The American Surgeon™

Acute mesenteric ischemia is a morbid disease process that is most common in elderly patients who often have multiple medical comorbidities. Intervention can progress to costly and futile care. The goal of this study was to develop a tool for practitioners to assess the risk of mortality. Patients treated at our institution over the past decade diagnosed with acute mesenteric ischemia were identified. Patients aged less than 65 years were excluded. Data were collected by retrospective chart review. Univariate analysis was used to identify significant risk factors for death. Decision tree analysis yielded a prognostic tool to assess death risk. Univariate analysis demonstrated that lactate (P ≤ 0.001) and pressor requirement (P ≤ 0.001) were predictive of death. Decision tree analysis showed that 79 per cent of patients with day of surgery (DOS) lactate ≥5.4 died postoperatively. Seventy per cent of patients with DOS lactate <5.4 progressed to death if they required pressors and had a creatinine >1.18. Only 6.1 per cent patients with a DOS lactate <5.4, creatinine <1.54, and no pressor requirement progressed to death. Several variables can be used to set expectations for families and help guide decision-making. Our tool was predictive of outcomes in 82 per cent of our study population.

Uniformity in surgical education is challenging because surgical experience is based on rotation assignments. With work hour restrictions, the likelihood of residents being exposed to rare or unusual cases is diminished. Telemedicine may create a new learning paradigm for surgical education and supplement exposure for rare or unusual cases. A retrospective review (2010–2016) of teleconferences involving trauma centers worldwide was conducted. Participating hospitals included centers from underdeveloped countries to first world nations. Trauma cases were discussed among surgeons with different levels of experience and resource availability. Data collected included types of cases, anatomic injury patterns, hospital location, and the number of telemedicine centers and viewers participating. Seventy-three hospitals in 64 cities, spanning 27 countries, participated in 276 telemedicine grand round conferences. Cases discussed included penetrating trauma (47%), blunt trauma (42%), and blast injury (4%). The anatomic regions included were the thorax (28%), abdomen (26%), thoracoabdominal region (13%), neck (7%), and pelvis (6%). The most common injury discussed was vascular in nature (18%), followed by the lung, liver, diaphragm, and heart. The most common vascular lesion was in the aorta (18%), followed by the iliac vessels (8%) and the vena cava (7%). Telemedicine is a valuable tool, allowing the dissemination of diverse experiences. Most cases presented evaluated rare injuries or complex surgical approaches, which are not commonly seen on trauma sites. Learning different approaches in the management of complex trauma will make surgeons more prepared to deal with challenging cases.

Accelerated partial breast irradiation (APBI) using the implanted brachytherapy device MammoSite® was approved for routine use by the Food and Drug Administration in 2002. The American Society of Breast Surgeons MammoSite® Breast Brachytherapy Registry served as a guideline for our institution to begin offering this treatment in 2005. This report reviews our available data to provide an analysis of patient outcomes over 12 years of use at a single institution. A retrospective review was conducted of records of 150 patients who underwent APBI or attempted APBI after breast-sparing surgeries between 2006 and 2017. These charts were analyzed for documentation of patient age, cancer stage, incidence of recurrence, and posttreatment complications. Of the patients evaluated, 99 per cent (149/150) completed treatment. The median time since treatment completion is now 8.9 years. One hundred eleven patients (74%) are now greater than five years posttreatment. Ipsilateral breast recurrence was found in 2.7 per cent of patients (4/149), and 1.3 per cent of patients (2/149) developed new primary breast tumors. Acute complications, mostly skin erythema (21%), were uncommon and self-limited. Subacute effects were generally fibrosis (13%) and mild local pain (9.4%). APBI for breast cancer after breast-conserving surgery continues to be used at our institution for select patients with good outcomes. Local control and toxicity are similar to that reported in the literature. Five-year local recurrence rates compare favorably with national trials. Occasional complications included fibrosis, persistent pain, and skin irritation.

