The association between initial calculated driving pressure at the induction of general anesthesia and composite postoperative oxygen support.

PURPOSE: Early discontinuation of postoperative oxygen support (POS) would partially depend on the innate pulmonary physics. We aimed to examine if the initial driving pressure (dP) at the induction of general anesthesia (GA) predicted POS prolongation. METHODS: We conducted a single-center retrospective study using the facility's database. Consecutive subjects over 2 years were studied to determine the change in odds ratio (OR) for POS prolongation of different dP classes at GA induction. The dP (cmH2O) was calculated as the ratio of tidal volume (mL) over dynamic Crs (mL/cmH2O) regardless of the respiratory mode. The adjusted OR was calculated using the logistic regression model of multivariate analysis. Moreover, we performed a secondary subgroup analysis of age and the duration of GA. RESULTS: We included 5,607 miscellaneous subjects. Old age, high scores of American Society of Anesthesiologist physical status, initial dP, and long GA duration were associated with prolonged POS. The dP at the induction of GA (7.78 [6.48, 9.45] in median [interquartile range]) was categorized into five classes. With the dP group of 6.5-8.3 cmH2O as the reference, high dPs of 10.3-13 cmH2O and ≥ 13 cmH2O were associated with significant prolongation of POS (adjusted OR, 1.62 [1.19, 2.20], p = 0.002 and 1.92 [1.20, 3.05], p = 0.006, respectively). The subgroup analysis revealed that the OR for prolonged POS of high dPs disappeared in the aged and ≥ 6 h anesthesia time subgroup. CONCLUSIONS: High initial dPs ≥ 10 cmH2O at GA induction predicted longer POS than those of approximately 7 cmH2O. High initial dPs were, however, a secondary factor for prolongation of postoperative hypoxemia in old age and prolonged surgery.

Pediatric Fulminant Myocarditis in Japan: A Retrospective Nationwide Database Study of Hospital Volume, Management Practices, and Mortality.

OBJECTIVES: Pediatric fulminant myocarditis is a subset of pediatric acute myocarditis associated with critical illness. We aimed to compare mortality and other outcomes such as length of hospital stay between pediatric fulminant myocarditis and nonfulminant myocarditis. For the subgroup of patients with fulminant myocarditis, we also aimed to describe the current management practices and evaluate the impact of clinically relevant factors, including hospital case volume, on mortality. DESIGN: Retrospective observational study using the Diagnosis Procedure Combination database from April 2012 to March 2018. SETTING: Over 1,000 acute care hospitals in Japan. PATIENTS: Patients with acute myocarditis less than 18 years old, including patients with fulminant myocarditis (i.e., those who received at least one of the following by day 7 of hospitalization: inotropes/vasopressors, mechanical circulatory support, or cardiopulmonary resuscitation). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Multivariable logistic regression analysis was conducted to investigate the association between clinically relevant factors and in-hospital mortality of patients with fulminant myocarditis. Furthermore, post hoc propensity score analyses (propensity score-adjusted, propensity score-matched, and inverse probability of treatment-weighted analyses) were performed to confirm the effect of hospital case volume on in-hospital mortality. In total, 866 pediatric patients with acute myocarditis were included, and 382 (44.1%) were categorized as fulminant myocarditis. In-hospital mortality for those with fulminant myocarditis was 24.1%. fulminant myocarditis was associated with 41.3-fold greater odds of mortality than nonfulminant myocarditis (95% CI, 14.7-115.9; p < 0.001). In the subgroup of patients with fulminant myocarditis, a higher in-hospital mortality was significantly associated with younger age (≤ 5 yr; odds ratio, 3.41; 95% CI, 1.75-6.64) and the need for either mechanical ventilation (odds ratio, 2.39; 95% CI, 1.03-5.57), cardiopulmonary resuscitation (odds ratio, 10.63; 95% CI, 5.52-20.49), or renal replacement therapy (odds ratio, 2.53; 95% CI, 1.09-5.87) by day 7. A lower in-hospital mortality rate was significantly associated with treatment at hospitals in the highest pediatric fulminant myocarditis case volume tertile (≥ 6 cases in 6 yr; odds ratio, 0.30; 95% CI, 0.13-0.68) compared with treatment at hospitals in the lowest tertile (1-2 cases in 6 yr). Post hoc propensity score analyses consistently supported the primary results. CONCLUSIONS: In-hospital mortality of pediatric fulminant myocarditis in Japan remains high. Treatment at hospitals in the highest pediatric fulminant myocarditis case volume tertile (≥ 6 cases in 6 yr) was associated with a 70% relative reduction in odds of in-hospital mortality compared with treatment at hospitals in the lowest tertile (1-2 cases in 6 yr). The reasons for such differences need further study.

