Death In Intensive Care Unit Patients Research Articles

Association of glycemic variability and prognosis in patients with traumatic brain injury: A retrospective study from the MIMIC-IV database

BackgroundElevated glycemic variability (GV) often occurs in intensive care unit (ICU) patients and is associated with patient prognosis. However, the association between GV and prognosis in ICU patients with traumatic brain injury (TBI) remains unclear. MethodClinical data of ICU patients with TBI were obtained from the Medical Information Mart for Intensive Care (MIMIC) -IV database. The coefficient of variation (CV) was utilized to quantify GV, while the Glasgow Coma Scale (GCS) was employed to evaluate the consciousness status of TBI patients. Pearson linear correlation analysis, linear regression, COX regression and restricted cubic spline (RCS) were used to investigate the relationship between CV and consciousness impairment, as well as the risk of in-hospital mortality. ResultA total of 1641 ICU patients with TBI were included in the study from the MIMIC-IV database. Pearson linear correlation and restricted cubic spline (RCS) analysis results showed a negative linear relationship between CV and the last GCS (P = 0.002) with no evidence of nonlinearity (P for nonlinear = 0.733). Multivariable linear regression suggested a higher CV was associated with a lower discharge GCS [β (95 %CI) = −1.86 (−3.08 ∼ −0.65), P = 0.003]. Furthermore, multivariable COX regression indicated that CV ≥ 0.3 was a risk factor for in-hospital death in TBI patients [HR (95 %CI) = 1.74 (1.15–2.62), P = 0.003], and this result was also consistent across sensitivity and subgroup analyses. ConclusionHigher GV is related to poorer consciousness outcomes and increased risk of in-hospital death in ICU patients with TBI. Additional research is needed to understand the logical relationship between GV and TBI progression.

Prospective multicenter study to describe the prevalence, outcomes, and management of phosphate disorders in intensive care patients: Study protocol for part B of the international GUTPHOS study

BackgroundAberrations in blood phosphate (Pi) levels, whether presenting as hypo- or hyperphosphatemia, appear to be associated with clinical complications and adverse outcomes in patients admitted to an intensive care unit (ICU). However, the prevalence of Pi disorders and the association with subsequent factors and organ failures leading to death in ICU patients are poorly described. Despite endeavors to understand the etiology and treatment of low Pi levels from systematic reviews and meta-analyses, the literature lacks comprehensive guidance for managing hypophosphatemia. Hyperphosphatemia, on the other hand, appears to be associated with higher mortality among critically ill patients, yet its prevalence among ICU patients, particularly following phosphate repletion, remains unknown. The present study aims to investigate the prevalence of Pi abnormalities upon ICU admission and their incidence during the first week of ICU stay, the factors associated with Pi alterations, and the effect of phosphate repletion on the normalization of Pi levels, and its associations with clinical outcomes. MethodsThis multicentre, prospective, non-interventional cohort study will include at least 1000 consecutive adult ICU patients (≥18 years) as part B of the GUTPHOS study. Sites are eligible if an anticipated minimal inclusion of 50 eligible patients during eight weeks from January 2024 until June 2024 and daily phosphate measurements during the first seven days of ICU stay are expected. All consecutive adult patients admitted to a participating ICU during the recruitment period, lasting up to eight weeks, or up to 120 patients if enrollment reaches that limit earlier, will be included. Study parameters include study site characteristics, patient demographics, daily assessment of Pi levels, Pi-related treatment, feeding details, renal replacement therapy details, the incidence of refeeding-associated hypophosphatemia and administered medication (during the first seven calendar days of ICU stay). There will be a follow-up period of a maximum of 90 days to document 28- and 90-day all-cause mortality as the primary outcome. Multiple logistic regression will be used to assess independent associations with mortality in addition to Receiver Operating Characteristics curves to identify cut-off Pi values associated with mortality and overcorrection. Linear mixed models will be conducted to assess Pi treatment effects. Subgroup analyses will be performed based on Pi abnormalities observed during ICU admission, categorized as normo-, hypo-, hyper-, or mixed, along with its severity (mild, moderate, or severe). DiscussionThe GUTPHOS study will be the first multicentre, prospective observational cohort study to investigate the prevalence, management practices, and consequent outcomes associated with Pi abnormalities during the first week of ICU admission. Its results may bridge the current evidence gap in repletion protocols while establishing the groundwork for a subsequent randomized controlled trial. Clinical trial registryNCT05909722.

