Safety of the injectable expanding biopolymer foam for non-compressible truncal bleeding in swine.

BACKGROUND: A novel hydrophobically modified chitosan (hm-chitosan) polymer has been previously shown to improve survival in a non-compressible intra-abdominal bleeding model in swine. We performed a 28-day survival study to evaluate the safety of the hm-chitosan polymer in swine. METHODS: Female Yorkshire swine (40-50 kg) were used. A mild, non-compressible, closed-cavity bleeding model was created with splenic transection. The hm-chitosan polymer was applied intra-abdominally through an umbilical nozzle in the same composition and dose previously shown to improve survival. Animals were monitored intraoperatively and followed 28 days postoperatively for survival, signs of pain, and end-organ function. Gross pathological and microscopic evaluations were performed at the conclusion of the experiment. RESULTS: A total of 10 animals were included (hm-chitosan = 8; control = 2). The 2 control animals survived through 28 days, and 7 of the 8 animals from the hm-chitosan group survived without any adverse events. One animal from the hm-chitosan group required early termination of the study for signs of pain, and superficial colonic ulcers were found on autopsy. Laboratory tests showed no signs of end-organ dysfunction after exposure to hm-chitosan after 28 days. On gross pathological examination, small (<0.5 cm) peritoneal nodules were noticed in the hm-chitosan group, which were consistent with giant-cell foreign body reaction in microscopy, presumably related to polymer remnants. Microscopically, no signs of systemic polymer embolization or thrombosis were noticed. CONCLUSION: Prolonged intraperitoneal exposure to the hm-chitosan polymer was tolerated without any adverse event in the majority of animals. In the single animal that required early termination, the material did not appear to be associated with end-organ dysfunction in swine. Superficial colonic ulcers that would require surgical repair were identified in 1 out of 8 animals exposed to hm-chitosan.

Use of artificial intelligence for nonlinear benchmarking of surgical care.

BACKGROUND: Existent methodologies for benchmarking the quality of surgical care are linear and fail to capture the complex interactions of preoperative variables. We sought to leverage novel nonlinear artificial intelligence methodologies to benchmark emergency surgical care. METHODS: Using a nonlinear but interpretable artificial intelligence methodology called optimal classification trees, first, the overall observed mortality rate at the index hospital's emergency surgery population (index cohort) was compared to the risk-adjusted expected mortality rate calculated by the optimal classification trees from the American College of Surgeons National Surgical Quality Improvement Program database (benchmark cohort). Second, the artificial intelligence optimal classification trees created different "nodes" of care representing specific patient phenotypes defined by the artificial intelligence optimal classification trees without human interference to optimize prediction. These nodes capture multiple iterative risk-adjusted comparisons, permitting the identification of specific areas of excellence and areas for improvement. RESULTS: The index and benchmark cohorts included 1,600 and 637,086 patients, respectively. The observed and risk-adjusted expected mortality rates of the index cohort calculated by optimal classification trees were similar (8.06% [95% confidence interval: 6.8-9.5] vs 7.53%, respectively, P = .42). Two areas of excellence and 4 for improvement were identified. For example, the index cohort had lower-than-expected mortality when patients were older than 75 and in respiratory failure and septic shock preoperatively but higher-than-expected mortality when patients had respiratory failure preoperatively and were thrombocytopenic, with an international normalized ratio ≤1.7. CONCLUSION: We used artificial intelligence methodology to benchmark the quality of emergency surgical care. Such nonlinear and interpretable methods promise a more comprehensive evaluation and a deeper dive into areas of excellence versus suboptimal care.

Development of a Machine Learning-Based Prescriptive Tool to Address Racial Disparities in Access to Care After Penetrating Trauma.