^99mTechnetium sestamibi scans (MIBI) can provide negative or inaccurate results in patients with biochemical primary hyperparathyroidism. Reliance on MIBI as a diagnostic modality rather than as a localization tool leads to misdiagnosis and inappropriate care. The aim of this study was to determine the impact of negative MIBI scans on referral patterns and surgical intervention. Adults with MIBI scans at our institution from January 1, 2011, to May 31, 2017, were retrospectively reviewed. Data collected include demographics, study date and results, ordering physician specialty, pre/postoperative laboratories, and operative and final pathology. Statistical analysis was performed with SPSS v24 (IBM Corp., Armonk, NY). Three hundred fifty-seven patients had a MIBI scan; 10 were excluded for incomplete data or incorrect diagnosis. One hundred eighty-six were interpreted as positive (53.6%) and 161 were interpreted as negative (46.4%). Of the 186 positive MIBI scans, 135 (72.6%) were seen by an endocrine surgeon (ES). Of these 135 patients, 111 (82.2%) underwent parathyroidectomy. Of the 161 negative MIBI scans, 69 (42.9%) were seen by an ES. Of these, 53 (76.8%) underwent parathyroidectomy. In all, 90/92 (97.8%) with a negative MIBI scan who were not seen by an ES did not have surgery. Nonendocrine surgeon physicians are more likely to use MIBI scans as diagnostic tools to assist with clinical decision-making. Patients with a negative MIBI scan not seen by an ES were significantly less likely to undergo para-thyroidectomy. Patients with primary hyperparathyroidism should be referred to an experienced parathyroid surgeon for evaluation, regardless of the MIBI result.

The increased use of CT scans has raised concerns regarding the risks of early radiation exposure in the Puerto Rico pediatric population. Available literature sustains that repeated exposure to imaging-related radiation in the pediatric population is associated with a 3-fold increase in the risk of developing pediatric hematogenous and central nervous system malignancies. It is for this reason that an international effort known as the Image Gently Campaign was created, mostly based on the “As Low As Reasonably Achievable” (ALARA) principle described by the Center for Disease Control. With this in mind, our aim was to identify whether there are any discrepancies in imaging tendencies outside our pediatric academic center in Puerto Rico and to determine whether our patients are at increased risk of over-radiation. There were 181 patients; five were excluded because of incomplete data. Our results show that children with appendicitis who are evaluated at nontertiary centers are more likely to have a CT scan performed (93%, OR: 4.054; 95% confidence interval: 2.6–6.4), as opposed to a nonradiating imaging study. In the Pediatric University Hospital, a CT scan was performed as the initial study in 23 per cent of the patients (OR: 0.09; 95% confidence interval: 0.05–0.18), favoring ultrasound as the diagnostic modality of choice. Our concern is that if this trend does not change, our pediatric population might have an increased risk of developing associated malignancies. We believe a local effort toward educational strategies should be implemented to prevent radiation overexposure in our pediatric patients.

Geriatric trauma patients with low-level falls often have multiple comorbidities and limited physiologic reserve. Our aim was to investigate postdischarge mortality in this population. We hypothesized that five-year mortality would be higher relative to other blunt mechanisms. The registry of our Level 1 trauma center was queried for patients evaluated between July 2008 and December 2012. Adult patients identified were matched with mortality data from 2008 to 2013 from the National Death Index. Low-level falls were identified by E Codes; other types of blunt trauma were based on registry classification. Patients with multiple admissions were excluded. Univariate analysis was performed using Fisher's exact and Wilcoxon tests. Kaplan-Meier curves were plotted to compare postdischarge mortality. Seven thousand nine hundred sixteen patients were evaluated, 35.1 per cent were females. Patients aged less than 65 years and penetrating trauma were excluded, yielding 1997 patients—63.7 per cent with low-level falls versus 36.3 per cent with other blunt traumas. Geriatric patients sustaining low-level falls were older, more likely female, had a higher inpatient mortality, and were less likely to return home at discharge. Injury severity score, hospital length of stay, and intensive care unit length of stay were similar. Survival analysis demonstrated increased postdischarge mortality in the low-level fall group with 25 per cent mortality at 120 days. Geriatric patients with other blunt trauma had a significantly lower postdischarge mortality. Geriatric patients injured in low-level falls have a higher inhospital mortality, are more likely to be functionally dependent on discharge, and have a high post-discharge mortality. Opportunities likely exist for injury prevention, consideration of palliative care, and postdischarge rehabilitation.

The American College of Surgeons Committee on Trauma defines undertriage (UT) as any major trauma patient (injury severity score ≥ 16) not undergoing treatment at the highest level of trauma team activation. This methodology does not account for many important factors that may impact outcome. We performed a retrospective review of the Pennsylvania State Trauma Registry to determine the impact of treatment interventions on mortality. Patients were stratified by triage category as follows: UT, appropriate triage, and overtriage. Multiple prehospital (PH) and ED interventions were assessed. Increased mortality was observed in all triage groups in patients requiring intervention. A logistic regression analysis was performed to assess the independent effect of individual interventions on mortality for patients triaged to partial activation or consult. PH CPR (OR 66.13 [47.07–92.93]), ED CPR (OR 16.87 [8.82–32.27]), PH or ED intubation (OR 16.68 [13.90–20.03]), PH or ED packed red blood cell transfusion (OR 1.89 [1.54–2.33]), emergent operative intervention (OR 3.58 [3.07–4.19]), ED central venous access (OR 5.04 [2.31–10.97]) were all associated with worsening mortality. The American College of Surgeons Committee on Trauma methodology overestimates mortality risk when emergent interventions are not required and underestimates risk where such interventions are necessary. Future methodologies for assessing UT should include these interventions.