A multicenter clinical study of critically ill patients with sepsis complicated with acute kidney injury in Beijing: incidence, clinical characteristics and outcomes

To investigate the epidemiological characteristics and prognosis of critically ill patients with sepsis combined with acute kidney injury (AKI) in intensive care unit (ICU) in Beijing, and to analyze the risk factors associated with in-hospital mortality among these critically ill patients. Data were collected from the Beijing AKI Trial (BAKIT) database, including 9 049 patients consecutively admitted to 30 ICUs in 28 tertiary hospitals in Beijing from March 1 to August 31, 2012. Patients were divided into non-AKI and non-sepsis group, AKI and non-sepsis group, non-AKI and sepsis group, AKI and sepsis group. Clinical data recorded included demographic characteristics, primary reasons for ICU admission, comorbidities, sequential organ failure assessment (SOFA), acute physiology and chronic health evaluation II(APACHE II) within 24 hours of ICU admission, physiological and laboratory indexes, treatment in the ICU, AKI staging based on the Kidney Disease: Improving Global Outcomes (KDIGO), as well as the prognostic indicators including length of stay in ICU, length of stay in hospital, ICU and in-hospital mortality. The primary endpoint was discharge or in-hospital death. Multivariate Logistic regression analysis was used to investigate the risk factors for hospital death in ICU patients. Kaplan-Meier survival curve was drawn to analyze the cumulative survival of ICU patients during hospitalization. A total of 3 107 critically ill patients were ultimately enrolled, including 1 259 cases in the non-AKI and non-sepsis group, 931 cases in the AKI and non-sepsis group, 264 cases in the non-AKI and sepsis groups, and 653 cases in the AKI and sepsis group. Compared with the other three group, patients in the AKI and sepsis group were the oldest, had the lowest mean arterial pressure (MAP), and the highest APACHE II score, SOFA score, blood urea nitrogen (BUN), and serum creatinine (SCr) levels, and they also had the highest proportion of receiving mechanical ventilation, requiring vasopressor support, and undergoing renal replacement therapy (RRT), all P < 0.01. Of these 3 107 patients, 1 584 (51.0%) were diagnosed with AKI, and the incidence of AKI in patients with sepsis was significantly higher than in those without sepsis [71.2% (653/917) vs. 42.5% (931/2 190), P < 0.01]. The highest proportion of KDIGO 0 stage was observed in the non-sepsis group (57.5%), while the highest proportion of KDIGO 3 stage was observed in the sepsis group (32.2%). Within the same KDIGO stage, the mortality of patients with sepsis was significantly higher than that of non-sepsis patients (0 stage: 17.8% vs. 3.1%, 1 stage: 36.3% vs. 7.4%, 2 stage: 42.7% vs. 17.1%, 3 stage: 54.6% vs. 28.6%, AKI: 46.1% vs. 14.2%). The ICU mortality (38.7%) and in-hospital mortality (46.1%) in the AKI and sepsis group were significantly higher than those in the other three groups. Kaplan-Meier survival curves further showed that the cumulative survival rate of patients with AKI and sepsis during hospitalization was significantly lower than that of the other three groups (53.9% vs. 96.9%, 85.8%, 82.2%, Log-Rank: χ 2 = 379.901, P < 0.001). Subgroup analysis showed that among surviving patients, length of ICU stay and total length of hospital stay were significantly longer in the AKI and sepsis group than those in the other three groups (both P < 0.01). Multivariate regression analysis showed that age, APACHE II score and SOFA score within 24 hours of ICU admission, coronary heart disease, AKI, sepsis, and AKI combined with sepsis were independent risk factors for ICU mortality in patients (all P < 0.05). After adjusting for covariates, AKI, sepsis, and sepsis combined with AKI were significantly associated with higher ICU and in-hospital mortality, with the highest ICU mortality [adjusted odds ratio (OR) = 14.82, 95% confidence interval (95%CI) was 8.10-27.12; Hosmer-Lemeshow test: P = 0.816] and in-hospital mortality (adjusted OR = 7.40, 95%CI was 4.94-11.08; Hosmer-Lemeshow test: P = 0.708) observed in patients with sepsis combined with AKI. The incidence of AKI is high in sepsis patients, and those with both AKI and sepsis have a higher disease burden, more abnormalities in physiological and laboratory indicators, and significantly increased ICU and in-hospital mortality. Among surviving patients, the length of ICU stay and total length of hospital stay are also longer in the AKI and sepsis group. Age, APACHE II score and SOFA score within 24 hours of ICU admission, coronary heart disease, AKI, and sepsis are independent risk factors for in-hospital mortality in ICU patients.