Importance: The use of artificial intelligence (AI) in clinical medicine risks perpetuating existing bias in care, such as disparities in access to postinjury rehabilitation services. Objective: To leverage a novel, interpretable AI-based technology to uncover racial disparities in access to postinjury rehabilitation care and create an AI-based prescriptive tool to address these disparities. Design, Setting, and Participants: This cohort study used data from the 2010-2016 American College of Surgeons Trauma Quality Improvement Program database for Black and White patients with a penetrating mechanism of injury. An interpretable AI methodology called optimal classification trees (OCTs) was applied in an 80:20 derivation/validation split to predict discharge disposition (home vs postacute care [PAC]). The interpretable nature of OCTs allowed for examination of the AI logic to identify racial disparities. A prescriptive mixed-integer optimization model using age, injury, and gender data was allowed to "fairness-flip" the recommended discharge destination for a subset of patients while minimizing the ratio of imbalance between Black and White patients. Three OCTs were developed to predict discharge disposition: the first 2 trees used unadjusted data (one without and one with the race variable), and the third tree used fairness-adjusted data. Main Outcomes and Measures: Disparities and the discriminative performance (C statistic) were compared among fairness-adjusted and unadjusted OCTs. Results: A total of 52 468 patients were included; the median (IQR) age was 29 (22-40) years, 46 189 patients (88.0%) were male, 31 470 (60.0%) were Black, and 20 998 (40.0%) were White. A total of 3800 Black patients (12.1%) were discharged to PAC, compared with 4504 White patients (21.5%; P < .001). Examining the AI logic uncovered significant disparities in PAC discharge destination access, with race playing the second most important role. The prescriptive fairness adjustment recommended flipping the discharge destination of 4.5% of the patients, with the performance of the adjusted model increasing from a C statistic of 0.79 to 0.87. After fairness adjustment, disparities disappeared, and a similar percentage of Black and White patients (15.8% vs 15.8%; P = .87) had a recommended discharge to PAC. Conclusions and Relevance: In this study, we developed an accurate, machine learning-based, fairness-adjusted model that can identify barriers to discharge to postacute care. Instead of accidentally encoding bias, interpretable AI methodologies are powerful tools to diagnose and remedy system-related bias in care, such as disparities in access to postinjury rehabilitation care.

Artificial intelligence versus surgeon gestalt in predicting risk of emergency general surgery.

BACKGROUND: Artificial intelligence (AI) risk prediction algorithms such as the smartphone-available Predictive OpTimal Trees in Emergency Surgery Risk (POTTER) for emergency general surgery (EGS) are superior to traditional risk calculators because they account for complex nonlinear interactions between variables, but how they compare to surgeons' gestalt remains unknown. Herein, we sought to: (1) compare POTTER to surgeons' surgical risk estimation and (2) assess how POTTER influences surgeons' risk estimation. STUDY DESIGN: A total of 150 patients who underwent EGS at a large quaternary care center between May 2018 and May 2019 were prospectively followed up for 30-day postoperative outcomes (mortality, septic shock, ventilator dependence, bleeding requiring transfusion, pneumonia), and clinical cases were systematically created representing their initial presentation. POTTER's outcome predictions for each case were also recorded. Thirty acute care surgeons with diverse practice settings and levels of experience were then randomized into two groups: 15 surgeons (SURG) were asked to predict the outcomes without access to POTTER's predictions while the remaining 15 (SURG-POTTER) were asked to predict the same outcomes after interacting with POTTER. Comparing to actual patient outcomes, the area under the curve (AUC) methodology was used to assess the predictive performance of (1) POTTER versus SURG, and (2) SURG versus SURG-POTTER. RESULTS: POTTER outperformed SURG in predicting all outcomes (mortality-AUC: 0.880 vs. 0.841; ventilator dependence-AUC: 0.928 vs. 0.833; bleeding-AUC: 0.832 vs. 0.735; pneumonia-AUC: 0.837 vs. 0.753) except septic shock (AUC: 0.816 vs. 0.820). SURG-POTTER outperformed SURG in predicting mortality (AUC: 0.870 vs. 0.841), bleeding (AUC: 0.811 vs. 0.735), pneumonia (AUC: 0.803 vs. 0.753) but not septic shock (AUC: 0.712 vs. 0.820) or ventilator dependence (AUC: 0.834 vs. 0.833). CONCLUSION: The AI risk calculator POTTER outperformed surgeons' gestalt in predicting the postoperative mortality and outcomes of EGS patients, and when used, improved the individual surgeons' risk prediction. Artificial intelligence algorithms, such as POTTER, could prove useful as a bedside adjunct to surgeons when preoperatively counseling patients. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level II.

POTTER-ICU: An artificial intelligence smartphone-accessible tool to predict the need for intensive care after emergency surgery.

BACKGROUND: Delays in admitting high-risk emergency surgery patients to the intensive care unit result in worse outcomes and increased health care costs. We aimed to use interpretable artificial intelligence technology to create a preoperative predictor for postoperative intensive care unit need in emergency surgery patients. METHODS: A novel, interpretable artificial intelligence technology called optimal classification trees was leveraged in an 80:20 train:test split of adult emergency surgery patients in the 2007-2017 American College of Surgeons National Surgical Quality Improvement Program database. Demographics, comorbidities, and laboratory values were used to develop, train, and then validate optimal classification tree algorithms to predict the need for postoperative intensive care unit admission. The latter was defined as postoperative death or the development of 1 or more postoperative complications warranting critical care (eg, unplanned intubation, ventilator requirement ≥48 hours, cardiac arrest requiring cardiopulmonary resuscitation, and septic shock). An interactive and user-friendly application was created. C statistics were used to measure performance. RESULTS: A total of 464,861 patients were included. The mean age was 55 years, 48% were male, and 11% developed severe postoperative complications warranting critical care. The Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application was created as the user-friendly interface of the complex optimal classification tree algorithms. The number of questions (ie, tree depths) needed to predict intensive care unit admission ranged from 2 to 11. The Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application had excellent discrimination for predicting the need for intensive care unit admission (C statistics: 0.89 train, 0.88 test). CONCLUSION: We recommend the Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application as an accurate, artificial intelligence-based tool for predicting severe complications warranting intensive care unit admission after emergency surgery. The Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application can prove useful to triage patients to the intensive care unit and to potentially decrease failure to rescue in emergency surgery patients.