To determine whether monitored anesthesia care (MAC) results in shorter colonoscopy time. A retrospective chart review from electronic medical records at Sentara Norfolk General Hospital was performed of all patients seen by the Eastern Virginia Medical School Department of Surgery who underwent a screening or diagnostic colonoscopy from December 2015 to July 2017. The primary end point is procedure time, with secondary end point of sedation time. There is a statistically significant difference in time to cecum between moderate sedation (MOD) and MAC (P = 0.002). Operator perceived difficulty is statistically associated with increased time to cecum (P < 0.0001). Time to cecum between MOD and MAC over the levels of difficulty was not significant (P = 0.403). A subanalysis looking at time to cecum between MOD and MAC for each level of difficulty showed a significant effect when difficulty was described as no difficulty. There is a statistically significant difference in time to scope insertion between MOD and MAC (P < 0.0001). Our data show that, taken as a conglomerate, the procedure and sedation time is shorter in MAC than in MOD. The use of MAC is associated with decrease time to scope insertion and overall time to cecum.

Previous literature demonstrates the safety of primary repair in penetrating colon injury requiring resection, without the creation of a diverting ostomy. It is unknown whether a similar approach can be applied to patients with blunt colon injury. The aim of this study was to measure outcomes in patients who underwent colon resection with and without ostomy creation after blunt trauma injury to help direct future management. Using the National Trauma Data Bank for years 2008 to 2012, we identified patients with blunt trauma mechanisms who underwent colectomy. Patients were stratified into two groups: primary anastomosis and diversion with ostomy. Primary outcome was inpatient mortality. Secondary outcomes included length of stay and perioperative complications. All risk-adjusted analyses were performed using logistic regression with consideration of interactions. Five hundred eighty-one observations met our inclusion criteria. Baseline characteristics between the two groups were similar with the exception of age (37.3 vs 42.2 years, P < 0.001) and admission Glasgow coma score (13.2 vs 12.1, P = 0.002). Risk-adjusted mortality for the two groups was not statistically significant (2.3% vs 3.0%, P = 0.63); however, patients with primary anastomosis had a shorter length of stay (18.2 vs 28.1, P < 0.001), fewer days in the intensive care unit (10.9 vs 16.2, P < 0.001), and fewer ventilator days (10.5 vs 14.6, P = 0.01). In patients requiring colon resection after blunt trauma, mortality is not different for those who receive a primary anastomosis versus ostomy. Patients without diversion had shorter hospital stays, intensive care unit days, and ventilator days. These data support that primary anastomosis is safe in this patient population.

In 2014, ACS-NSQIP® targeted pancreatectomies to improve outcome reporting and risk calculation related to pancreatectomy. At the same time, our department began prospectively collecting data for pancreatectomy in the Enhanced Recovery After Surgery® Interactive Audit System (EIAS). The purpose of this study is to compare reported outcomes between two major auditing databases for the same patients undergoing pancreatectomy. The same 171 patients were identified in both databases. Clinical outcomes were then obtained from each database and compared to determine whether reported complication rates were statistically different between auditing databases. A combination of Wilcoxon rank sum and Pearson's chi-squared tests were used to calculate statistical significance. No significant difference was appreciated in captured demographics between EIAS and NSQIP. Significant differences in reported rates for renal dysfunction, postoperative pancreatic fistula, return to the operative room, and urinary tract infection were noted between EIAS and NSQIP. Although significant differences in reported complication rates were demonstrated between EIAS and NSQIP for pancreatectomy, much of the discrepancy is attributable to subtle differences in definitions for postoperative occurrences between the two auditing databases. It is vital for surgeons to understand the exact definition that determines the complication rate for a given database.