The association between serum chloride and mortality in ICU patients with heart failure: The impact of bicarbonate

ObjectiveTo assess whether serum chloride predicts risk of death in intensive care unit (ICU) patients with heart failure (HF) and the effect of bicarbonate on the efficacy of serum chloride in predicting risk of death in ICU patients. MethodsA total of 9364 HF patients hospitalized in the ICU were enrolled. Patients were divided into hypochloremia (< 96 mEq/L), normal chloride (96–108 mEq/L), and hyperchloremia (> 108 mEq/L) groups. Similarly, we divided the serum bicarbonate level into three groups: low bicarbonate (< 22 mEq/L), medium bicarbonate (22–26 mEq/L), and high bicarbonate (> 26 mEq/L). The outcome of this study was in-hospital mortality. Then, we analyzed the association between abnormal serum chloride and mortality according to the category of serum bicarbonate and assessed the interaction effect. Restricted cubic spline (RCS) was used to show possible nonlinear associations. ResultsIn the overall study population, hypochloremia was associated with a higher risk of in-hospital mortality than normal chloride (odds ratio [OR] 1.54, 95% confidence interval [CI] 1.26–1.86, P < 0.001), hyperchloremia was not significantly related to in-hospital mortality (OR 1.00, 95% CI 0.85–1.19, P = 0.962). However, a linear association between serum chloride and in-hospital mortality was found in the low and normal bicarbonate groups (all P for nonlinear >0.05). ConclusionsHypochloremia is associated with in-hospital mortality and longer hospital stay in critically ill patients with HF. In addition, risk of death in the low and medium serum bicarbonate groups decreased with increasing serum chloride level.

Glycemic variability and in-hospital death of critically ill patients and the role of ventricular arrhythmias

BackgroundAbnormal glycemic variability is common in the intensive care unit (ICU) and is associated with increased in-hospital mortality and major adverse cardiovascular events, but little is known about whether adverse outcomes are partly mediated by ventricular arrhythmias (VA). We aimed to explore the association between glycemic variability and VA in the ICU and whether VA related to glycemic variability mediate the increased risk of in-hospital death.MethodsWe extracted all measurements of blood glucose during the ICU stay from The Medical Information Mart for Intensive Care IV (MIMIC-IV) database version 2.0. Glycemic variability was expressed by the coefficient of variation (CV), which was calculated by the ratio of standard deviation (SD) and average blood glucose values. The outcomes included the incidence of VA and in-hospital death. The KHB (Karlson, KB & Holm, A) is a method to analyze the mediation effect for nonlinear models, which was used to decompose the total effect of glycemic variability on in-hospital death into a direct and VA-mediated indirect effect.ResultsFinally, 17,756 ICU patients with a median age of 64 years were enrolled; 47.2% of them were male, 64.0% were white, and 17.8% were admitted to the cardiac ICU. The total incidence of VA and in-hospital death were 10.6% and 12.8%, respectively. In the adjusted logistic model, each unit increase in log-transformed CV was associated with a 21% increased risk of VA (OR 1.21, 95% CI: 1.11–1.31) and a 30% increased risk (OR 1.30, 95% CI: 1.20–1.41) of in-hospital death. A total of 3.85% of the effect of glycemic variability on in-hospital death was related to the increased risk of VA.ConclusionHigh glycemic variability was an independent risk factor for in-hospital death in ICU patients, and the effect was caused in part by an increased risk of VA.

Development and validation of a prognostic nomogram model for ICU patients with alcohol-associated cirrhosis.

The prognosis of patients with alcohol-associated cirrhosis (ALC) admitted to the intensive care unit (ICU) is poor. We developed and validated a nomogram (NIALC) for ICU patients with ALC. Predictors of mortality were defined by a machine learning method in a cohort of 394 ICU patients with ALC from the Medical Information Mart for Intensive Care database. Then the nomogram (NIALC) was constructed and evaluated using the AUC. The MELD, MELD-sodium, Child-Pugh, and CLIF-SOFA scores were then compared with NIALC. Two datasets of 394 and 501 ICU patients with ALC were utilized for model validation. In-hospital mortality was 41% and 21% in the training and external validation sets. Predictors included were blood urea nitrogen, total bilirubin, prothrombin time, serum creatinine, lactate, partial thromboplastin time, phosphate, mean arterial pressure, lymphocytes, fibrinogen, and albumin. The AUCs for the NIALC were 0.767 and 0.760 in the two validation cohorts, which were better than those of the MELD, MELD-sodium, Child-Pugh, and CLIF-SOFA. We developed a nomogram for ICU patients with ALC, which demonstrated better discriminative ability than previous prognostic scores. This nomogram could be conveniently used to facilitate the individualized prediction of death in ICU patients with ALC.

Serum mAST/ALT ratio had high predictive value for adverse outcome of severe fever with thrombocytopenia syndrome with severe condition

