Search

Post-operative delirium (POD) is common yet often underdiagnosed following vascular surgery. Elderly patients with advanced peripheral artery disease may be at particular risk for POD yet understanding of the clinical predictors and impact of POD is incomplete. We sought to identify POD predictors and associated resource utilization after infrainguinal lower extremity bypass.This single center retrospective analysis included all infrainguinal bypass cases performed for peripheral arterial disease from 2012-2020. The primary outcome was inpatient POD. Delirium sequelae were also evaluated. Key secondary outcomes were length of stay, nonhome discharge, readmission, 30-day amputation, post-operative myocardial infarction, mortality, and 2-year survival. Regression analysis was used to evaluate risk factors for delirium in addition to association with 2-year survival and amputation free survival.Among 420 subjects undergoing infrainguinal lower extremity bypass, 105 (25%) developed POD. Individuals with POD were older and more likely to have non-elective surgery (P 0.05). On multivariable analysis, independent predictors of POD were age 60-89 years old, chronic limb threatening ischemia, female sex, and nonelective procedure. Consultations for POD took place for 25 cases (24%); 13 (52%) were with pharmacists, and only 4 (16%) resulted in recommendations. The average length of stay for those with POD was higher (17 days vs. 9 days; P 0.001). POD was associated with increased non-home discharge (61.8% vs. 22.1%; P 0.001), 30-day major amputation (6.7% vs. 1.6%; P 0.01), 30-day postoperative myocardial infarction (11.4% vs. 4.1%; P 0.01), and 90-day mortality (7.6% vs. 2.9%; P = 0.03). Survival at 2 years was lower in those with delirium (89% vs. 75%; P 0.001). In a Cox proportional hazards model, delirium was independently associated with decreased survival (HR = 2.0; 95% CI = 1.15-3.38; P = 0.014) and decreased major-amputation free survival (HR = 1.9; 95% CI = 1.18-2.96; P = 0.007).POD is common following infrainguinal lower extremity bypass and is associated with other adverse post-operative outcomes and increased resource utilization, including increased hospital length of stay, nonhome discharge, and worse 2-year survival. Future studies should evaluate the role of routine multidisciplinary care for high-risk patients to improve perioperative outcomes for vulnerable older adults undergoing infrainguinal lower extremity bypass.

Racial and ethnic disparities have been well-documented in the outcomes for chronic limb threatening ischemia (CLTI). One purported explanation has been the disease severity at presentation. We hypothesized that the disparities in major adverse limb events (MALE) after peripheral vascular intervention (PVI) for CLTI would persist despite controlling for disease severity at presentation using the WIfI (wound, ischemia, foot infection) stage.The Vascular Quality Initiative PVI dataset (2016-2021) was queried for CLTI. Patients were excluded if they were missing the WIfI stage. The primary end point was the incidence of 1-year MALE, defined as major amputation (through the tibia or fibula or more proximally) or reintervention (endovascular or surgical) of the initial treatment limb. A multivariate hierarchical Fine-Gray analysis was performed, controlling for hospital variation, competing risk of death, and presenting WIfI stage, to assess the independent association of Black/African American race and Latinx/Hispanic ethnicity with MALE. A Cox proportional hazard regression model was used for the 1-year survival analysis.Overall, 47,830 patients (60%) had had WIfI scores reported (73% White, 20% Black, and 7% Latinx). The 1-year unadjusted cumulative incidence of MALE was 13.1% (95% confidence interval [CI], 12.6%-13.5%) for White, 14.3% (95% CI, 13.5%-15.3%) for Black, and 17.0% (95% CI, 15.3%-18.9%) for Latinx patients. On bivariate analysis, the occurrence of MALE was significantly associated with younger age, Black race, Latinx ethnicity, coronary artery disease, cerebrovascular disease, congestive heart failure, hypertension, diabetes, dialysis, intervention level, any prior minor or major amputation, and WIfI stage (P .001). The cumulative incidence of 1-year MALE increased by increasing WIfI stage: stage 1, 11.7% (95% CI, 10.9%-12.4%); stage 2, 12.4% (95% CI, 11.8%-13.0%); stage 3, 14.8% (95% CI, 13.8%-15.8%); and stage 4, 15.4% (95% CI, 14.3%-16.6%). The cumulative incidence also increased by intervention level: inflow, 10.7% (95% CI, 9.8%-11.7%), femoropopliteal, 12.3% (95% CI, 11.7%-12.9%); and infrapopliteal, 14.1% (95% CI, 13.5%-14.8%). After adjustment for WIfI stage only, Black race (subdistribution hazard ratio [SHR], 1.30; 95% CI, 1.17-1.44; P .001) and Latinx ethnicity (SHR, 1.58; 95% CI, 1.37-1.81; P .001) were associated with an increased 1-year hazard of MALE compared with White race. On adjusted multivariable analysis, MALE disparities persisted for Black/African American race (SHR, 1.12; 95% CI, 1.01-1.25; P = .028) and Latinx/Hispanic ethnicity (SHR, 1.34; 95% CI, 1.16-1.54; P .001) compared with White race.Black/African American and Latinx/Hispanic patients had a higher associated hazard of MALE after PVI for CLTI compared with White patients despite an adjustment for WIfI stage at presentation. These results suggest that disease severity at presentation does not account for disparities in outcomes. Further work should focus on better understanding the underlying mechanisms for disparities in historically marginalized racial and ethnic groups presenting with CLTI.

