or both. In addition, the sensitivity of the commonly described ultrasonographic findings was assessed., Results: From 2010 through 2012, 406 patients among 9087 reviewed charts presented to the emergency department with acute upper abdominal pain and met inclusion criteria. 32.5% (N = 132) of patients underwent AUS only, 11.3% (N = 46) underwent HIDA scan only, and 56.2% (N = 228) had both studies performed for workup. 52.7% (N = 214) of patients had histopathologically confirmed acute cholecystitis. The sensitivities of AUS, HIDA, and AUS combined with HIDA for acute cholecystitis were 73.3% (95% confidence interval [CI] = 66.3%-79.5%), 91.7% (95% CI = 86.2%-95.5%), and 97.7% (95% CI = 93.4%-99.5%), respectively. Although of limited sensitivity, AUS findings of sonographic Murphy sign, gallbladder distension, and gallbladder wall thickening were associated with a diagnosis of acute cholecystitis., Conclusions: The sensitivity of AUS for diagnosing acute cholecystitis in patients with acute upper abdominal pain is limited. The addition of a HIDA scan in the diagnostic workup significantly improves sensitivity and can add valuable information in the appropriate clinical setting., (Copyright © 2014 Elsevier Inc. All rights reserved.)

Background: Colon resections are associated with substantial risk for morbidity and readmissions, and these have become markers for quality of care., Objective: The purpose of this study was to determine risk factors for readmissions after elective colectomies to improve patient care and better understand the complex issues associated with readmissions., Design: This was an analysis of the prospective, statewide, multicenter Michigan Surgical Quality Collaborative database., Settings: The analysis was conducted at academic and community medical centers in the state of Michigan., Patients: Elective laparoscopic and open ileocolic and segmental colectomies from 2008 through 2010 were included., Main Outcome Measures: Univariate analysis and a multivariate logistic regression model were used to determine influence of patient characteristics, operative factors, and postoperative complications on the incidence of 30-day postoperative readmission., Results: The readmission rate among 4013 cases was 7.3% (N = 293). On the basis of multivariate logistic regression, the top 3 significant risk factors associated with readmission were stroke (OR, 10.0 [95% CI, 2.70-37.0]; p = 0.001), venous thromboembolism (OR, 6.5 [95% CI, 3.7-11.3]; p < 0.0001), and organ-space surgical site infection (OR, 5.6 [95% CI, 3.4-9.4]; p < 0.0001). Important factors that contributed to readmission risk but were not found to be independent predictors of readmission included diabetes mellitus, preoperative steroids, smoking, cardiac comorbidities, age >80 years, anastomotic leaks, fascial dehiscence, sepsis, pneumonia, unplanned intubation, and length of stay., Limitations: The Michigan Surgical Quality Collaborative is a large database, and true causal relations are difficult to determine; reason for readmission is not recorded in the database., Conclusions: Postoperative complications account for the majority of risk factors behind readmissions after elective colectomy, whereas preoperative risk factors have less direct influence. Current strategies addressing readmission rates should focus on reducing preventable complications.

Purpose: Peripheral tissue oxygen saturation (Sto2) has shown promise as an early indicator of tissue hypoperfusion and as a risk stratification tool in various forms of shock. The purpose of this study was to determine if Sto2 would predict admission to an intensive (ICU) or progressive care unit in patients with early signs of sepsis., Methods: In this prospective observational study, a rapid response team measured Sto2 levels in patients screening positive for sepsis. Using a logistic regression model, the value of Sto2 as a predictor for ICU admission within 72 hours of the initial assessment was determined., Results: The 31 (47%) of 66 patients who required ICU admission within 72 hours of evaluation had a significantly lower Sto2 value (median, 78% vs 81%; P = .05). All patients with Sto2 less than 70% required ICU admission. A 1-point increase in Sto2 was associated with a 7% decrease in the odds of requiring ICU admission, and the area under the curve for Sto2 was 0.64 (0.51-0.77, P = .01)., Conclusions: Low Sto2 levels in patients screening positive for sepsis are associated with an increased risk of ICU admission, but their reliability as a predictor is rather low. An Sto2 below 70% might be an interesting cutoff value for further study., (© 2013.)

Background: The classic Whipple operation carries substantial risk of complications. A pylorus-preserving pancreaticoduodenectomy might confer the benefit of decreased perioperative morbidity, but existing data comparing both techniques are inconclusive., Methods: Using a propensity score model to adjust for potentially confounding differences in patient characteristics, 30-d mortality, operative time, red blood cell transfusion requirements, major complications, and length of hospital stay were compared between both techniques in the American College of Surgeons' National Surgical Quality Improvement Program database. Separate analyses were carried out for underlying malignancy or benign disease, as defined by International Classification of Diseases, Ninth Revision codes., Results: A total of 6988 pancreaticoduodenectomies from 2005 through 2010 were included. In 5424 patients (77.6%) with underlying malignancy, there were no significant differences for 30-d mortality (2.4% versus 2.8%, P = 0.33) and major organ system complications (all P > 0.10). Patients undergoing the classic Whipple operation had a significantly longer operative time (389 versus 366 min, P < 0.01), longer length of hospital stay (13.1 versus 12.0 days, P < 0.01), and higher red blood cell transfusion requirements (1.0 versus 0.8 units, P < 0.01). Results were similar for 1564 patients (22.4%) with underlying benign disease, except for a higher occurrence of postoperative pulmonary (P = 0.02) and renal (P = 0.05) complications in patients undergoing the classic Whipple operation., Conclusions: Short-term outcomes after classic and pylorus-preserving pancreaticoduodenectomy in this large, multicenter database are excellent, without significant differences in postoperative mortality and most major organ system complications. However, small advantages in resource and blood utilization may be accomplished with the pylorus-preserving technique., (Copyright © 2013 Elsevier Inc. All rights reserved.)