BackgroundSevere fever with thrombocytopenia syndrome (SFTS) usually demonstrates multi-organ injury with a high mortality rate. This study aimed to investigate associations of serum aspartate/alanine aminotransferase (AST)/ALT, cytosolic AST (cAST)/ALT and mitochondrial AST (mAST)/ALT ratios with the prognosis of SFTS patients.MethodsA total of 355 confirmed SFTS patients were included. Clinical and laboratory data were compared between survivors and nonsurvivors. Logistic regression analysis was used to assess the independent risk factors for fatality in all patients and those admitted to the intensive care unit (ICU). The predictive values of the risk factors and constructed risk models were evaluated.ResultsMean age and biochemical parameters were significantly greater in nonsurvivors than in survivors. In ICU patients, the three ratios, high-sensitivity troponin I (hsTnI), creatine kinase (CK), lactate dehydrogenase (LDH) and α-hydroxybutyrate dehydrogenase (α-HBDH) were elevated markedly in nonsurvivors than in survivors. Multivariate logistic regression analysis showed that age, three ratios and α-HBDH were independent risk factors for mortality in all patients. Only the three ratios were independent risk factors for death in ICU patients. Risk Models (M1, M2 and M3) and simplified models (sMs) containing the three ratios respectively had comparatively high predictive values for fatality in all patients with area under ROC curves (AUCs) > 0.85. In ICU patients, mAST/ALT ratio had the highest predictive value, sensitivity and odds ratio (OR) for mortality among three ratios.ConclusionAST/ALT, cAST/ALT and mAST/ALT ratios were associated with unfavorable clinical outcome of SFTS. The prognostic value of mAST/ALT ratio was higher in severe cases.

Age-adjusted Charlson Comorbidity Index as effective predictor for in-hospital mortality of patients with cardiac arrest: a retrospective study

BackgroundCardiac arrest is currently one of the leading causes of mortality in clinical practice, and the Charlson Comorbidity Index (CCI) is widely utilized to assess the severity of comorbidities. We aimed to evaluate the relationship between the age-adjusted CCI score and in-hospital mortality in intensive care unit (ICU) patients with the diagnosis of cardiac arrest, which is important but less explored previously.MethodsThis was a retrospective study including patients aged over 18 years from the MIMIC-IV database. We calculated the age-adjusted CCI using age information and ICD codes. The univariate analysis for varied predictors’ differences between the survival and the non-survival groups was performed. In addition, a multiple factor analysis was conducted based on logistic regression analysis with the primary result set as hospitalization death. An additional multivariate regression analysis was conducted to estimate the influence of hospital and ICU stay.ResultsA total of 1772 patients were included in our study, with median age of 66, among which 705 (39.8%) were female. Amongst these patients, 963 (54.3%) died during the hospitalization period. Patients with higher age-adjusted CCI scores had a higher likelihood of dying during hospitalization (P < 0.001; OR: 1.109; 95% CI: 1.068–1.151). With the age-adjusted CCI incorporated into the predictive model, the area under the receiver operating characteristic curve was 0.794 (CI: 0.773–0.814), showing that the prediction model is effective. Additionally, patients with higher age-adjusted CCI scores stayed longer in the hospital (P = 0.026, 95% CI: 0.056–0.896), but there was no significant difference between patients with varied age-adjusted CCI scores on the days of ICU stay.ConclusionThe age-adjusted CCI is a valid indicator to predict death in ICU patients with cardiac arrest, which can offer enlightenment for both theory literatures and clinical practice.

Risk Factors Associated with in-Hospital COVID-19 Mortality in a Public Hospital in Brazil