Despite the shared pathogenesis of peripheral arterial disease (PAD) and vascular dementia, there is little data on cognitive impairment in PAD patients. We hypothesized that cognitive impairment will be common and previously unrecognized.Cognitive impairment screening was prospectively performed for Veterans presenting to a single Veterans Affairs outpatient vascular surgery clinic from 2020-2021 for PAD consultation or disease surveillance. Overall, 125 Veterans were screened. Cognitive impairment was defined as a score of26 on the Montreal Cognitive Assessment (MoCA) survey. A multivariable logistic regression assessed for independent risk factors for cognitive impairment.Overall, 77 (61%) had cognitive impairment, 92% was previously unrecognized. Cognitive impairment was associated with increased age (74.4 vs. 71.8 years, p=0.03), Black vs. White race (94% vs. 54%, p0.01), hypertension (66% vs. 31%, p=0.01), prior stroke/TIA (79% vs. 58%, p=0.03), diabetes treated with insulin (79% vs. 58%, p=0.05), and post-traumatic stress disorder (PTSD) (80% vs. 57%, p=0.04). On multivariable analysis, risk factors for newly diagnosed cognitive impairment included age ≥ 70 years, diabetes treated with insulin, PTSD, and Black race.Many Veterans with PAD have evidence of cognitive impairment and is overwhelmingly underdiagnosed. This study suggests cognitive impairment is an unrecognized issue in a VA population with PAD, requiring more study to determine cognitive impairment's impact on surgical outcomes, and how it can be mitigated and incorporated into clinical care.