Background: The purpose of this study was to compare the incidence of postoperative surgical site infections (SSIs), operative times (OTs), and length of hospital stay (LOS) after open and laparoscopic ventral/incisional hernia repair (VIHR) using multicenter, prospectively collected data., Methods: The incidence of postoperative SSIs, OTs, and LOS was determined for cases of VIHR in the American College of Surgeons' National Surgical Quality Improvement Program database in 2009 and 2010. Open and laparoscopic techniques were compared using a propensity score model to adjust for differences in patient demographics, characteristics, comorbidities, and laboratory values., Results: A total of 26,766 cases met the inclusion criteria; 21,463 cases were open procedures (reducible, n = 15,520 [72 %]; incarcerated/strangulated, n = 5,943 [28 %]), and 5,303 cases were laparoscopic procedures (reducible, n = 3,883 [73 %]; incarcerated/strangulated, n = 1,420 [27 %]). Propensity score adjusted odds ratios (ORs) were significantly different between open and laparoscopic VIHR for reducible and incarcerated/strangulated hernias with regard to superficial SSI (OR 5.5, p < 0.01 and OR 3.1, p < 0.01, respectively), deep SSI (OR 6.9, p < 0.01, and OR 8.0, p < 0.01, respectively) and wound disruption (OR 4.6, p < 0.01 and OR 9.3, p = 0.03, respectively). The risk for organ/space SSI was significantly greater for open operations among reducible hernias (OR 1.9, p = 0.02), but there was no significant difference between the open and laparoscopic repair groups for incarcerated/strangulated hernias (OR 0.8, p = 0.41). The OT was significantly longer for laparoscopic procedures, both for reducible (98.5 vs. 84.9 min, p < 0.01) and incarcerated/strangulated hernias (96.4 vs. 81.2 min, p < 0.01). LOS (mean, 95 % confidence interval) was significantly longer for open repairs for both reducible (open = 2.79, 2.59-3.00; laparoscopic = 2.39, 2.20-2.60; p < 0.01) and incarcerated/strangulated (open = 2.64, 2.55-2.73; laparoscopic = 2.17, 2.02-2.33; p < 0.01) hernias., Conclusions: Laparoscopic VIHR for reducible and incarcerated/strangulated hernias is associated with shorter LOS and decreased risk for superficial SSI, deep SSI, and wound disruption, but longer OTs when compared to open repair.

Purpose: Many clinical trials comparing the outcomes of open surgical repair (OSR) versus endovascular aneurysm repair (EVAR) for abdominal aortic aneurysms (AAAs) have been conducted, with varying results. Surprisingly, few outcomes studies have closely examined perceived physical and mental health-related quality of life (HRQOL) factors through a validated survey tool. The purpose of this prospective observational study was to describe the trajectory of HRQOL measures, from baseline to 1 year after surgery, in patients undergoing OSR or EVAR for AAA, and to explore for differences in physical and mental composite scores and their construct domains (subscales) using the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36(®)) tool., Patients and Methods: Over an 18-month period, a small sample of patients undergoing elective AAA repair in a community hospital setting were prospectively enrolled. Fifteen patients undergoing OSR and twenty patients undergoing EVAR were studied. Physical and mental HRQOL parameters were assessed using the SF-36., Results: No significant differences in demographic and clinical variables were found between the OSR and EVAR groups. In the multivariable linear models with repeated measures, both groups showed a significant decline in physical health composite scores 30 days after the surgical procedure (P < 0.01). However, although the OSR group showed a statistically significant decline in three of the four physical health domains, the EVAR group declined in only one physical health domain. Only the OSR group showed a significant decline in three of the four mental health domains at 30 days; however, the decline of these domains was not reflected in the group's mental health composite scores. By 90 days after surgery, both groups were not significantly different from their baseline in physical or mental health composite scores, or in any of their respective physical health domains., Conclusion: In this small sample of patients undergoing AAA repair, EVAR resulted in less physical and emotional decline than OSR in the early postoperative period. However, patients in both groups may return to near baseline status at 90 days.