We investigated the patterns of injury associated with major midface trauma. Our hypothesis is that midface injuries are associated with a decrease in certain traumatic brain injuries as well as major torso injuries. The registry of our Level I trauma center was queried for all adult patients treated over 25 years from 1989 to 2013. Patients with midface fractures were identified based on the ICD-9 code. Associated injuries were defined based both on individual ICD-9 codes as well as the Barell Injury Matrix. Injury etiology was defined based on e-codes. Univariate analysis was performed using chi-squared test, Fisher's exact test, and Wilcoxon test. A total of 29,152 patients were identified. Excluding pediatric patients, those with exclusively penetrating trauma, and patients with incomplete data, 20,971 patients were included for subsequent analysis. Midface fractures were identified in 752 patients. Patients with Le Fort fractures were more likely to be male, have a higher Injury Severity Score, a lower arrival Glasgow Coma Scale, and more likely to require intensive care unit admission and mechanical ventilation, with a longer hospital length of stay. Patients with midface fractures had significantly fewer subdural hematomas, subarachnoid hemorrhages, spine fractures, and were less likely to have associated abdominal and pelvic injuries. Patients with midface fractures were more likely to require facial reconstruction procedures and craniotomy. Patients presenting with midface fractures after blunt trauma have a distinctly different pattern of injuries. One potential mechanism for this is a deceleration effect, where midface impact and resulting fractures dissipate some of the energy.

The purpose of this study was to assess resource utilization after implementation of a mild traumatic brain injury (TBI) treatment protocol. A retrospective review was conducted of patients with isolated mild TBI before and after implementation of a mild TBI treatment protocol in May 2015. Patients admitted from June 2014 to February 2017, aged 18 to 89 years, presenting with a Glasgow coma score of 13 to 15, with an isolated small intracerebral hemorrhage on CT without midline shift, and not coagulopathic were evaluated. According to the protocol, patients were admitted to a non-intensive care unit (ICU) ward, without routine neurosurgical consultation or repeat head CT unless clinically indicated. Hospital length of stay (LOS), ICU LOS, rate of neurosurgical consultation, rate of repeat head CT within 24 hours of admission, and associated costs were evaluated. Forty-six patients were identified in the preprotocol group and 97 in the protocol group. The protocol group had a shorter hospital LOS (1.46 vs 2.04 days, P = 0.0034), shorter ICU LOS (0.02 vs 0.37 days, P < 0.0001), lower rates of repeat head CT (2.06% vs 39.13%, P < 0.0001), and neurosurgical consultations (1.03% vs 28.26%, P < 0.0001). Decreased charges derived from fewer repeat head CT and neurosurgical consultations were observed from $43.98 to $844.04 per patient. There were no inpatient mortalities and no progressions of injury requiring unplanned admission to the ICU or operative intervention. Efficient delivery of care is paramount in modern medicine and this study demonstrates that the mild TBI treatment protocol significantly decreased resource utilization without jeopardizing patient safety.

In 2015, the United States Navy hospital ship (USNS) COMFORT, deployed to 11 Caribbean and Latin American countries over a six-month period to provide humanitarian civic assistance. Personnel from the United States Navy and multiple nongovernmental organizations collaborated to offer surgical and medical care. Data from past deployments aid in planning for future missions by prioritizing finite resources and maximizing care. The data analyzed included all patients evaluated and treated by the Directorate of Surgical Services of the USNS COMFORT between April and September 2015. Comparative and descriptive statistics were performed to analyze patient demographics, surgical subspecialty performing the procedures, types of general and pediatric surgical procedures performed, operative times, and complication rates. Of the 1256 surgical cases performed aboard USNS COMFORT during CP15, 24.8 per cent were general surgery cases, followed by 16 per cent ophthalmology, 10.6 per cent pediatric surgery, 10 per cent plastic surgery, and eight additional specialties with <10 per cent of the cases each. Total operative time was 1253 hours with a total room time of 1896.5 hours. The identified complication rate was 1.99 per cent across all specialties. The USNS COMFORT platform offers the unique capability to provide humanitarian surgical assistance. Reporting these data demonstrate that there is a need for humanitarian assistance and this can be provided safely through the Continuing Promise mission. Future deployments may target resources toward the surgical services with higher volumes, which were general surgery, ophthalmology, pediatric surgery, and plastic surgery.

Identifying disparities in surgical outcomes among patient populations may help hospitals target patients at highest risk for complications. The South Carolina Surgical Quality Collaborative (SCSQC) is a regional collaborative made up of eight facilities whose goal is to improve the quality and value of general surgical care in South Carolina. Using SCSQC data, we reviewed colon surgery outcomes to determine whether disparities exist between specific patient populations. SCSQC colon surgery data were reviewed from August 2015 to August 2017. SSI, length of stay, return to the ED, and reoperation rates were used as outcome measures. They were evaluated in patient populations stratified by gender, race (white, black, and other), and age (<50, 50–70, and >70 years). A total of 2611 patients were included in this study. Statistically significant differences in outcomes were identified between white and black patients in length of stay (6.0 vs 7.5 days, P < 0.0001) and return to the ED (8.1% vs 14.7%, P < 0.0001), but not in SSI (6.4% vs 6.8%, P = 0.8839) or reoperation rates (6.4% vs 8.4%, P = 0.1886). Length of stay increased with increasing age (4.1 vs 7.1 vs 8.8, P < 0.0001). SSI varied by age (4.0% vs 8.2% vs 6.4%, P = 0.0005), as did return to the ED (11.2% vs 9.7% vs 769%, P = 0.0987) and reoperation rates (4.5% vs 8.1% vs 8.2%, P = 0.0034). SCSQC data indicate that race and age may place patients at risk for negative outcomes after colorectal surgery.