With the evolution in vascular surgery toward increased endovascular therapy and decreased open surgical training, comfort with open procedures by current trainees is declining. A proposed method to improve this discomfort is simulator training. We hypothesized that open, cadaver, and endovascular surgery simulation would be associated with increased self-perceived comfort in performing corresponding procedures.Integrated (0 + 5) vascular surgery residents and recent graduates in the United States were asked to complete a survey quantifying comfort via a Likert scale with procedures and experience with simulation training. Simulation groups were then matched using coarsened exact matching. Ordinal logistic regression assessed the association between simulation experience and comfort in performing procedures.Surveys were completed by 68 trainees and 20 attending surgeons in their first 5 years of practice. On unmatched analyses, there were no significant differences in comfort in performing any open or endovascular aorto-mesenteric or peripheral vascular procedures between respondents who reported experience with open or endovascular simulation, respectively. However, respondents who reported cadaver simulation experience (58%, 51/88) had a significantly higher reported comfort score performing open juxtarenal aortic repair (2.4 vs. 1.7), superior mesenteric artery thrombectomy or bypass (2.5 vs. 1.9), inferior vena cava or iliac vein repair (2.2 vs. 1.7), axillary-femoral artery bypass (3.4 vs. 2.5), femoral-popliteal artery bypass (3.7 vs. 2.8), and inframalleolar artery bypass (2.8 vs. 2.1; all P 0.05). After matching on training level, number of abdominal cases completed, and number of open vascular cases completed, ordinal logistic regression demonstrated that previous cadaver simulation was significantly associated with increased comfort in performing open aortic repairs, venous repair, visceral revascularization, and peripheral bypasses.In this nationally representative sample, cadaver, but not open or endovascular, simulation was associated with increased comfort in performing open vascular surgery. Providing cadaver simulation to trainees may help to improve comfort levels in performing open surgery. Integrated vascular surgery training programs should consider implementing these experiences into their curriculum.

The purpose was to examine the incidence of positive cultures in a highly susceptible subset of trauma patients admitted to the surgical intensive care unit (SICU) for mechanical ventilation and to examine the impact of their timing on outcomes.A retrospective review was conducted of blunt trauma patients admitted to the SICU for mechanical ventilation at a level I trauma center over a five-year period. All urine, blood, and sputum cultures were abstracted. Patients with at least one positive culture were compared with those with negative or no cultures. The primary outcome was mortality. A Cox regression model with a time-dependent variable was utilized to calculate the adjusted hazard ratio (AHR).The median age of 635 patients meeting inclusion criteria was 46 and 74.2% were male. A total of 298 patients (46.9%) had at least one positive culture, with 28.9% occurring within two days of admission. Patients with positive cultures were more likely to be severely injured with an injury severity score (ISS) ≥16 (68.5% vs. 45.1%, p 0.001). Overall mortality was 22%. Patients who had their first positive culture within two and three days from admission had a significantly higher AHR for mortality (AHR: 14.46, p 0.001 and AHR: 10.59, p = 0.028, respectively) compared to patients with a positive culture at day six or later.Early positive cultures are common among trauma patients requiring mechanical ventilation and are associated with higher mortality. Early identification with "damage control cultures" obtained on admission to aid with early targeted treatment might be justified.

Background Massive transfusion protocols (MTPs) are necessary for hemodynamically unstable trauma patients with active bleeding. Thrombotic events have been associated with blood transfusion; however, the risk factors for the development of venous thromboembolism (VTE) in trauma patients receiving MTP are unknown. Methods A retrospective review was conducted by reviewing the electronic medical records of all trauma patients admitted to a Level I trauma center who received MTP from 2011 to 2016. Data were collected on patient demographics, mechanism of injury, injury severity scores, quantity of blood products transfused during MTP activation, incidence of VTE, intensive care unit length of stay (LOS), hospital LOS, and ventilator days. The primary outcome was VTE. Results Of the 59 patients who had MTP activated, 15 (25.4%) developed a VTE during their hospital admission. Patients who developed VTE were compared with those who did not. Age (40 y versus 35 y, P = 0.59), sex (60% versus 73% male, P = 0.52), and mechanism of injury (47% versus 59% blunt, P = 0.40) were similar. Intensive care unit LOS, hospital LOS, and ventilator days were longer in the patients who were diagnosed with a VTE. Multivariable analysis revealed an increase in the odds for developing a VTE with increasing packed red blood cell transfusion (adjusted odds ratio = 2.61, P = 0.03). Conclusions The risk for VTE in trauma patients requiring massive transfusion is proportional to the number of packed red blood cells transfused. Liberal screening protocols and maintenance of a high index of suspicion for VTE in these high-risk patients is justified.