Objective: Despite multiple studies over more than 3 decades, there still is no consensus about the influence of anesthesia type on postoperative outcomes following carotid endarterectomy (CEA). The objective of this study was to investigate whether anesthesia type, either general anesthesia (GA) or regional anesthesia (RA), independently contributes to the risk of postoperative cardiovascular complications or death using the American College of Surgeons' National Surgical Quality Improvement Program (ACS NSQIP) database., Methods: Retrospective analysis of elective cases of CEA from 2005 through 2009 was performed. A propensity score model using 45 covariates, including demographic factors, comorbidities, stroke history, measures of general health, and laboratory values, was used to adjust for bias and to determine the independent influence of anesthesia type on postoperative stroke, myocardial infarction (MI), and death., Results: Of 26,070 cases listed in the ACS NSQIP database, GA and RA were used in 22,054 (84.6%) and 4016 (15.4%) cases, respectively. Postoperative stroke, MI, and death occurred in 360 (1.63%), 133 (0.6%), and 154 (0.70%) patients of the GA group, respectively, and in 58 (1.44%), 11 (0.27%), and 27 (0.67%) patients of the RA group, respectively. Stratification by propensity score quintile and adjustment for covariates demonstrated GA to be a significant risk factor for postoperative MI with an adjusted odds ratio (OR) and confidence interval (CI) of 2.18 (95% CI, 1.17-4.04), P = .01 in the entire study population. The OR for MI was 5.41 (95% CI, 1.32-22.16; P = .019) in the subgroup of patients with preoperative neurologic symptoms, and 1.44 (95% CI, 0.71-2.90; P = .31) in the subgroup of patients without preoperative neurologic symptoms., Conclusions: This analysis of a large, prospectively collected and validated multicenter database indicates that GA for CEA is an independent risk factor for postoperative MI, particularly in patients with preoperative neurologic symptoms., (Copyright © 2012 Society for Vascular Surgery. Published by Mosby, Inc. All rights reserved.)

Background: Anastomotic leakage is a morbid and potentially fatal complication of colorectal surgery. Determination of pre- and intraoperative risk factors may identify patients requiring increased postoperative surveillance for this major complication., Objective: The purpose of this study was to identify risk factors associated with anastomotic leakage after colectomy with primary intra-abdominal anastomosis., Design: The prospective, statewide multicenter Michigan Surgical Quality Collaborative database was analyzed., Setting: This study was performed at academic and community medical centers in the state of Michigan., Patients: Included were all cases of open and laparoscopic colectomy with primary intra-abdominal anastomosis from 2007 through 2010., Main Outcome Measures: Univariate analysis followed by a multivariate logistic regression model was used to determine the influence of patient factors and operative events with respect to the incidence of postoperative anastomotic leakage., Results: Inclusion criteria were met by 4340 cases. Anastomotic leakage occurred in 85 (3.2%) of the 2626 (60.5%) open colectomies, and in 51 (3.0%) of the 1714 (39.5%) laparoscopic procedures, which was not significantly different (p = 0.63). Significant risk factors associated with anastomotic leakage based on the multivariate logistic regression model were fecal contamination with OR 2.51, 95% CI, 1.16 to 5.45, p = 0.02; and intraoperative blood loss of more than 100 mL and 300 mL, with OR 1.62, 95% CI, 1.10 to 2.40, p = 0.02; and OR 2.22, 95% CI, 1.32 to 3.76, p = 0.003., Limitations: The Michigan Surgical Quality Collaborative colectomy project excluded high-risk rectal resections and low pelvic anastomoses. Information about operative technique and intraoperative events is limited, and anastomotic leakage was determined through chart review., Conclusion: Fecal contamination and increased blood loss during colectomy should raise suspicion for potential postoperative anastomotic leakage.

Background: Complications associated with blood transfusions have resulted in widespread acceptance of low hematocrit levels in surgical patients. However, preoperative anemia seems to be a risk factor for adverse postoperative outcomes in certain surgical patients. This study investigated the National Surgical Quality Improvement Program (NSQIP) database to determine if preoperative anemia in patients undergoing open and laparoscopic colectomies is an independent predictor for an adverse composite outcome (CO) consisting of myocardial infarction, stroke, progressive renal insufficiency or death within 30 days of operation, or for an increased hospital length of stay (LOS)., Study Design: Hematocrit levels were categorized into 4 classes: severe, moderate, mild, and no anemia. From 2005 to 2008, the NSQIP database recorded 23,348 elective open and laparoscopic colectomies that met inclusion criteria. Analyses using multivariable models, controlling for potential confounders and stratifying on propensity score, were performed., Results: Compared with nonanemic patients, those with severe, moderate, and mild anemia were more likely to have the adverse CO with odds ratios of 1.83 (95% CI 1.05 to 3.19), 2.19 (95 % CI 1.63 to 2.94), and 1.49 (95% CI 1.20 to 1.86), respectively. Patients with a normal hematocrit had a reduced hospital LOS, compared with those with severe, moderate, and mild anemia (p < 0.01). A history of cardiovascular disease did not significantly influence these findings., Conclusions: This large multicenter database analysis suggests that the presence of severe and moderate and even mild preoperative anemia is an independent risk factor for complications and a longer hospital stay after colon surgery., (Copyright © 2011 American College of Surgeons. Published by Elsevier Inc. All rights reserved.)