Hypothermia, Poorly Recognized by clinicians, is associated with Higher Mortality Among Critically Ill Patients with infections: A Retrospective Observational Study.

We investigated the relationship between the presence of hypothermia in infection and mortality in 233 infectious critically ill patients. The adjusted hazard ratio for death at 28 days in the low body temperature group was 3.30 compared with the high body temperature group. The proportion of appropriate antimicrobial therapy significantly decreased with decreasing body temperature. The proportion of medical records that documented body temperature abnormality in the low body temperature group (33%) was significantly lower than that in the high body temperature group (69%). Delayed antimicrobial therapy in patients with hypothermia, which may be due to poor recognition by physicians, could result in mortality.

Effects of acute ethanol intoxication in an ovine peritonitis model.

BACKGROUND: Acute ethanol intoxication has been shown to have contrasting effects on outcomes in sepsis. The aim of this study was to explore the effects of acute ethanol intoxication on hemodynamics, renal function, brain perfusion and lactate/pyruvate in an ovine sepsis model. METHODS: Anesthetized, mechanically ventilated female sheep were randomized to an ethanol group (n = 7), which received 1 g/kg ethanol diluted in intravenous (i.v.) saline infusion or a control group (n = 7), which received the same volume of i.v. saline. Both groups received the treatment for a period of 2 h prior to induction of sepsis by intraperitoneal injection of feces. Other treatment included fluid resuscitation but no vasopressors or antibiotics. Global hemodynamics, renal blood flow, brain cortex laser Doppler flowmetry and microdialysis analyses were recorded hourly. RESULTS: In the ethanol group, blood ethanol concentrations were 137 ± 29 mg/dL at the time of feces injection and decreased to become undetectable by 12 h. Arterial hypotension occurred earlier in the ethanol than in the control group (8 [7-12] vs. 14 [11-20] hours, p = 0.03). Lactate levels increased to > 2 mmol/L earlier in the ethanol group. Renal dysfunction (9 [6-13] vs. 13 [12-15] hours, p = 0.05) and oliguria (urine output < 0.5 mL/kg/h; 10 [7-12] vs. 13 [12, 13] hours, p = 0.01) developed earlier in the ethanol than in the control group. Brain blood flow and lactate/pyruvate were unaffected. There was no significant difference in survival time. CONCLUSIONS: Acute ethanol intoxication in this model of peritonitis resulted in earlier development of shock and renal dysfunction but did not alter brain perfusion and metabolism or short-term survival.

Rapid-onset plasma leakage of extracorporeal oxygenation membranes possibly due to hyperbilirubinemia.

Extracorporeal membrane oxygenation (ECMO) is an emerging tool for supporting cardiopulmonary function in patients with cardiorespiratory failure or arrest. The oxygenator of the ECMO circuit requires effective oxygenation and removal of carbon dioxide from the blood. Major problems that can occur with the oxygenator include plasma leakage, one of the late-onset serious complications necessitating device replacement. However, the rapid onset of plasma leakage is rare. We present a 1-year-old boy with acute respiratory failure due to Pneumocystis and Aspergillus pneumonia. He presented with tachypnea, tachycardia, and hypoxemia despite the ventilatory support, and was therefore placed on venoarterial ECMO with a drainage catheter from the right internal jugular vein (12 Fr) and a return catheter to the right internal carotid artery (10 Fr). Extracorporeal circulation was initiated at a blood flow of 1 L/min (145 mL/kg/min) and a sweep gas flow of 1 L/min with FiO2 of 0.7. Although he was successfully weaned from the venoarterial ECMO on day 15 with an improvement of cardiopulmonary function, he was later placed on venoarterial ECMO again because of the progression of pulmonary hypertension. Laboratory tests showed increased concentrations of hepatic enzymes and hyperbilirubinemia (total bilirubin 31.6 mg/dL). Six hours after starting ECMO circulation, plasma leakage from the oxygenator occurred. Although we replaced the oxygenator with a new one, the replacement showed plasma leakage after 6 h. Disassembly of the oxygenator revealed congestion from bilirubin in the membrane fibers. We described a case of repeated, rapid-onset plasma leakage after implementation of ECMO. Hyperbilirubinemia was likely associated with the plasma leakage of this patient.