Background: Early studies reported laboratory changes and hematologic abnormalities in severe cases of COVID-19, although the relation between these findings and the course of the disease is not well established. Methods: In this observational, retrospective study we aimed to analyze clinical, biochemical, and hematological findings in hospitalized adult patients with COVID-19 that were admitted in the University Hospital of Brasilia from April 2020 to December 2021.The first (1st) and second (2nd) waves dates were defined in accordance with Johns Hopkins University. Study was approved by the local ethics committee. All records of hospitalized patients with positive RT-PCR for SARS-CoV-2 were analyzed (N=760). A total of 662 were included in this study (table 1). Patients were dived in two groups according to outcome (death or hospital discharge). The analysis plan (table 2) included continuous variables (e.g., age), as well as categorical variables (e.g., intensive care unit (ICU) admission, comorbidities, bleeding, thrombosis, transfusion need, mechanical ventilation (MV) and renal replacement therapy (RRT)). We also analyzed hemoglobin (Hb), leukocyte levels, erythroblastosis, platelet count, coagulation tests, d-dimer (DD) and ferritin. Univariate and multivariate analysis were performed to identify variables that were significantly associated (p<0.05) with outcome. Results: Study results (table 2) demonstrated a significant association between death outcome and the following variables: age (68 vs 58 years, p<0.001), lower Hb levels (p<0.0001), erythroblastosis (p<0.0001, OR 8.3), leukocytosis (p<0.0001, OR 7.9), lymphopenia (p<0.0001), thrombocytopenia (p<0.0001, OR 5.2), elevated DD (p=0.0001, OR 7.5), transfusion necessity (p<0.0001, OR 3.9) and high ferritin levels (p=0.0002), especially above 1000ng/mL (p<0.0001, OR 3.7). Patients admitted to the ICU were 22.8 and 10 times more likely to die, respectively, in the 1st and 2nd waves. Reduction in the risk of death in ICU patients between waves are probably due to technical improvement and vaccination. The presence of erythroblastosis was associated with a 12.5- and 6.4-times greater risk of death, respectively, in the 1st and 2nd waves (p<0.0001). The medians of the lowest and highest leukocyte (respectively, p<0.0001, p=0.0395) and neutrophil levels (respectively, p<0.0001, p=0.0369) were significantly higher in the group that died. Thrombocytopenia below 50,000/mm³ increased the risk of death in the 1st and 2nd waves (both with p<0.0001; OR: 16.1 vs 7.2). Thrombocytosis reduced the risk of death by 40%. DD levels 4 or 6 times above the reference limit increased the odds of death by 6.6 times (both with p<0.0001). Proven arterial and/or venous thrombosis were more common in the group that died (24.5% vs 16%, p=0.0103, OR: 1.7). Likewise, therapeutic heparin was used significantly more in this group (p<0.0001, OR: 3.5), and therapeutic heparin increased the risk of bleeding in the overall population (p=0.0170, OR: 1.7). Bleeding, including those that required transfusion, were significantly more frequent in the group that died (respectively, 32.3% vs 10.4%, p<0.0001, OR: 4.1 and 20.8% vs 3.6%, p<0.0001, OR: 7.0). Transfusion of red blood cells, platelet concentrates (both with p<0.0001), and fresh frozen plasma (p=0.0003) were significantly more frequent in the group that died. Comorbidities were also significantly higher in this group (91.7% vs 85.6%, p=0.0432, OR: 1.8) with emphasis on heart failure (in the 2nd wave, p=0.04, OR: 1.79), previous coronary artery disease/stroke (p=0.0324, OR: 1.6), arrhythmia (p=0.0005, OR: 2.8), chronic liver disease (p=0.0191, OR: 3.4), hematologic neoplasm (p=0.024, OR: 3.5), chronic kidney disease (p<0.0001, OR: 2.3), kidney transplant (p=0.021, OR: 2.5) and pre-hospitalization RRT (2nd wave, p=0.0002, OR 2.9). Prevalence of hypertension, obesity, diabetes, and lung disease was not different between groups. In multivariate analysis, factors significantly associated with death were thrombocytopenia, erythroblastosis, DD greater than 6 times the normal limit, neutrophilia and MV. Conclusion: Data from this large retrospective single-center study demonstrated that thrombocytopenia, erythroblastosis, DD greater than 6 times the upper reference limits and neutrophilia were predictive factors of in-hospital mortality related to COVID-19. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Prediction of hospital mortality in intensive care unit patients from clinical and laboratory data: A machine learning approach.

BACKGROUNDIntensive care unit (ICU) patients demand continuous monitoring of several clinical and laboratory parameters that directly influence their medical progress and the staff’s decision-making. Those data are vital in the assistance of these patients, being already used by several scoring systems. In this context, machine learning approaches have been used for medical predictions based on clinical data, which includes patient outcomes.AIMTo develop a binary classifier for the outcome of death in ICU patients based on clinical and laboratory parameters, a set formed by 1087 instances and 50 variables from ICU patients admitted to the emergency department was obtained in the “WiDS (Women in Data Science) Datathon 2020: ICU Mortality Prediction” dataset. METHODSFor categorical variables, frequencies and risk ratios were calculated. Numerical variables were computed as means and standard deviations and Mann-Whitney U tests were performed. We then divided the data into a training (80%) and test (20%) set. The training set was used to train a predictive model based on the Random Forest algorithm and the test set was used to evaluate the predictive effectiveness of the model.RESULTSA statistically significant association was identified between need for intubation, as well predominant systemic cardiovascular involvement, and hospital death. A number of the numerical variables analyzed (for instance Glasgow Coma Score punctuations, mean arterial pressure, temperature, pH, and lactate, creatinine, albumin and bilirubin values) were also significantly associated with death outcome. The proposed binary Random Forest classifier obtained on the test set (n = 218) had an accuracy of 80.28%, sensitivity of 81.82%, specificity of 79.43%, positive predictive value of 73.26%, negative predictive value of 84.85%, F1 score of 0.74, and area under the curve score of 0.85. The predictive variables of the greatest importance were the maximum and minimum lactate values, adding up to a predictive importance of 15.54%.CONCLUSIONWe demonstrated the efficacy of a Random Forest machine learning algorithm for handling clinical and laboratory data from patients under intensive monitoring. Therefore, we endorse the emerging notion that machine learning has great potential to provide us support to critically question existing methodologies, allowing improvements that reduce mortality.