Decompressive craniectomy (DC) is a surgical modality sometimes used in the management of elevated intracranial pressure. Questions remain as to its long-term benefits in traumatic brain injury patients. The extended Glasgow Outcome Scale (eGOS) is a scoring system based on a structured interview that allows for consistent and reproducible measurement of long-term functional outcomes. The purpose of this study was to determine the eGOS score of post-craniectomy patients after discharge and stratify survivors based on outcome. A multicenter review of patients who underwent DC was performed. Survivors underwent a phone survey at which time the eGOS was calculated. Patients with an eGOS ≥ 5 were considered to have a good functional outcome. Fifty-four patients underwent DC. Age (OR 1.038; confidence interval 1.003–1.074) and Glasgow Coma Scale (OR 0677; confidence interval 0.527–0.870) were predictors of mortality. Patients who were available for follow-up (n = 13) had poor functional outcomes at discharge (eGOS = 3); however, this improved at the time of follow-up survey (eGOS = 5; P = 0.005). DC is a controversial operation with high mortality and uncertain benefit. Among our cohort, the eGOS score was significantly higher at follow-up survey than it was at discharge. Although the mortality was high, if patients survived to discharge, most had a good functional outcome at follow-up survey.

The significance and management of melanoma in situ (MIS) at the margin of excision of invasive melanoma is debated. Patients undergoing excision of invasive melanoma from 2000 to 2016 with MIS at the margin were identified. A cohort without MIS was matched for age, gender, location, and Breslow depth. Thirty-two patients with 33 cases of MIS at the margin were identified. Melanoma was located on the head/neck (66.7%), extremities (24.2%), and trunk (9.1%). Median Breslow depth was 1.0 mm (range 0.25–10.80). Margin treatment included re-excision (45.5%), re-excision plus imiquimod (3.0%), imiquimod alone (9.1%), and observation alone (42.4%). At a median follow-up of 91 months (range 28–126), five patients (15.2%) with a median Breslow depth of 4.75 mm (range 1.10–6.70) developed local recurrence (LR). Three underwent re-excision of the positive margin and two were observed. Intervention for positive margins did not decrease LR compared with observation (P = 0.905, OR = 1.125, 95% confidence interval [CI] 0.162–7.824). All five patients with LR were alive at the last follow-up. There were two recurrences in the matched cohort (6.1%); both were alive at the last follow-up. Risk of LR is higher with MIS at the margin, but this does not seem to impact survival. Larger studies may elucidate predictive factors and interventions that decrease risk for LR.

Appendectomy for presumed appendicitis is the most common surgical emergency during pregnancy. Delayed diagnosis and treatment of appendicitis carries risk for the fetus and mother. We sought to evaluate the accuracy of MRI in pregnant patients with suspected appendicitis. All pregnant patients with suspected appendicitis between January 2014 and April 2016 were included. MRI reports were categorized into positive, negative, and inconclusive groups. Diagnosis of appendicitis was based on pathology report. Fifty-two patients were included in the study. The MRI was positive in two, negative in 29, and inconclusive in 21 patients. Twelve patients had surgery, 11 of which had positive appendicitis on pathology. Both positive MRI patients had appendicitis. In the negative MRI group, 3 of 29 (10%) had appendicitis. In the inconclusive MRI group, 6 of 21 (29%) had appendicitis. A positive MRI result was very specific with a 100 per cent positive predictive value; however, the sensitivity was as low as 18 per cent (diagnosed only 2 of 11 cases). Although a positive MRI finding was reliable in making a decision to operate, a negative or inconclusive MRI was not. In patients with a high clinical suspicion of appendicitis, surgery should still be considered even without definitive positive MRI findings.

There is no standard method for closure of an appendiceal stump during laparoscopic appendectomy. This study compares stump closure using a stapler with closure using an Endoloop ligature. The charts of all patients who underwent laparoscopic appendectomy at a single tertiary care center over a two-year period were reviewed for demographics, comobidities, operative details and costs, and outcomes. There were 325 patients who underwent a laparoscopic appendectomy. The majority, 250 (77%), underwent stump closure with a stapler. They were equivalent in demographics and postoperative complication rates. Cases using an Endoloop were slightly faster in terms of procedure time and room time, and less expensive in terms of operative supply cost. The price difference is not explained by time saved in the operating room and more likely by the equipment price.