Changes in kidney perfusion and renal cortex metabolism in septic shock: an experimental study.

BACKGROUND: The etiology of renal dysfunction in sepsis is currently attributed to altered perfusion, microcirculatory abnormalities and cellular alterations. To clarify these mechanisms, we characterized the changes in renal perfusion and cortex metabolism in a large animal model of sepsis. METHODS: We studied 12 adult female sheep randomized to peritonitis-induced sepsis (n = 8) or to sham procedure (n = 4). A flow probe was positioned around the renal artery to measure renal blood flow (RBF). Laser Doppler was used to measure regional flow in the kidney cortex and medulla. A microdialysis probe was inserted into the renal cortex to measure cortical glucose, lactate, and pyruvate. Fluid resuscitation was provided to keep pulmonary artery occlusion pressure at baseline levels. All animals were observed for 18 h. RESULTS: Hypotension occurred after 9 h in the septic animals (P = 0.02 versus baseline). RBF and cortical flow were significantly lower than at baseline from 12 h in the septic animals (P = 0.01 and P = 0.03, respectively). Cortical lactate and pyruvate levels increased in the septic animals from 3 and from 6 h, respectively (both P = 0.02 versus baseline), and the L/P ratio from 15 h (P = 0.01). There was a correlation between cortical flow and cortical L/P ratio after shock onset (r = -0.60, P = 0.002) but not before. CONCLUSIONS: In this peritonitis model, sepsis was associated with metabolic alterations that may reflect early induction of cortical glycolysis. Septic shock was associated with reduced renal perfusion and decreased cortical and medullary blood flow, followed by signs of anaerobic metabolism in the cortex when flow reductions became critical.

Esmolol Administration to Control Tachycardia in an Ovine Model of Peritonitis.

BACKGROUND: Excessive adrenergic signaling may be harmful in sepsis. Using ß-blockers to reduce sympathetic overactivity may modulate sepsis-induced cardiovascular, metabolic, immunologic, and coagulation alterations. Using a randomized ovine fecal peritonitis model, we investigated whether administration of a short-acting ß-blocker, esmolol, could control tachycardia without deleterious effects on hemodynamics, renal perfusion, cerebral perfusion, cerebral metabolism, or outcome. METHODS: After induction of fecal peritonitis, 14 anesthetized, mechanically ventilated, and hemodynamically monitored adult female sheep were randomly assigned to receive a continuous intravenous infusion of esmolol to control heart rate between 80 and 100 bpm (n = 7) or a saline infusion (control group, n = 7). Esmolol was discontinued when the mean arterial pressure decreased below 60 mm Hg. Fluid resuscitation was titrated to maintain pulmonary artery occlusion pressure at baseline values. Left renal blood flow and cerebral cortex perfusion and metabolism were monitored in addition to standard hemodynamic variables. RESULTS: Esmolol was infused for 11 (9-14) hours; the target heart rate (80-100 bpm) was achieved between 3 and 8 hours after feces injection. In the first 5 hours after the start of the infusion, the decrease in heart rate was compensated by an increase in stroke volume index; later, stroke volume index was not statistically significantly different in the 2 groups, so that the cardiac work index was lower in the esmolol than in the control group. Hypotension (mean arterial pressure <60 mm Hg) occurred earlier (10 [8-12] vs 14 [11-20] hours; P= .01) in the esmolol group than in the control animals. Renal blood flow decreased earlier in the esmolol group, but there were no differences in urine output, cerebral cortex perfusion, metabolism, or survival between the groups. CONCLUSIONS: In this ovine model of abdominal sepsis, early control of tachycardia by esmolol was associated with a transient increase in stroke volume, followed by earlier hypotension. There were no significant effects of esmolol on cerebral perfusion, metabolism, urine output, or survival.

The effects of acute renal denervation on kidney perfusion and metabolism in experimental septic shock.