Comparison of pneumonia and nonpneumonia-related Acinetobacter baumannii complex bacteremia: A single-center retrospective study

Acinetobacter baumannii complex (ABC) is a group of increasingly prevalent opportunistic pathogens that cause a variety of life-threatening nosocomial infections, especially in the intensive care unit (ICU). This study assessed the differences between pneumonia- and non-pneumonia-related ABC bacteremia and possible independent risk factors for 30-day mortality. The clinical data of ICU patients diagnosed with ABC bacteremia at a tertiary care hospital from January 2009 to December 2020 were collected, and sorted into groups of ABC bacteremia with and without pneumonia. Significant changes in the incidence of ABC bacteremia and antibiotic resistance were observed over the 12-year study. Compared with nonpneumonia-related ABC bacteremia, pneumonia-related ABC bacteremia was associated with a higher rate of hypertension, less prior tigecycline use, more carbapenem-resistant (CR) strains, and a higher 30-day mortality rate. In multivariate analysis, immunosuppression, higher APACHE II score, and SOFA score were independent risk factors for 30-day mortality. Moreover, the risk of death was 1.919 times higher in the pneumonia-related group. Although pneumonia-related ABC bacteremia had worse outcomes, it was not an independent risk factor for death statistically. Immunosuppression and disease severity levels increased the risks of death in ICU patients with ABC bacteremia.

Angiotensin II induces reactive oxygen species, DNA damage, and T-cell apoptosis in severe COVID-19

BackgroundLymphopenia is predictive of survival in patients with coronavirus disease 2019 (COVID-19).ObjectiveThe aim of this study was to understand the cause of the lymphocyte count drop in severe forms of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.MethodsMonocytic production of reactive oxygen species (ROSs) and T-cell apoptosis were measured by flow cytometry, DNA damage in PBMCs was measured by immunofluorescence, and angiotensin II (AngII) was measured by ELISA in patients infected with SARS-CoV-2 at admission to an intensive care unit (ICU) (n = 29) or not admitted to an ICU (n = 29) and in age- and sex-matched healthy controls.ResultsWe showed that the monocytes of certain patients with COVID-19 spontaneously released ROSs able to induce DNA damage and apoptosis in neighboring cells. Of note, high ROS production was predictive of death in ICU patients. Accordingly, in most patients, we observed the presence of DNA damage in up to 50% of their PBMCs and T-cell apoptosis. Moreover, the intensity of this DNA damage was linked to lymphopenia. SARS-CoV-2 is known to induce the internalization of its receptor, angiotensin-converting enzyme 2, which is a protease capable of catabolizing AngII. Accordingly, in certain patients with COVID-19 we observed high plasma levels of AngII. When looking for the stimulus responsible for their monocytic ROS production, we revealed that AngII triggers ROS production by monocytes via angiotensin receptor I. ROSs released by AngII-activated monocytes induced DNA damage and apoptosis in neighboring lymphocytes.ConclusionWe conclude that T-cell apoptosis provoked via DNA damage due to the release of monocytic ROSs could play a major role in COVID-19 pathogenesis.

Healthcare-associated infections: a threat to the survival of patients with COVID-19 in intensive care units

BackgroundWide variation in mortality rates among critically ill patients with coronavirus disease 2019 (COVID-19) has been reported. This study evaluated whether healthcare-associated infections (HAI) are a risk factor for death among patients with severe COVID-19 in the intensive care unit (ICU).MethodsThis retrospective cohort study included patients with severe COVID-19 hospitalized in the ICU of four hospitals in the city of Curitiba, Brazil. Patients with COVID-19 who died during ICU hospitalization were compared with those who were discharged. A second analysis compared patients who developed HAI in the ICU with those who did not. Multiple logistic regression models were used to control for confounders.ResultsIn total, 400 patients were included, and 123 (31%) patients developed HAI. The most common HAI was lower respiratory tract infection (67%). Independent risk factors for death were: age [odds ratio (OR) 1.75, 95% confidence interval (CI) 1.43–2.15; P<0.0001]; clinical severity score (OR 2.21, 95% CI 1.70–2.87; P<0.0001); renal replacement therapy (OR 12.8, 95% CI 5.78–28.6; P<0.0001); and HAI (OR 5.9, 95% CI 3.31–10.5; P<0.0001). A longer interval between symptom onset and hospital admission was protective against death (OR 0.93, 95% CI 0.88–0.98; P=0.017). The only independent factors associated with HAI were high C-reactive protein and low PaO2/FiO2 ratio.ConclusionsNo factors that could point to a high-risk group for HAI acquisition were identified. However, age, dialysis and HAI increased the risk of death in ICU patients with severe COVID-19; of these, HAI is the only preventable risk factor.

The effects of midazolam or propofol plus fentanyl on ICU mortality: a retrospective study based on the MIMIC-IV database.