The use of vasopressors (VP) in the resuscitation of massively transfused trauma patients might be considered a marker of inadequate resuscitation. We sought to characterize the utilization of VP in patients receiving massive transfusion and examine the association of their use with mortality.Trauma patients admitted from January 2011 to October 2016 receiving massive transfusion, defined as 3 units of pRBC within the first hour from admission, were selected for analysis. Demographics, admission vital signs and labs, use of VP, surgical interventions and outcomes were collected. Standard statistical tools were utilized.Over the 5-year study period, 120 trauma patients met inclusion criteria. The median age was 39 years with 77% being male and 41% sustaining a penetrating injury. Patients who received VP [VP (+)] were more likely to have a lower admission GCS (median 4.5 vs. 14.0, p0.01) and less likely to have a penetrating injury (31% vs. 54%, p=0.02). The overall mortality was 49% and significantly higher in the VP (+) cohort (60% vs. 34%, AHR: 9.9, adjusted p=0.03). Mortality increased in a stepwise fashion with increasing number of VP utilized, starting at 34% for no VP, to 78% for 3 VP, and 100% for 5 or more. The majority of deaths in the VP (-) group (88%) occurred within one day from admission. For the VP (+) group, 57% of deaths occurred within one day, with the remaining 43% occurring at a later time.In the era of massive transfusion protocols, vasopressors are commonly utilized in exsanguinating trauma patients and their use is associated with a higher mortality risk. Deaths in patients receiving vasopressors are more likely to occur later compared to those in patients who do not receive vasopressors. Further research to characterize the role of these agents in the resuscitation of trauma patients is required.

Introduction Rigid cervical collars are routinely placed in the pre-hospital setting after significant blunt trauma. Patients who are deemed competent by field personnel (Glasgow Coma Scale (GCS) ≥13, no major distracting injury and not grossly intoxicated) may refuse cervical collar placement. Material and methods A retrospective review was conducted of all adult trauma patients presenting to a Level 1 trauma center after blunt trauma with a GCS≥13 and no distracting injury or gross intoxication from January 2014 to December 2014. Pre-hospital data was collected from emergency medical service reports and hospital data from patient charts. Cervical spine injury was identified by International Classification of Disease-9th Revision codes. Patients refusing cervical spine immobilization prior to arrival are compared to those who were compliant. Results A total of 629 patients met inclusion criteria. Cervical spine immobilization was refused by 28 patients, while 601 complied. There were 16 cervical spine injuries (2.5%), with 3 (10.7%) in noncompliant patients and 13 (2.2%) among those who were complaint (p = 0.03). Conclusion The incidence of cervical spine injuries in patients refusing cervical collar immobilization is higher than in compliant patients. Patients arriving for initial evaluation having refused cervical collar immobilization should be treated with caution.

Patients with traumatic brain injury (TBI) are often resuscitated with crystalloids in the emergency department (ED) to maintain cerebral perfusion. The purpose of this study was to evaluate whether crystalloid resuscitation volume impacts mortality in TBI patients. This was a retrospective study of trauma patients with head abbreviated injury scale score ≥2, who received crystalloids during ED resuscitation between 2004 and 2013. Clinical characteristics and volume of crystalloids received in the ED were collected. Patients who received