Propofol infusion syndrome (PIS) is a potentially lethal complication of propofol marked by rhabdomyolysis, metabolic acidosis, and cardiac arrhythmias or collapse. The objective of this study was to determine the effectiveness of a prospective screening protocol to prevent PIS. All trauma patients admitted who received propofol as a continuous infusion were prospectively screened from November 1, 2013 to December 31, 2015. Variables studied included demographics, injury severity, laboratory values, infusion rates, and mortality. Serum creatine phosphokinase (CPK) and lactate were drawn daily. Propofol was stopped for a positive screen defined as an increase in CPK to greater than 5000 IU/L or lactate greater than 4 mmol/L. Positive and negative cohorts were compared. Two hundred and twenty-five patients met the inclusion criteria and 12 patients (5.3%) had propofol stopped because of elevated CPK. No differences were identified in demographics, transfusions, injury severity, hospital length of stay, or propofol dose. The positive screened group had longer intensive care unit length of stay (20 vs 13 days; P = 0.002) and increased vent days (14.5 vs 10 days; P = 0.008). Max serum osmolality (334 vs 305 mosm/kg; P = 0.049) and max serum CPK (6782 vs 1058 IU/L; P < 0.0001) were higher in the positive cohort. No cases of PIS occurred, and mortality (16.7 vs 15.5%; P = 0.999) was not different between the cohorts. The screening protocol was effective in eliminating PIS. Serial CPK evaluations provided an effective screening tool and serum lactate can be dropped from screening.

Bacteremia is a worrisome postoperative complication and blood cultures (BCx) are often nondiagnostic. We previously reported a 4 per cent overall yield of positive cultures in postoperative patients. To reduce unnecessary testing, we present a predictive model to identify patients in whom growth of pathogens is unlikely and provide a clinical decision-making guide. Retrospective analysis of nonpregnant patients ≥18 years who had BCx within 10 days postoperatively was performed. Generalized linear mixed models identified clinical predictors of high- and low-yield cultures. A clinical algorithm was created using significant predictors, and positive predictive value, negative predictive value, sensitivity, and specificity calculated. Among 1759 BCx, hypotension, maximum temperature ≥101.5 °F within 24 hours of culture, and culture collected after postoperative day (POD) two were statistically significant predictors of positive cultures. Forty nine per cent of BCx were sent ≤ POD 2, and <1 per cent of these were positive. When all three criteria were met, the probability of a positive culture increased to 17 per cent. When absent, the probability of a negative culture was 99 per cent. When applied to the initial data set, the model resulted in 85 per cent reduction of cultures with 9 per cent yield of positive cultures. Drawing BCx based on a single predictor is inadequate. Reducing the number of cultures reflexively ordered within the first two POD could significantly reduce the number of unnecessary BCx. Several clinical features identified patients most likely to have positive BCx within the first 10 POD and could reduce unnecessary BCx. This model should be validated in an independent, prospective cohort.

With the advent of proton pump inhibitors and H. Pylori treatment, the old dogma “the most common cause of lower GI bleeding is upper GI bleeding” may no longer be valid. We sought to determine the most common causes of GI bleeding in patients without an obvious source and their clinical outcomes. We queried our hospital database for GI hemorrhage during 2015, excluding patients with obvious sources such as hematemesis or anal pathology. We collected data from patients with GI bleeding defined as bright red blood per rectum, melena, or a positive fecal occult blood test. The primary endpoints were etiology of GI bleed, amount of transfusions required, and types of interventions performed. Ninety-three patients were admitted with GI bleeding as defined previously: mean age was 74 years and mean hemoglobin was 8.2. Seventy-four per cent received blood transfusions with an average of 2 units transfused per patient; 22 per cent received 3 or more units of blood. The etiology of bleeding was 17 per cent upper GI source, 15 per cent lower GI source, and in 68 per cent, the source remained unknown. Bleeding stopped spontaneously in 86 per cent of patients and 9 per cent died. Endoscopy was performed in 71 per cent, but only 6 per cent underwent therapeutic endoscopic intervention. No patient had surgical or interventional radiologic procedures related to their GI bleed. Gastrointestinal bleeding, without an obvious source on presentation, rarely requires operative intervention or interventional radiologic procedure. Blood transfusions were not predictive of the need for therapeutic endoscopic intervention which was required in only 6 per cent of patients.