BackgroundPatients in the intensive care unit (ICU) are often under stress and fail to cooperate well with invasive treatments. Analgesia and sedation are of great significance for reducing the suffering of patients and ensuring the application and effectiveness of treatment. For better clinical choice, we aimed to explore the effect of the combination of propofol + fentanyl or midazolam + fentanyl on the short-term prognosis of hospitalized patients in the ICU.MethodsAccording to the inclusion and exclusion criteria, we retrospectively included patients in the MIMIC-IV database receiving midazolam + fentanyl or propofol + fentanyl analgesic and sedative treatment using Structured Query Language (SQL) to extract clinical data from the MIMIC-IV database. The primary endpoint was the death rate within 28 days after the patient was admitted to the ICU. Doubly robust estimation was used to infer the relationship between sedation and analgesia and 28 days outcome. The gradient boosted model (GBM) was used to estimate the propensity score (PS) of the patient’s sedation and analgesia program, PS was used as the weight, and the inverse probabilities weighting (IPW) model was used to generate a weighted cohort.ResultsIn total, 4,188 cases were included, with 2,174 (51.9%) in the propofol group and 2,014 (48.1%) in the midazolam group. In the PS score matching cohort, the 28-day mortality of patients in the midazolam group was 30.8%, and the 28-day mortality of patients in the propofol group was 25.5%. The adjusted odds ratio (OR) value was 1.421 [95% confidence interval (CI): 1.118–1.806, P<0.001]. Patients in the propofol group did not use vasoactive drugs for a longer period of time than the midazolam group, and patients in the propofol group received significantly more fluids than those in the midazolam group in the first three days after admission to the ICU.ConclusionsCompared with midazolam combined with fentanyl, propofol combined with fentanyl for sedation and analgesia can reduce the risk of short-term death in ICU patients.

Application of Deep Learning Technology in Predicting the Risk of Inpatient Death in Intensive Care Unit.

The Intensive Care Unit (ICU) is an important unit for the rescue of critically ill patients in hospitals, and patient mortality is an important indicator to measure the level of ICU treatment. Currently, a variety of clinical scoring systems are used to evaluate the patient's condition and predict survival, but these systems require a lot of resources. However, due to the rapid development of artificial intelligence and deep learning, machine learning based methods have been used to study the survival prediction of ICU patients. Additionally, these methods have made significant progress, but there is still a distance from clinical application, and equally metric interpretability of the deep learning method is not very mature. Therefore, in this paper, we have proposed a predicting model for the life and death of ICU patients, which is primarily based on the Fuzzy ARTMAP model. With a thorough analysis of the existing ICU patient condition assessment and life and death prediction methods, we have observed that patient's ICU monitoring information performs integrated analysis and extracts features according to the clinical characteristics of physiological indicators. Finally, fuzzy ARTMAP neural network is used to predict the life and death of patients. Likewise, prediction results are combined with the clinical scoring system and logistic regression, artificial neural network, support vector machine, and AdaBoost. Experimental results of these algorithms were compared, which verifies that the proposed method has outperformed the existing model. The main purpose of the proposed mode is to design a life and death prediction method for ICU patients, which has high predictive performance and is an acceptable method for clinical medical staff, where ICU monitoring data is used. Experimental results show that the method proposed has achieved better prediction performance and accuracy ratio, which provide theoretical reference for clinical application.

Association between CHA2DS2-VASc score and in-hospital death in ICU patients with COVID-19

Background: CHA2DS2-VASc score is a scientifically proven risk assessment score for patients with atrial fibrillation. It may be a good predictor of in-hospital death in COVID-19 patients. The present study aimed to evaluate the association between CHA2DS2-VASc score and in-hospital mortality in the prognosis of intensive care unit (ICU) patients with COVID-19.&#x0D; Methods: Eighty-four COVID-19 patients who were hospitalized in the ICU were retrospectively analyzed in a tertiary health care center, and the CHA2DS2-VASc score was determined. All analyses were performed using SPSS statistical software (SPSS Inc., Chicago, IL, USA, 20.0). A p-value &lt;0.05 was considered statistically significant.&#x0D; Results: The median age of patients was 60.0 years, and most were males (75.0%). Findings of the study showed that the CHA2DS2-VASc score was considerably higher among the hospitalized patients than discharged patients (3.08 ± 1.72 vs. 1.38 ± 1.16; p&lt;0.001), and patients who required mechanical ventilation compared to those who did not require mechanical ventilation (3.03 ± 1.68 vs. 1.15 ± 0.97; P &lt;0.001), respectively. Patients with CHA2DS2-VASc score of ≥3 had substantially higher age [67(45-87) vs. 58(19-75); P ˂0.001], computed tomography involvement score [67.5(20-90) vs. 35(15-90); P ˂0.001] and need for mechanical ventilation [29(90.6%) vs. 22(42.3%); P ˂0.001]. A significant difference was found in oxygen saturation on admission (P =0.001) between the two groups. In-hospital death was significantly higher among patients with a CHA2DS2-VASc score of ≥3 (P &lt;0.001). The CHA2DS2-VASc score was positively correlated with white blood cells count (r=0.257, P =0.018) and negatively correlated with the number of days spent in the hospital (r=-0.184, P=0.130) due to higher in-hospital death in ICU patients with COVID-19.&#x0D; Conclusion: CHA2DS2-VASc score may be an effective tool to estimate in-hospital death in COVID-19 patients who were hospitalized in the ICU.