Background High-flow nasal cannula (HFNC) is increasingly used to reduce reintubations in patients with respiratory failure. Benefits include providing positive end expiratory pressure, reducing anatomical dead space, and decreasing work of breathing. We sought to compare outcomes of critically ill surgical patients extubated to HFNC versus conventional therapy. Methods A retrospective review was conducted in the surgical intensive care unit of an academic center during August 2015 to February 2016. Data including demographics, ventilator days, oxygen therapy after extubation, reintubation rates, surgical intensive care unit and hospital length of stay, and mortality were collected. Self and palliative extubations were excluded. Characteristics and outcomes, with the primary outcome being reintubation, were compared between those extubated to HFNC versus cool mist/nasal cannula (CM/NC). Results Of the 184 patients analyzed, 46 were extubated to HFNC and 138 to CM/NC. Mean age and days on ventilation before extubation were 57.8 years and 4.3 days, respectively. Both cohorts were similar in age, sex, and had a similar prevalence of cardiopulmonary diagnoses at admission. Although prior to extubation HFNC had lengthier ventilation requirements (7.1 versus 3.4 days, P Conclusions Ventilated patients at risk for recurrent respiratory failure have reduced reintubation rates when extubated to HFNC. Patients with prolonged intubation or those with high-risk comorbidities may benefit from extubation to HFNC.

Purpose The purpose of the study is to determine if excessive fluid administration is associated with a prolonged hospital course and worse outcomes. Materials and methods In July 2013, all normotensive trauma patients admitted to the surgical intensive care unit (ICU) were administered crystalloids at 30 mL/h (“to keep open [TKO]”) and were compared to patients admitted during the preceding 6 months who were placed on a rate between 125 mL/h to 150 mL/h (non-TKO). The primary outcomes were ICU, hospital, and ventilator days. Results A total of 101 trauma patients met inclusion criteria: 56 (55.4%) in the TKO and 45 (44.6%) in the non-TKO group. Overall, the 2 groups were similar in regard to age, Injury Severity Score, Acute Physiology and Chronic Health Evaluation IV scores, and the need for mechanical ventilation. TKO had no effect on renal function compared to non-TKO with similarities in maximum hospital creatinine. TKO patients had lower ICU stay (2.7 ± 1.5 vs 4.1 ± 4.6 days; P = .03) and ventilator days (1.4 ± 0.5 vs 5.5 ± 4.8 days; P Conclusions A protocol that encourages admission basal fluid rate of TKO or 30 mL/h in normotensive trauma patients is safe, reduces fluid intake, and may be associated with a shorter intensive care unit course and fewer ventilator days.

Background Adequate venous thromboembolism (VTE) prophylaxis is essential after trauma, especially in patients with lower extremity and/or pelvic fractures. We sought to investigate if prophylactic enoxaparin dosed by anti -Xa trough levels could reduce clinically evident VTE in trauma patients with lower extremity or pelvic injury. Methods Prospective data was collected on trauma patients admitted for at least two days with any lower extremity and/or pelvic fracture and who received enoxaparin for VTE prophylaxis between October 2013 and January 2016. Patients in the control cohort received enoxaparin at 30 mg twice daily. Patients in the adjustment cohort had anti -Xa trough levels measured after three or more consecutive doses of enoxaparin. Those with a trough level of 0.1 IU/mL or lower had their dosage increased by 10-mg increments. Results Of the 159 patients included, 58 (36.5%) were monitored with anti -Xa trough levels. The cohorts were similar in age, sex, regional AIS, ISS score, ICU and hospital length of stay, proportion of patients with diagnostic testing for VTE, and time to first enoxaparin dose. Initial enoxaparin dosing in the majority of patients (84.5%) who had anti -Xa trough levels measured was subprophylactic. Patients receiving enoxaparin dosed by anti -Xa trough level had a significantly lower VTE rate than those who did not (1.7% v. 13.9%, p = 0.03). Conclusions Prophylactic enoxaparin adjusted by anti -factor Xa level may lead to a decreased rate of clinically evident VTE among trauma patients with lower extremity and/or pelvic fractures. Our findings indicate that the initial dose of enoxaparin was frequently too low.