We aim to investigate the relation between depression, posttraumatic stress disorder (PTSD), and inflammatory factors in patients with severe burn injury. Psychological assessment was carried out using PTSD checklist (PCL) involving a 17-item self-report questionnaire (PCL-17) and the Hamilton Rating Scale for depression (HAMD-24). The serum IL-1β, IL-6, IL-8, and tumor necrosis factor-α (TNF-α) were determined using enzyme-linked immunosorbent assay. Correlation analysis was performed to analyze the correlation between the factors and scores of PTSD and depression. Compared with the PCL-17 score, HAMD-24 score, and inflammatory factors at month 3, a significant decrease was noticed in the PCL-17 score, HAMD-24 score, and inflammatory factors at months 6 and 9 (P < 0.01). For the HAMD-24 score, significant improvements were noticed in the anxiety/somatization, cognitive disorder, blocking, sleep disorders, and depression at months 3, 6, and 9. The levels of IL-1β, IL-8, and TNF-α were positively correlated with the PCL-17 score (P < 0.05). The levels of IL-1β, IL-6, IL-8, and TNF-α were positively correlated with the HAMD-24 score (P < 0.05). Patients with severe burn injury showed obvious stress alternation displaying specific depression-related characteristics, and inflammation may involve in the pathogenesis of PTSD and depression in burn patients.

The intent of this study was to characterize the management and subsequent complications of combat injury tube thoracostomies and to determine risk factors for the development of pneumonia (PNA) and retained hemothorax (RH). One hundred fifteen patients with 173 tube thoracostomies met the inclusion criteria and were analyzed. The mean injury severity score was 30.8 1 11.6, 23.5 per cent had traumatic amputations, 49.7 per cent had a hemothorax, and 50.3 per cent had a pneumothorax as indications for tube thoracostomy (TT) placement. Within 24 hours of injury, 89.6 per cent were intubated, the majority (54%) were injured by improvised explosive devices, 35.6 per cent sustained rib fractures, and 12.2 per cent had a diaphragm injury. A mean of 1.5 1 0.7(range 1–4) tube thoracostomies were placed, 18.3 per cent of patients had bilateral tube thoracostomies, and the average TT duration was 6.7 1 3.9 days. The incidence of PNA was 27 per cent (n = 31), RH was 9.6 per cent (n = 11), and empyema was 1.7 per cent (n = 2). Multivariable analysis identified the duration of ventilation [OR 1.2, 95% confidence interval (CI): 1.097–1.313, P < 0.001] as independently associated with the development of PNA. Bilateral TT placement (OR 3.848, 95% CI: 1.219–12.143, P = 0.0216) and injury severity score (OR 1.050, 95% CI: 1.001–1.102, P = 0.0443) were independently associated with PNA development when a patient was intubated for eight days or less. The number of tube thoracostomies placed (OR 3.08, 95% CI: 1.03–9.18, P = 0.0439) was independently associated with the development of RH. Further research is warranted to identify modifiable risk factors to reduce the incidence of PNA and RH in patients with TT placed for traumatic injuries.

This study was performed to assess our institution's experience with stab injuries to the posterior mediastinal box. We examine the value of performing CT of the chest and esophagram in conjunction with a chest X-ray (CXR) over performing CXR(s) alone in evaluating this group of patients. We performed a retrospective study covering a 10-year period consisting of patients with stab wounds to the posterior mediastinal box. Age, gender, and injury severity score as demographic data points were collected. CXR, CT, and esophagram results; identified injuries; and subsequent interventions were analyzed. Of 78 patients who met the inclusion criteria, a total of 55 patients underwent esophagram, one had a false-positive result, and zero had their course altered by the study. Sixty-six patients underwent CT imaging, and there were nine missed findings on initial CXR. Five of these were clinically insignificant and the remaining four were managed with a chest tube alone. There were no tracheobronchial, esophageal, cardiac, or great vessel injuries. Hemodynamically stable, asymptomatic patients with stab wounds to the posterior mediastinal box do not require routine CT and esophagram in the absence of CXR and cardiac ultrasonographic abnormalities.