BACKGROUND: Perfusion deficits likely play an important role in the development of renal dysfunction in sepsis. Renal denervation may improve kidney perfusion and metabolism. METHODS: We randomized 14 female sheep to undergo bilateral surgical renal denervation (n = 7) or sham procedure (n = 7) prior to induction of sepsis. Renal blood flow (RBF) was measured with a pre-calibrated flowprobe. Laser Doppler probes were implanted to measure cortical and medullary perfusion. Cortical glucose, lactate and pyruvate levels were measured using the microdialysis technique. Creatinine clearance was determined. Sepsis was induced by peritonitis and fluid resuscitation was provided to avoid hypovolemia. RESULTS: RBF and cortical perfusion were higher in the denervated group during the first 6 h after induction of sepsis (P < 0.001 and P < 0.05, respectively), while medullary perfusion decreased similarly in both groups. After hypotension developed, RBF decreased to similar levels in both groups. Cortical pyruvate and lactate levels were lower in the denervated animals (P < 0.001 and P < 0.001, respectively). There were no differences between groups in creatinine clearance, urine output or time to oliguria. CONCLUSION: Denervation thus caused an early increase in RBF that was distributed towards the kidney cortex. Although associated with an attenuation of early cortical metabolic alterations, denervation failed to prevent the deterioration in renal function.

Effects of Different Crystalloid Solutions on Hemodynamics, Peripheral Perfusion, and the Microcirculation in Experimental Abdominal Sepsis.

BACKGROUND: Crystalloid solutions are used to restore intravascular volume in septic patients, but each solution has limitations. The authors compared the effects of three crystalloid solutions on hemodynamics, organ function, microcirculation, and survival in a sepsis model. METHODS: Peritonitis was induced by injection of autologous feces in 21 anesthetized, mechanically ventilated adult sheep. After baseline measurements, animals were randomized to lactated Ringer's (LR), normal saline (NS), or PlasmaLyte as resuscitation fluid. The sublingual microcirculation was assessed using sidestream dark field videomicroscopy and muscle tissue oxygen saturation with near-infrared spectroscopy. RESULTS: NS administration was associated with hyperchloremic acidosis. NS-treated animals had lower cardiac index and left ventricular stroke work index than LR-treated animals from 8 h and lower mean arterial pressure than LR-treated animals from 12 h. NS-treated animals had a lower proportion of perfused vessels than LR-treated animals after 12 h (median, 82 [71 to 83] vs. 85 [82 to 89], P = 0.04) and greater heterogeneity of proportion of perfused vessels than PlasmaLyte or LR groups at 18 h. Muscle tissue oxygen saturation was lower at 16 h in the NS group than in the other groups. The survival time of NS-treated animals was shorter than that of the LR group (17 [14 to 20] vs. 26 [23 to 29] h, P < 0.01) but similar to that of the PlasmaLyte group (20 [12 to 28] h, P = 0.74). CONCLUSIONS: In this abdominal sepsis model, resuscitation with NS was associated with hyperchloremic acidosis, greater hemodynamic instability, a more altered microcirculation, and more severe organ dysfunction than with balanced fluids. Survival time was shorter than in the LR group.

Timing of tracheotomy in ICU patients: a systematic review of randomized controlled trials.

INTRODUCTION: The optimal timing of tracheotomy in critically ill patients remains a topic of debate. We performed a systematic review to clarify the potential benefits of early versus late tracheotomy. METHODS: We searched PubMed and CENTRAL for randomized controlled trials that compared outcomes in patients managed with early and late tracheotomy. A random-effects meta-analysis, combining data from three a priori-defined categories of timing of tracheotomy (within 4 versus after 10 days, within 4 versus after 5 days, within 10 versus after 10 days), was performed to estimate the weighted mean difference (WMD) or odds ratio (OR). RESULTS: Of the 142 studies identified in the search, 12, including a total of 2,689 patients, met the inclusion criteria. The tracheotomy rate was significantly higher with early than with late tracheotomy (87 % versus 53 %, OR 16.1 (5.7-45.7); p <0.01). Early tracheotomy was associated with more ventilator-free days (WMD 2.12 (0.94, 3.30), p <0.01), a shorter ICU stay (WMD -5.14 (-9.99, -0.28), p = 0.04), a shorter duration of sedation (WMD -5.07 (-10.03, -0.10), p <0.05) and reduced long-term mortality (OR 0.83 (0.69-0.99), p = 0.04) than late tracheotomy. CONCLUSIONS: This updated meta-analysis reveals that early tracheotomy is associated with higher tracheotomy rates and better outcomes, including more ventilator-free days, shorter ICU stays, less sedation, and reduced long-term mortality, compared to late tracheotomy.

Ultrasound Imaging Reduces Failure Rates of Percutaneous Central Venous Catheterization in Children.