Do not forget the platelets: The independent impact of red blood cell to platelet ratio on mortality in massively transfused trauma patients.

BACKGROUND: Balanced blood component administration during massive transfusion is standard of care. Most literature focuses on the impact of red blood cell (RBC)/fresh frozen plasma (FFP) ratio, while the value of balanced RBC:platelet (PLT) administration is less established. The aim of this study was to evaluate and quantify the independent impact of RBC:PLT on 24-hour mortality in trauma patients receiving massive transfusion. METHODS: Using the 2013 to 2018 American College of Surgeons Trauma Quality Improvement Program database, adult patients who received massive transfusion (≥10 U of RBC/24 hours) and ≥1 U of RBC, FFP, and PLT within 4 hours of arrival were retrospectively included. To mitigate survival bias, only patients with consistent RBC:PLT and RBC:FFP ratios between 4 and 24 hours were analyzed. Balanced FFP or PLT transfusions were defined as having RBC:PLT and RBC:FFP of ≤2, respectively. Multivariable logistic regression was used to compare the independent relationship between RBC:FFP, RBC:PLT, balanced transfusion, and 24-hour mortality. RESULTS: A total of 9,215 massive transfusion patients were included. The number of patients who received transfusion with RBC:PLT >2 (1,942 [21.1%]) was significantly higher than those with RBC:FFP >2 (1,160 [12.6%]) (p < 0.001). Compared with an RBC:PLT ratio of 1:1, a gradual and consistent risk increase was observed for 24-hour mortality as the RBC:PLT ratio increased (p < 0.001). Patients with both FFP and PLT balanced transfusion had the lowest adjusted risk for 24-hour mortality. Mortality increased as resuscitation became more unbalanced, with higher odds of death for unbalanced PLT (odds ratio, 2.48 [2.18-2.83]) than unbalanced FFP (odds ratio, 1.66 [1.37-1.98]), while patients who received both FFP and PLT unbalanced transfusion had the highest risk of 24-hour mortality (odds ratio, 3.41 [2.74-4.24]). CONCLUSION: Trauma patients receiving massive transfusion significantly more often have unbalanced PLT rather than unbalanced FFP transfusion. The impact of unbalanced PLT transfusion on 24-hour mortality is independent and potentially more pronounced than unbalanced FFP transfusion, warranting serious system-level efforts for improvement. LEVEL OF EVIDENCE: Therapeutic/Care Management; Level IV.

Treatment and outcomes of anticoagulated geriatric trauma patients with traumatic intracranial hemorrhage after falls.

INTRODUCTION: Emergency physicians and trauma surgeons are increasingly confronted with pre-injury direct oral anticoagulants (DOACs). The objective of this study was to assess if pre-injury DOACs, compared to vitamin K antagonists (VKA), or no oral anticoagulants is independently associated with differences in treatment, mortality and inpatient rehabilitation requirement. METHODS: We performed a review of the prospectively maintained institutional trauma registry at an urban academic level 1 trauma center. We included all geriatric patients (aged ≥ 65 years) with tICH after a fall, admitted between January 2011 and December 2018. Multivariable logistic regression analysis controlling for demographics, comorbidities, vital signs, and tICH types were performed to identify the association between pre-injury anticoagulants and reversal agent use, neurosurgical interventions, inhospital mortality, 3-day mortality, and discharge to inpatient rehabilitation. RESULTS: A total of 1453 tICH patients were included (52 DOAC, 376 VKA, 1025 control). DOAC use was independently associated with lower odds of receiving specific reversal agents [odds ratio (OR) 0.28, 95% confidence interval (CI) 0.15-0.54] than VKA patients. DOAC use was independently associated with requiring neurosurgical intervention (OR 3.14, 95% CI 1.36-7.28). VKA use, but not DOAC use, was independently associated with inhospital mortality, or discharge to hospice care (OR 1.62, 95% CI 1.15-2.27) compared to controls. VKA use was independently associated with higher odds of discharge to inpatient rehabilitation (OR 1.41, 95% CI 1.06-1.87) compared to controls. CONCLUSION: Despite the higher neurosurgical intervention rates, patients with pre-injury DOAC use were associated with comparable rates of mortality and discharge to inpatient rehabilitation as patients without anticoagulation exposure. Future research should focus on risk assessment and stratification of DOAC-exposed trauma patients.