BACKGROUND Traumatic brain injury (TBI) is a leading cause of morbidity and mortality in the United States, especially in the elderly, who have the highest rates of TBI-related hospitalizations and deaths among all age groups. Sepsis is one of many risk factors that is associated with higher mortality and longer length of hospital stay in this population partially due to the immunosuppressive effects of TBI. The significance of early indicators of infection, such as a positive blood, sputum, or urine culture, is not well described. The purpose of this study was to determine if early positive cultures predict higher mortality in elderly patients with TBI. METHODS All trauma patients aged ≥65 years with TBI, admitted between January 1, 2009 and December 31, 2013 to the surgical intensive care unit, were retrospectively reviewed. Clinical data including results from sputum, blood, and urine cultures were reviewed. RESULTS Overall, 288 elderly patients with TBI were identified, and 92 (32%) had a positive culture. Patients with positive cultures had longer intensive care unit (median 6.0 versus 2.0 days, P

While assault is commonly reported or suspected in children with traumatic wounds, a recent overview of these injuries, especially those requiring trauma surgery consultation is lacking in the literature.Explore the incidence, demographics and injury patterns of children presenting to trauma centers following an assault.Retrospective review of the National Trauma Data Bank 2007 to 2011.Subjects up to 18 years old with "assault" reported as the intent of injury. Patients were divided into infants (2 years), young children (2-5 years), children (6-11 years), and adolescents (12-18 years).Mechanism of injury, injury severity and mortality based on age groups and race.Of 609,207 children, 58,299 (9.6%) were victims of an assault. The median age was 16 years and 81% were male, with a median injury severity score (ISS) of 8. The majority of patients were adolescents (76%), followed by infants (17%) and young children (4%). There was a stepwise increase in the proportion of assaulted Black children with increasing age (23.2% for infants and up to 46.7% for adolescents, trend p 0.01, effect size: 0.175) while the opposite applied for White children (46.0% for infants and down to 19.5% for adolescents, trend p 0.01, effect size: -0.230). With increasing age, White subjects had the highest trend of being assaulted during an unarmed fight or brawl (p 0.01, effect size: 0.393), while for Black victims the highest trend was noted for assault with a firearm (p 0.01, effect size: 0.323). Almost 2 out of 3 infants sustained severe head trauma (59%). The overall mortality was 8%, highest among young children, where it reached 16% (p 0.01).Up to 10% of children admitted following trauma are victims of assault with traumatic brain injuries predominant in infants and firearm injuries predominant in adolescents. Injury patterns largely correlate to age and race. Assault in children is associated with a high mortality risk. These data highlight the magnitude of the problem and calls for further involvement of trauma surgeons to improve outcomes, bring awareness and promote preventative strategies to eliminate assault in children.

Liberal emergency department (ED) resuscitation after trauma may lead to uncontrolled hemorrhage, reduced organ perfusion, and compartment syndrome. Recent guidelines reduced the standard starting point for crystalloid resuscitation from 2 L to 1 L and emphasized "balanced" resuscitation. The purpose of this study was to characterize how an urban, Level 1 trauma center has responded to changes in crystalloid resuscitation practices over time and to describe associated patient outcomes.This is a retrospective review of trauma patients who sustained moderate to severe injury (ISS 9) and received crystalloid resuscitation in the ED during 1/2004-12/2013 at an urban, Level 1 trauma center. Patient data collected included age, gender, Glasgow Coma Scale (GCS) score, initial systolic blood pressure (SBP), mechanism of injury, regional Abbreviated Injury Scale (AIS) score, Injury Severity Score (ISS), volume of blood products and crystalloids administered in the ED. Patients who received2 L of crystalloid were considered low-volume while those who received ≥2 L were high-volume patients. Clinical characteristics and outcomes were compared between high- and low-volume cohorts, and multivariate regression was used to adjust for confounders. Trend analysis examined changes in variables over time.1571 moderate to severely injured patients received crystalloid resuscitation; 1282 (82%) were low-volume and 289 (18%) were high-volume. Compared to high-volume patients, low-volume patients presented with a higher median SBP (134 vs. 122 mmHg, p 0.001) and GCS (15 vs. 14, p 0.001). Low-volume patients also had lower median ISS (15 vs. 19, p 0.001). Unadjusted mortality was lower in the low-volume cohort (7% vs. 19%, p 0.001). Multivariate analysis demonstrated that high-volume patients had increased odds of mortality compared to low-volume patients (AOR 1.88, p = 0.008). Decreased rates of high-volume resuscitation and overall mortality were demonstrated over the 10-year study period.The observed decrease in high-volume crystalloid resuscitations in the ED paralleled a reduction in mortality over the ten-year period. In addition, adjusted mortality was higher in those receiving high-volume resuscitation.