OBJECTIVE: Ultrasound imaging has been shown to be beneficial for percutaneous central venous cannulation in systematic reviews of randomized controlled trials in adult patients, but not in pediatrics. The aim of this updated review was to determine whether percutaneous central venous catheterization with the aid of ultrasound reduces cannulation failure in children. DATA SOURCES: PubMed was searched using the terms: ultrasound, catheterization, central vein (including internal jugular and femoral veins), and pediatrics. STUDY SELECTION: Both nonrandomized comparative studies and randomized controlled trials were eligible for inclusion if they assessed the rate of cannulation failure using real-time, dynamic ultrasound guidance, ultrasound-assisted vein prelocation, and/or anatomic landmark technique. DATA EXTRACTION: Five nonrandomized studies and nine randomized controlled trials were included. The rates of cannulation failure and arterial puncture were retrieved. DATA SYNTHESIS: Random-effects meta-analysis was applied. CONCLUSIONS: The meta-analysis of five nonrandomized studies showed that the rate of cannulation failure was significantly lower with real-time ultrasound guidance than anatomic landmark technique (odds ratio, 0.44 [95% CI, 0.27-0.72]; p = 0.001). The combination of nine randomized controlled trials also showed lower failure rates with either the real-time ultrasound guidance or the prelocation technique over the landmark technique (odds ratio, 0.22 [95% CI, 0.07-0.69]; p = 0.0003) and fewer arterial punctures in the ultrasound group (odds ratio, 0.31 [95% CI, 0.09-1.08]; p = 0.07). However, seven out of nine studies were assessed as having high risk of bias. Since the lower cannulation failure and less frequent chance of arterial puncture with ultrasound were predominantly shown in studies at high risk of bias, further definitive and adequately powered studies with clear outcomes are needed.

Clinical neurophysiological assessment of sepsis-associated brain dysfunction: a systematic review.

INTRODUCTION: Several studies have reported the presence of electroencephalography (EEG) abnormalities or altered evoked potentials (EPs) during sepsis. However, the role of these tests in the diagnosis and prognostic assessment of sepsis-associated encephalopathy remains unclear. METHODS: We performed a systematic search for studies evaluating EEG and/or EPs in adult (≥ 18 years) patients with sepsis-associated encephalopathy. The following outcomes were extracted: a) incidence of EEG/EP abnormalities; b) diagnosis of sepsis-associated delirium or encephalopathy with EEG/EP; c) outcome. RESULTS: Among 1976 citations, 17 articles met the inclusion criteria. The incidence of EEG abnormalities during sepsis ranged from 12% to 100% for background abnormality and 6% to 12% for presence of triphasic waves. Two studies found that epileptiform discharges and electrographic seizures were more common in critically ill patients with than without sepsis. In one study, EEG background abnormalities were related to the presence and the severity of encephalopathy. Background slowing or suppression and the presence of triphasic waves were also associated with higher mortality. A few studies demonstrated that quantitative EEG analysis and EP could show significant differences in patients with sepsis compared to controls but their association with encephalopathy and outcome was not evaluated. CONCLUSIONS: Abnormalities in EEG and EPs are present in the majority of septic patients. There is some evidence to support EEG use in the detection and prognostication of sepsis-associated encephalopathy, but further clinical investigation is needed to confirm this suggestion.

A Pilot Estimation of Ventricular-Arterial Coupling Using a Vascular Screening Device (VaSera®).

BACKGROUND: Non-invasive cardiovascular assessment has become an alternative to invasive techniques. VaSera®, a vascular screening device, measures arterial stiffness with the cardio-ankle vascular index (CAVI); it also measures cardiophysiological variables of ejection time (ET) and pre-ejection period (PEP). We aimed to apply the parameters obtained by VaSera® to estimate heart function based on left ventricular end-systolic elastance/arterial elastance (Ees/Ea) and to assess the minimal required number of measurements for estimation. METHODS: We conducted an experimental laboratory study for healthy volunteers. Using the previously established formula, the Ees/Ea value of each participant was estimated using ET and PEP values measured by VaSera®. The intraclass correlation coefficient (ICC) assessed the minimum required number of measurements. Concordance correlation coefficient (CCC) and Bland and Altman analysis assessed variation of Ees/Ea estimation against the trimmed average. RESULTS: A total of 660 measurements from 132 participants were included. The Ees/Ea estimates from the VaSera® were 1.5 [1.2, 1.9]. The ICC for Ees/Ea was 0.71 (95% confidence interval: 0.65-0.77), suggesting that four measurements were required. The CCC between the trimmed average of Ees/Ea and the mean of four Ees/Ea estimates was 0.99. Bland and Altman analysis showed excellent agreement for the mean of four Ees/Ea estimates and the trimmed average of Ees/Ea. CONCLUSIONS: For screening of heart failure, the Ees/Ea estimated using non-invasive vascular-stiffness assessment device would be tolerable and four sequential measurements were required.