Chitosan-based lifefoam improves survival in lethal noncompressible abdominal bleeding in swine.

BACKGROUND: In military combat settings, noncompressible closed cavity exsanguination is the leading cause of potentially survivable deaths, with no effective treatment available at point of injury. The aim of this study was to assess whether an expanding foam based on hydrophobically modified chitosan (hm-chitosan) may be used as a locally injectable hemostatic agent for the treatment of noncompressible bleeding in a swine model. METHODS: A closed-cavity, grade V hepato-portal injury was created in all animals resulting in massive noncoagulopathic, noncompressible bleeding. Animals received either fluid resuscitation alone (control, n = 8) or fluid resuscitation plus intraperitoneal hm-chitosan agent through an umbilical port (experimental, n = 18). The experiment was terminated at 180 minutes or death (defined as end-tidal CO2 <8mmHg or mean arterial pressure [MAP] <15mmHg), whichever came first. RESULTS: All animals had profound hypotension and experienced a near-arrest from hypovolemic shock (mean MAP = 24 mmHg at 10 minutes). Mean survival time was higher than 150 minutes in the experimental arm versus 27 minutes in the control arm (P < .001). Three-hour survival was 72% in the experimental group and 0% in the control group (P = .002). Hm-chitosan stabilized rising lactate, preventing acute lethal acidosis. MAP improved drastically after deployment of the hm-chitosan and was preserved at 60 mmHg throughout the 3 hours. Postmortem examination was performed in all animals and the hepatoportal injuries were anatomically similar. CONCLUSION: Intraperitoneal administration of hm-chitosan-based foam for massive, noncompressible abdominal bleeding improves survival in a lethal, closed-cavity swine model. Chronic safety and toxicity studies are required.

Risk factors for ischemic gastrointestinal complications in patients undergoing open cardiac surgical procedures: A single-center retrospective experience.

BACKGROUND: Ischemic gastrointestinal complications (IGIC) following cardiac surgery are associated with high morbidity and mortality and remain difficult to predict. We evaluated perioperative risk factors for IGIC in patients undergoing open cardiac surgery. METHODS: All patients that underwent an open cardiac surgical procedure at a tertiary academic center between 2011 and 2017 were included. The primary outcome was IGIC, defined as acute mesenteric ischemia necessitating a surgical intervention or postoperative gastrointestinal bleeding that was proven to be of ischemic etiology and necessitated blood product transfusion. A backward stepwise regression model was constructed to identify perioperative predictors of IGIC. RESULTS: Of 6862 patients who underwent cardiac surgery during the study period, 52(0.8%) developed IGIC. The highest incidence of IGIC (1.9%) was noted in patients undergoing concomitant coronary artery, valvular, and aortic procedures. The multivariable regression identified hypertension (odds ratio [OR] = 5.74), preoperative renal failure requiring dialysis (OR = 3.62), immunocompromised status (OR = 2.64), chronic lung disease (OR = 2.61), and history of heart failure (OR = 2.03) as independent predictors for postoperative IGIC. Pre- or intraoperative utilization of intra-aortic balloon pump or catheter-based assist devices (OR = 4.54), intraoperative transfusion requirement of >4 RBC units(OR = 2.47), and cardiopulmonary bypass > 180 min (OR = 2.28) were also identified as independent predictors for the development of IGIC. CONCLUSIONS: We identified preoperative and intraoperative risk factors that independently increase the risk of developing postoperative IGIC after cardiac surgery. A high index of suspicion must be maintained and any deviation from the expected recovery course in patients with the above-identified risk factors should trigger an immediate evaluation with the involvement of the acute care surgical team.

Intraoperative Deaths: Who, Why, and Can We Prevent Them?