Background Spinal motion restriction (SMR) after traumatic injury has been a mainstay of prehospital trauma care for more than 3 decades. Recent guidelines recommend a selective approach with cervical spine clearance in the field when criteria are met. Materials and methods In January 2014, the Department of Health Services of the City of Los Angeles, California, implemented revised guidelines for cervical SMR after blunt mechanism trauma. Adult patients (aged ≥18 y) with an initial Glasgow Coma Scale (GCS) score of ≥13 presented to a single level I trauma center after blunt mechanism trauma over the following 1-y period were retrospectively reviewed. Demographics, injury data, and prehospital data were collected. Cervical spine injury (CSI) was identified by International Classification of Disease, Ninth Revision, codes. Results Emergency medical services transported 1111 patients to the emergency department who sustained blunt trauma. Patients were excluded if they refused c-collar placement or if documentation was incomplete. A total of 997 patients were included in our analysis with 172 (17.2%) who were selective cleared of SMR per protocol. The rate of Spinal Cord Injury was 2.2% (22/997) overall and 1.2% (2/172) in patients without SMR. The sensitivity and specificity of the protocol are 90.9% (95% confidence interval: 69.4-98.4) and 17.4% (95% confidence interval: 15.1-20.0), respectively, for CSI. Patients with CSI who arrived without immobilization having met field clearance guidelines, were managed without intervention, and had no neurologic compromise. Conclusions Guidelines for cervical SMR have high sensitivity and low specificity to identify CSI. When patients with injuries were not placed on motion restrictions, there were no negative clinical outcomes.

Importance Continuous renal replacement therapy (CRRT) benefits patients with renal failure who are too hemodynamically unstable for intermittent hemodialysis. The duration of therapy beyond which continued use is futile, particularly in a population of patients admitted to and primarily cared for by a surgical service (hereinafter referred to as surgical patients), is unclear. Objective To analyze proportions of and independent risk factors for survival to discharge after initiation of CRRT among patients in a surgical intensive care unit (SICU). Design, Setting, and Participants This retrospective cohort study included all patients undergoing CRRT from July 1, 2012, through January 31, 2016, in an SICU of an urban tertiary medical center. The population included patients treated before or after general surgery and patients admitted to a surgical service during inpatient evaluation and care before liver transplant. The pretransplant population was censored from further survival analysis on receipt of a transplant. Exposures Continuous renal replacement therapy. Main Outcomes and Measures Hospital mortality among patients in an SICU after initiation of CRRT. Results Of 108 patients (64 men [59.3%] and 44 women [40.7%]; mean [SD] age, 62.0 [12.7] years) admitted to the SICU, 53 were in the general surgical group and 55 in the pretransplant group. Thirteen of the 22 patients in the pretransplant group who required 7 or more days of CRRT died (in-hospital mortality, 59.1%); among the 12 patients in the general surgery group who required 7 or more days of CRRT, 12 died (in-hospital mortality, 100%). In the general surgical group, each day of CRRT was associated with an increased adjusted odds ratio of death of 1.39 (95% CI, 1.01-1.90; P = .04). Conclusions and Relevance Continuous renal replacement therapy is valuable for surgical patients with an acute and correctable indication; however, survival decreases significantly with increasing duration of CRRT. Duration of CRRT does not correlate with survival among patients awaiting liver transplant.