GENETIC DELETION OF TRANSLOCATOR PROTEIN EXACERBATES POST-SEPSIS SYNDROME WITH ACTIVATION OF THE C1Q PATHWAY IN SEPTIC MOUSE MODEL.

ABSTRACT: Significant numbers of patients who survive sepsis exhibit psychiatric and cognitive impairments, termed post-sepsis syndrome. Understanding the underlying pathophysiology is essential to develop effective therapies. Translocator protein 18 kDa (TSPO) is a multifaceted mitochondrial protein implicated in inflammation, oxidative stress, and steroidogenesis in the central nervous system. Despite accumulated evidence demonstrating TSPO is a biomarker in psychiatric and neurodegenerative disorders, the role of this protein in post-sepsis syndrome remains elusive. The aim of this study was to investigate the role of TSPO in the long-term impairment of mouse behavior associated with psychiatric and cognitive impairments following sepsis induced by cecal ligation and puncture (CLP) surgery. Animals were divided into three groups: (i) wild type (WT) + sham, (ii) WT + CLP, and (iii) TSPO knock out + CLP. Survival rate and body weight change were assessed up to 17 days after surgeries. Then, we also assessed anxiety-like behavior, depression-like behavior, cognitive function, locomotor activity, and forelimb muscle strength in surviving mice by elevated plus maze, tail suspension test, y-maze, open field test, and grip strength test, respectively. Deletion of the TSPO gene led to high mortality and prolonged weight loss and exacerbated anxiety-like and depressive-like behavior with cognitive impairment 17 days after, but not before, CLP surgery. RNA-seq analysis of the hippocampus revealed the upregulation of genes (C1qb, C1qc, and Tyrobp) in C1q complement pathways correlated significantly with anxiety-like behavior that appeared long after CLP surgery. The expressions of these genes predicted other behavioral traits, including depressive-like behavior in the tail suspension test and grip power impairment, supporting the role of the C1q pathway in post-sepsis syndrome. Because the C1q pathway has recently attracted interest as a tag for pathological synaptic elimination, the current study suggests the C1q pathway is involved in the psychiatric and cognitive impairments observed in post-sepsis syndrome.

Association of body temperature and antipyretic treatments with mortality of critically ill patients with and without sepsis: multi-centered prospective observational study.

INTRODUCTION: Fever is frequently observed in critically ill patients. An independent association of fever with increased mortality has been observed in non-neurological critically ill patients with mixed febrile etiology. The association of fever and antipyretics with mortality, however, may be different between infective and non-infective illness. METHODS: We designed a prospective observational study to investigate the independent association of fever and the use of antipyretic treatments with mortality in critically ill patients with and without sepsis. We included 1,425 consecutive adult critically ill patients (without neurological injury) requiring >48 hours intensive care admitted in 25 ICUs. We recorded four-hourly body temperature and all antipyretic treatments until ICU discharge or 28 days after ICU admission, whichever occurred first. For septic and non-septic patients, we separately assessed the association of maximum body temperature during ICU stay (MAXICU) and the use of antipyretic treatments with 28-day mortality. RESULTS: We recorded body temperature 63,441 times. Antipyretic treatment was given 4,863 times to 737 patients (51.7%). We found that treatment with non-steroidal anti-inflammatory drugs (NSAIDs) or acetaminophen independently increased 28-day mortality for septic patients (adjusted odds ratio: NSAIDs: 2.61, P=0.028, acetaminophen: 2.05, P=0.01), but not for non-septic patients (adjusted odds ratio: NSAIDs: 0.22, P=0.15, acetaminophen: 0.58, P=0.63). Application of physical cooling did not associate with mortality in either group. Relative to the reference range (MAXICU ≥ 39.5°C increased risk of 28-day mortality in non-septic patients (adjusted odds ratio 8.14, P=0.01), but not in septic patients (adjusted odds ratio 0.47, P=0.11) [corrected]. CONCLUSIONS: In non-septic patients, high fever (≥39.5°C) independently associated with mortality, without association of administration of NSAIDs or acetaminophen with mortality. In contrast, in septic patients, administration of NSAIDs or acetaminophen independently associated with 28-day mortality, without association of fever with mortality. These findings suggest that fever and antipyretics may have different biological or clinical or both implications for patients with and without sepsis. TRIAL REGISTRATION: ClinicalTrials.gov: NCT00940654.