INTRODUCTION: Intraoperative deaths (IODs) are rare but catastrophic. We systematically analyzed IODs to identify clinical and patient safety patterns. METHODS: IODs in a large academic center between 2015 and 2019 were included. Perioperative details were systematically reviewed, focusing on (1) identifying phenotypes of IOD, (2) describing emerging themes immediately preceding cardiac arrest, and (3) suggesting interventions to mitigate IOD in each phenotype. RESULTS: Forty-one patients were included. Three IOD phenotypes were identified: trauma (T), nontrauma emergency (NT), and elective (EL) surgery patients, each with 2 sub-phenotypes (e.g., ELm and ELv for elective surgery with medical arrests or vascular injury and bleeding, respectively). In phenotype T, cardiopulmonary resuscitation was initiated before incision in 42%, resuscitative thoracotomy was performed in 33%, and transient return of spontaneous circulation was achieved in 30% of patients. In phenotype NT, ruptured aortic aneurysms accounted for half the cases, and median blood product utilization was 2,694 mL. In phenotype ELm, preoperative evaluation did not include electrocardiogram in 12%, cardiac consultation in 62%, stress test in 87%, and chest x-ray in 37% of patients. In phenotype ELv, 83% had a single peripheral intravenous line, and vascular injury was almost always followed by escalation in monitoring (e.g., central/arterial line), alert to the blood bank, and call for surgical backup. CONCLUSIONS: We have created a framework for IOD that can help with intraoperative safety and quality analysis. Focusing on interventions that address appropriateness versus futility in care in phenotypes T and NT, and on prevention and mitigation of intraoperative vessel injury (e.g., intraoperative rescue team) or preoperative optimization in phenotype EL may help prevent IODs.

Validation of the Emergency Surgery Score (ESS) in a Greek patient population: a prospective bi-institutional cohort study.

PURPOSE: The Emergency Surgery Score (ESS) is a reliable point-based score that predicts mortality and morbidity in emergency surgery patients. However, it has been validated only in the U.S. PATIENTS: We aimed to prospectively validate ESS in a Greek patient population. METHODS: All patients who underwent an emergent laparotomy were prospectively included over a 15-month period. A systematic chart review was performed to collect relevant preoperative, intraoperative, and postoperative variables based on which the ESS was calculated for each patient. The relationship between ESS and 30-day mortality, morbidity (i.e., the occurrence of at least one complication), and the need for intensive care unit (ICU) admission was evaluated and compared between the Greek and U.S. patients using the c-statistics methodology. The study was registered on "Research Registry" with the unique identifying number 5901. RESULTS: A total of 214 patients (102 Greek) were included. The mean age was 64 years, 44% were female, and the median ESS was 7. The most common indication for surgery was hollow viscus perforation (25%). The ESS reliably and incrementally predicted mortality (c-statistics = 0.79 [95% CI 0.67-0.90] and 0.83 [95% CI 0.74-0.92]), morbidity (c-statistics = 0.83 [95% CI 0.76-0.91] and 0.79 [95% CI 0.69-0.88]), and ICU admission (c-statistics = 0.88 [95% CI 0.81-0.96] and 0.84 [95% CI 0.77-0.91]) in both Greek and U.S. CONCLUSION: The correlation between the ESS and the surgical outcomes was statistically significant in both Greek and U.S. patients undergoing emergency laparotomy. ESS could prove globally useful for preoperative patient counseling and quality-of-care benchmarking.

The Role of Transaminases in Predicting Choledocholithiasis. A Novel Predictive Composite Score Development in a Cohort of 1089 Patients Undergoing Laparoscopic Cholecystectomy.

BACKGROUND: Optimal use of interventional procedures and diagnostic tests for patients with suspected choledocholithiasis depends on accurate pretest risk estimation. We sought to define sensitivity/specificity of transaminases in identifying choledocholithiasis and to incorporate them into a biochemical marker composite score that could accurately predict choledocholithiasis. METHODS: All adult patients who underwent laparoscopic cholecystectomy by our Emergency Surgery Service between 2010 and 2018 were reviewed. Admission total bilirubin (TB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (ALP) was captured. Choledocholithiasis was confirmed via intraoperative cholangiogram, endoscopic retrograde cholangiopancreatography, or magnetic resonance cholangiopancreatography. Area under receiver operating characteristic curve (AUC) or C-statistic for AST, ALT, ALP, and TB as a measure of detecting choledocholithiasis was calculated. For score development, our database was randomly dichotomized to derivation and validation cohort and a score was derived. The score was validated by calculating its C-statistic. RESULTS: 1089 patients were included; 210 (20.3%) had confirmed choledocholithiasis. The AUC was .78 for TB, .77 for ALP and AST, and .76 for ALT. 545 and 544 patients were included in the derivation and the validation cohort, respectively. The elements of the derived score were TB, AST, and ALP. The score ranged from 0 to 4. The AUC was .82 in the derivation and .77 in the validation cohort. The probability of choledocholithiasis increased from 8% to 89% at scores 0 to 4, respectively. CONCLUSIONS: Aspartate aminotransferase predicted choledocholithiasis adequately and should be featured in choledocholithiasis screening algorithms. We developed a biochemical composite score, shown to be accurate in preoperative choledocholithiasis risk assessment in an emergency surgery setting.

Time is domain: factors affecting primary fascial closure after trauma and non-trauma damage control laparotomy (data from the EAST SLEEP-TIME multicenter registry).