Level I trauma centers serve as a community resource, with most centers using an inclusive transfer policy that may result in overtriage. The financial burden this imparts on an urban trauma system has not been well examined. We sought to examine the incidence of secondary overtriage (SOT) at an urban Level I trauma center. This was a retrospective study from an urban Level I trauma center examining patients admitted as trauma transfers (TT) from 2010 to 2014. SOT was defined as patients not meeting the “Orange Book” transfer criteria and who had a length of stay of <48 hours. Average ED and transport charges were calculated for total transfer charges. A total of 2397 TT were treated. The number of TT increased over the study interval. The mean age of TT was 59.7 years (SD ± 26.4 years); patients were predominantly male (59.2%), white (83.2%), with at least one comorbidity (71.5%). Blunt trauma accounted for 96.8 per cent of admissions with a median Injury Severity Score of nine (IQR: 5–16). Predominant injuries were isolated closed head trauma (61.4%), skin/soft tissue injury (18.9%), and spinal injury (17.6%). SOT was 48.2 per cent and increased yearly (P < 0.001). The median trauma center charge for SOT was ($27,072; IQR: $20,089–34,087), whereas ED charges were ($40,440; IQR: $26,150–65,125), resulting in a total cost of $67,512/patient. A liberal TT policy results in a high SOT rate adding significant unnecessary costs to the health-care system. Efforts to establish transfer guidelines may allow for significant cost savings without compromising care.

To date, no studies have examined the relationship between geographic and socioeconomic factors and the frequency of pedestrians sustaining traumatic injuries from a motor vehicle. The objective of this study was to analyze the impact of location on the frequency of pedestrian injury by motor vehicle. The University of Louisville Trauma Registry was queried for patients who had been struck by a motor vehicle from 2010 to 2015. Demographic and injury information as well as outcome measures were evaluated to identify those impacting risk of pedestrian versus motor vehicle accidents. Number of incidents was correlated with lower median household income. There was also a moderate correlation between number of incidents and population density. Multivariable analysis demonstrated a significant association between increased median household income and distance from downtown Louisville and decreased risk of death after pedestrian versus motor vehicle accident. Incidence of pedestrian injury by motor vehicles is influenced by regional socioeconomic status. Efforts to decrease the frequency of these events should include further investigation into the mechanisms underpinning this relationship.

Trauma centers are legally bound by Emergency Medical Treatment and Active Labor Act to provide equal treatment to trauma patients, regardless of payer source. However, evidence has suggested that disparities in trauma care exist. This study investigated the relationships between payer source and procedures (total, diagnostic, and surgical) and the number of medical consults in an adult trauma population. This is a 10-year retrospective trauma registry study at a Level I trauma facility. Payer source of adult trauma patients was identified, demographics and variables associated with trauma outcomes were abstracted, and multivariate logistic regression tests were used to determine statistical differences in the number of procedures and medical consults. Of the 12,870 records analyzed, 69.1 per cent of patients were commercially insured, 21.2 per cent were uninsured, and 9.6 per cent had Medicaid. After controlling for patient- and injury-related variables, the commercially insured received more total procedures (4.30) than the uninsured (3.35) or those with Medicaid (3.34), and more diagnostic (2.59) procedures than the uninsured (2.03) or those with Medicaid (2.04). There was not a difference in the number of surgical procedures or medical consults among payer sources. This study noted that disparities (measured by the number of procedures received) compared by payer source existed in the care of trauma patients. However, for medical consults and definitive care (measured by surgical procedures), disparities were not observed. Future research should focus on secondary factors that influence levels of care such as patient-level factors (health literacy) and trauma program policies.

The purpose of our study was to evaluate the rate of ventral hernia repair (VHR) after open abdominal aortic anneurysm in New York State compared with the rate of VHR after open abdominal aortic bypass procedures. The Statewide Planning and Research Cooperative System database was queried for all abdominal aortic aneurysm (AAA) and bypass procedures performed between 2000 and 2010. Social security death index was used to identify patients who died. The cause-specific Cox proportional hazard model was applied to compare the risk of having follow-up VHR between patients with AAA and bypass with death as a competing risk event. A multivariable model was used to explore independent relationship with the risk of having follow-up ventral hernia after adjusting for other factors. There were 9314 patients who underwent open AAA repair, 739 (7.93%) of which had subsequent VHR. Comparatively, 8280 patients underwent aortofemoral or aortoiliac bypass procedures, with 480 (5.8%) undergoing subsequent VHR. The observed one-year, five-year, and 10-year VHR rates for AAA versus bypass were 2.8 versus 1.8 per cent, 10.0 versus 8.0 per cent, 10.7 versus 9.38 per cent, respectively. After controlling for all other factors, patients undergoing AAA repair were more likely and elderly patients were less likely to undergo VHR (P < 0.0001). Patients with serious comorbid conditions such as valvular disease, diabetes mellitus, and neurologic disorders were less likely to undergo subsequent VHR controlling for other factors. VHR after AAA procedures is more common compared with bypass procedures for occlusive disease. Because this patient population has significant comorbidity, prophylactic mesh placement may play a role in preventing necessity for future procedures.