PURPOSE: Damage control laparotomy (DCL) is used for both traumatic and non-traumatic indications. Failure to achieve primary fascial closure (PFC) in a timely fashion has been associated with complications including sepsis, fistula, and mortality. We sought to identify factors associated with time to PFC in a multicenter retrospective cohort. METHODS: We reviewed retrospective data from 15 centers in the EAST SLEEP-TIME registry, including age, comorbidities (Charlson Comorbidity Index [CCI]), small and large bowel resection, bowel discontinuity, vascular procedures, retained packs, number of re-laparotomies, net fluid balance after 24 h, trauma, and time to first takeback in 12-h increments to identify key factors associated with time to PFC. RESULTS: In total, 368 patients (71.2% trauma, of which 50.6% were penetrating, median ISS 25 [16, 34], with median Apache II score 15 [11, 22] in non-trauma) were in the cohort. Of these, 92.9% of patients achieved PFC at 60.8 ± 72.0 h after 1.6 ± 1.2 re-laparotomies. Each additional re-laparotomy reduced the odds of PFC by 91.5% (95%CI 88.2-93.9%, p < 0.001). Time to first re-laparotomy was highly significant (p < 0.001) in terms of odds of achieving PFC, with no difference between 12 and 24 h to first re-laparotomy (ref), and decreases in odds of PFC of 78.4% (65.8-86.4%, p < 0.001) for first re-laparotomy after 24.1-36 h, 90.8% (84.7-94.4%, p < 0.001) for 36.1-48 h, and 98.1% (96.4-99.0%, p < 0.001) for > 48 h. Trauma patients had increased likelihood of PFC in two separate analyses (p = 0.022 and 0.002). CONCLUSION: Time to re-laparotomy ≤ 24 h and minimizing number of re-laparotomies are highly predictive of rapid achievement of PFC in patients after trauma- and non-trauma DCL. LEVEL OF EVIDENCE: 2B.

Dexmedetomidine and paralytic exposure after damage control laparotomy: risk factors for delirium? Results from the EAST SLEEP-TIME multicenter trial.

PURPOSE: To evaluate factors associated with ICU delirium in patients who underwent damage control laparotomy (DCL), with the hypothesis that benzodiazepines and paralytic infusions would be associated with increased delirium risk. We also sought to evaluate the differences in sedation practices between trauma (T) and non-trauma (NT) patients. METHODS: We reviewed retrospective data from 15 centers in the EAST SLEEP-TIME registry admitted from January 1, 2017 to December 31, 2018. We included all adults undergoing DCL, regardless of diagnosis, who had completed daily Richmond Agitation Sedation Score (RASS) and Confusion Assessment Method-ICU (CAM-ICU). We excluded patients younger than 18 years, pregnant women, prisoners and patients who died before the first re-laparotomy. Data collected included age, number of re-laparotomies after DCL, duration of paralytic infusion, duration and type of sedative and opioid infusions as well as daily CAM-ICU and RASS scores to analyze risk factors associated with the proportion of delirium-free/coma-free ICU days during the first 30 days (DF/CF-ICU-30) using multivariate linear regression. RESULTS: A 353 patient subset (73.2% trauma) from the overall 567-patient cohort had complete daily RASS and CAM-ICU data. NT patients were older (58.9 ± 16.0 years vs 40.5 ± 17.0 years [p < 0.001]). Mean DF/CF-ICU-30 days was 73.7 ± 96.4% for the NT and 51.3 ± 38.7% in the T patients (p = 0.030). More T patients were exposed to Midazolam, 41.3% vs 20.3% (p = 0.002). More T patients were exposed to Propofol, 91.0% vs 71.9% (p < 0.001) with longer infusion times in T compared to NT (71.2 ± 85.9 vs 48.9 ± 69.8 h [p = 0.017]). Paralytic infusions were also used more in T compared to NT, 34.8% vs 18.2% (p < 0.001). Using linear regression, dexmedetomidine infusion and paralytic infusions were associated with decreases in DF/CF-ICU-30, (- 2.78 (95%CI [- 5.54, - 0.024], p = 0.040) and (- 7.08 ([- 13.0, - 1.10], p = 0.020) respectively. CONCLUSIONS: Although the relationship between paralytic use and delirium is well-established, the observation that dexmedetomidine exposure is independently associated with increased delirium and coma is novel and bears further study.

Early versus late enteral nutrition in critically ill patients receiving vasopressor support.

BACKGROUND: Outcomes of early enteral nutrition (EEN) in critically ill patients on vasoactive medications remain unclear. We aimed to compare in-hospital outcomes for EEN vs late EN (LEN) in mechanically ventilated patients receiving vasopressor support. METHODS: This was a retrospective study using the national eICU Collaborative Research Database. Adult patients requiring vasopressor support and mechanical ventilation within 24 h of admission and for ≥2 days were included. Patients with an admission diagnosis that could constitute a contraindication for EEN (eg, gastrointestinal [GI] perforation, GI surgery) and patients with an intensive care unit (ICU) length of stay (LOS) <72 h were excluded. EEN and LEN were defined as tube feeding within 48 h and between 48 h and 1 week (nothing by mouth during the first 48 h) of admission, respectively. Propensity score matching was performed to derive two cohorts receiving EEN and LEN that were comparable for baseline patient characteristics. RESULTS: Among 1701 patients who met the inclusion criteria (EEN: 1001, LEN: 700), 1148 were included in propensity score-matched cohorts (EEN: 574, LEN: 574). Median time to EN was 29 vs 79 h from admission in the EEN and LEN groups, respectively. There was no significant difference in mortality or hospital LOS between the two nutrition strategies. EEN was associated with shorter ICU LOS, lower need for renal replacement therapy, and lower incidence of electrolyte abnormalities. CONCLUSION: This study showed no difference in 28-day mortality between EEN and LEN in critically ill patients receiving vasopressor support.

Machine Learning and Artificial Intelligence for Surgical Decision Making.

Background: The use of machine learning (ML) and artificial intelligence (AI) in medical research continues to grow as the amount and availability of clinical data expands. These techniques allow complex interpretation of data and capture non-linear relations not immediately apparent by classic statistical techniques. Methods: This review of the ML/AI literature provides a brief overview for practicing surgeons and clinicians of the current and future roles these methods will have within surgical infection research. Results: A conceptual overview of the techniques is provided along with concrete examples in the surgical infections literature. Further examples of ML/AI techniques in clinical decision support as well as therapy discovery with model-based deep reinforcement learning are illustrated. Conclusions: Artificial intelligence and ML are important and increasingly utilized techniques within the expanding body of surgical infection research. This article provides a minimal baseline literacy in ML/AI to be able to view such projects in an appropriately critical fashion.

Trauma and nontrauma damage-control laparotomy: The difference is delirium (data from the Eastern Association for the Surgery of Trauma SLEEP-TIME multicenter trial).

BACKGROUND: Damage-control laparotomy (DCL) has been used for traumatic and nontraumatic indications. We studied factors associated with delirium and outcome in this population. METHODS: We reviewed DCL patients at 15 centers for 2 years, including demographics, Charlson Comorbidity Index (CCI), diagnosis, operations, and outcomes. We compared 30-day mortality; renal failure requiring dialysis; number of takebacks; hospital, ventilator, and intensive care unit (ICU) days; and delirium-free and coma-free proportion of the first 30 ICU days (DF/CF-ICU-30) between trauma (T) and nontrauma (NT) patients. We performed linear regression for DF/CF-ICU-30, including age, sex, CCI, achievement of primary fascial closure (PFC), small and large bowel resection, bowel discontinuity, abdominal vascular procedures, and trauma as covariates. We performed one-way analysis of variance for DF/CF-ICU-30 against traumatic brain injury severity as measured by Abbreviated Injury Scale for the head. RESULTS: Among 554 DCL patients (25.8% NT), NT patients were older (58.9 ± 15.8 vs. 39.7 ± 17.0 years, p < 0.001), more female (45.5% vs. 22.1%, p < 0.001), and had higher CCI (4.7 ± 3.3 vs. 1.1 ± 2.2, p < 0.001). The number of takebacks (1.7 ± 2.6 vs. 1.5 ± 1.2), time to first takeback (32.0 hours), duration of bowel discontinuity (47.0 hours), and time to PFC were similar (63.2 hours, achieved in 73.5%). Nontrauma and T patients had similar ventilator, ICU, and hospital days and mortality (31.0% NT, 29.8% T). Nontrauma patients had higher rates of renal failure requiring dialysis (36.6% vs. 14.1%, p < 0.001) and postoperative abdominal sepsis (40.1% vs. 17.1%, p < 0.001). Trauma and NT patients had similar number of hours of sedative (89.9 vs. 65.5 hours, p = 0.064) and opioid infusions (106.9 vs. 96.7 hours, p = 0.514), but T had lower DF/CF-ICU-30 (51.1% vs. 73.7%, p = 0.029), indicating more delirium. Linear regression analysis indicated that T was associated with a 32.1% decrease (95% CI, 14.6%-49.5%; p < 0.001) in DF/CF-ICU-30, while achieving PFC was associated with a 25.1% increase (95% CI, 10.2%-40.1%; p = 0.001) in DF/CFICU-30. Increasing Abbreviated Injury Scale for the head was associated with decreased DF/CF-ICU-30 by analysis of variance (p < 0.001). CONCLUSION: Nontrauma patients had higher incidence of postoperative abdominal sepsis and need for dialysis, while T was independently associated with increased delirium, perhaps because of traumatic brain injury. LEVEL OF EVIDENCE: Therapeutic study, level IV.

The Emergency Surgery Score is a powerful predictor of outcomes across multiple surgical specialties: Results of a retrospective nationwide analysis.

BACKGROUND: The Emergency Surgery Score was recently validated in a prospective multicenter study as an accurate predictor of mortality in emergency general surgery patients. The Emergency Surgery Score is easily calculated using multiple demographic, comorbidity, laboratory, and acuity of disease variables. We aimed to investigate whether the Emergency Surgery Score can predict 30-day postoperative mortality across patients undergoing emergency surgery in multiple surgical specialties. METHODS: Our study is a retrospective cohort study using data from the national American College of Surgeons National Surgical Quality Improvement Program database (2007-2017). We included patients that underwent emergency gynecologic, urologic, thoracic, neurosurgical, orthopedic, vascular, cardiac, and general surgical procedures. The Emergency Surgery Score was calculated for each patient, and the correlation between the Emergency Surgery Score and 30-day mortality was assessed for each specialty using the c-statistics methodology. RESULTS: Of 6,485,915 patients, 173,890 patients were included. The mean age was 60 years, 50.6% were female patients, and the overall mortality was 9.7%. The Emergency Surgery Score predicted mortality best in emergency gynecologic, general, and urologic surgery (c-statistics: 0.97, 0.87, 0.81, respectively). The Emergency Surgery Score predicted mortality moderately well in emergency thoracic, neurosurgical, orthopedic, and vascular surgery (c-statistics 0.73-0.79). For example, the mortality of gynecology patients with an Emergency Surgery Score of 5, 9, and 13 was 2%, 27%, and 50%, respectively. The Emergency Surgery Score performed poorly in cardiac surgery. CONCLUSION: The Emergency Surgery Score accurately predicts mortality across patients undergoing emergency surgery in multiple surgical specialties, especially general, gynecologic, and urologic surgery. The Emergency Surgery Score can prove useful for perioperative patient counseling and for benchmarking the quality of surgical care.

Multisystem outcomes and predictors of mortality in critically ill patients with COVID-19: Demographics and disease acuity matter more than comorbidities or treatment modalities.

BACKGROUND: We sought to describe characteristics, multisystem outcomes, and predictors of mortality of the critically ill COVID-19 patients in the largest hospital in Massachusetts. METHODS: This is a prospective cohort study. All patients admitted to the intensive care unit (ICU) with reverse-transcriptase-polymerase chain reaction-confirmed severe acute respiratory syndrome coronavirus 2 infection between March 14, 2020, and April 28, 2020, were included; hospital and multisystem outcomes were evaluated. Data were collected from electronic records. Acute respiratory distress syndrome (ARDS) was defined as PaO2/FiO2 ratio of ≤300 during admission and bilateral radiographic pulmonary opacities. Multivariable logistic regression analyses adjusting for available confounders were performed to identify predictors of mortality. RESULTS: A total of 235 patients were included. The median (interquartile range [IQR]) Sequential Organ Failure Assessment score was 5 (3-8), and the median (IQR) PaO2/FiO2 was 208 (146-300) with 86.4% of patients meeting criteria for ARDS. The median (IQR) follow-up was 92 (86-99) days, and the median ICU length of stay was 16 (8-25) days; 62.1% of patients were proned, 49.8% required neuromuscular blockade, and 3.4% required extracorporeal membrane oxygenation. The most common complications were shock (88.9%), acute kidney injury (AKI) (69.8%), secondary bacterial pneumonia (70.6%), and pressure ulcers (51.1%). As of July 8, 2020, 175 patients (74.5%) were discharged alive (61.7% to skilled nursing or rehabilitation facility), 58 (24.7%) died in the hospital, and only 2 patients were still hospitalized, but out of the ICU. Age (odds ratio [OR], 1.08; 95% confidence interval [CI], 1.04-1.12), higher median Sequential Organ Failure Assessment score at ICU admission (OR, 1.24; 95% CI, 1.06-1.43), elevated creatine kinase of ≥1,000 U/L at hospital admission (OR, 6.64; 95% CI, 1.51-29.17), and severe ARDS (OR, 5.24; 95% CI, 1.18-23.29) independently predicted hospital mortality.Comorbidities, steroids, and hydroxychloroquine treatment did not predict mortality. CONCLUSION: We present here the outcomes of critically ill patients with COVID-19. Age, acuity of disease, and severe ARDS predicted mortality rather than comorbidities. LEVEL OF EVIDENCE: Prognostic, level III.