Development of a Nomogram Model for Predicting the Risk of In-Hospital Death in Patients with Acute Kidney Injury.

ObjectiveTo analyze the risk factors of in-hospital death in patients with acute kidney injury (AKI) in the intensive care unit (ICU), and to develop a personalized risk prediction model.MethodsThe clinical data of 137 AKI patients hospitalized in the ICU of Anhui provincial hospital from January 2018 to December 2020 were retrospectively analyzed. Patients were divided into two groups: those that survived to discharge (“survival” group, 100 cases) and those that died while in hospital (“death” group, 37 cases), and risk factors for in-hospital death analyzed.ResultsThe in-hospital mortality of AKI patients in the ICU was 27.01% (37/137). A multivariate logistic regression analysis indicated age, mechanical ventilation and vasoactive drugs were significant risk factors for in-hospital death in AKI patients, and a nomogram risk prediction model was developed. The Harrell’s C-index of the nomogram model was 0.891 (95% CI: 0.837–0.945), and the area under the receiver operating characteristic (ROC) curve was 0.886 (95% CI: 0.823–0.936) after internal validation, indicating that the nomogram model had good discrimination. The Hosmer–Lemeshow goodness of fit test and calibration curve indicated the predicted probability of the nomogram model was consistent with the actual frequency of death in ICU patients with AKI. The decision curve analysis (DCA) showed that the clinical net benefit level of the nomogram model is highest when the probability threshold of AKI is between 0.01 and 0.75.ConclusionPatients in the ICU with AKI had high in-hospital mortality and were affected by a variety of risk factors. The nomogram prediction model based on the risk factors of AKI showed good prediction efficiency and clinical applicability, which could help medical staff in the ICU to identify AKI patients with high-risk, allowing early prevention, detection and intervention, and reducing the risk of in-hospital deaths in ICU patients with AKI.

Association Between the Central Venous Pressure and All-Cause Mortality in Critically Ill Patients with Acute Kidney Injury.

PurposeElevated central venous pressure (CVP) plays an important role in the occurrence of acute kidney injury (AKI) and it is also independently associated with the prognosis of critically ill patients. However, the effect of CVP on critically ill AKI patients remains unclear. In this study, we analyzed the relationship between CVP and all-cause mortality of critically ill patients with AKI.Patients and MethodsThe clinical data of patients in intensive care unit (ICU) were retrieved from the Medical Information Mart for Intensive Care III (MIMIC-III) database and retrospectively analyzed. The all-cause mortality for up to 90 days was the main observed outcome. We used the minimum CVP value obtained during the first 72 hours after ICU admission for our analysis and patients were grouped according to this parameter. Patients were also analyzed after being further divided according to stages 1, 2 and 3 of AKI. Multiple Cox regression and Kaplan–Meier analyses were used to explore the association between CVP measurements and death of ICU patients with AKI.ResultsA total of 1986 ICU patients with AKI were studied. A total of 527 (26.5%) patients died by day 90. The high CVP group (patients with ≥10 mmHg) had the lowest 90-day survival rate (P =0.001 by log rank test) when according to Kaplan–Meier analysis. By using Cox regression analysis, high CVP was found to be linked to an increase in mortality (CVP ≥10 mmHg versus ≤5 mmHg, HR, 1.336, 95% CI, 1.064 to 1.677, P trend=0.014). Furthermore, when using in a multivariate Cox regression analysis with CVP as a continuous variable, the higher CVP levels were still an independent risk factor for 90-day all-cause mortality (per 1 mmHg increase, HR, 1.031, 95% CI, 1.013–1.049, P=0.001). In subgroup analysis, a similar trend was observed in patients with AKI stages 2 and 3.ConclusionThe minimum CVP level during the first 72h after ICU admission was positively associated with mortality in critically ill patients with AKI and this more marked in cases with severe AKI.

Predictive Value of Eosinophil Count on COVID-19 Disease Progression and Outcomes, a Retrospective Study of Leishenshan Hospital in Wuhan, China.

BackgroundThe potential protective role of eosinophils in the COVID-19 pandemic has aroused great interest, given their potential virus clearance function and the infection resistance of asthma patients to this coronavirus. However, it is unknown whether eosinophil counts could serve as a predictor of the severity of COVID-19.MethodsA total of 1004 patients with confirmed COVID-19 who were admitted to Leishenshan Hospital in Wuhan, China, were enrolled in this study, including 905 patients in the general ward and 99 patients in the intensive care unit (ICU). We reviewed their medical data to analyze the association between eosinophils and ICU admission and death.ResultsOf our 1004 patients with COVID-19, low eosinophil counts/ratios were observed in severe cases. After adjusting for confounders that could have affected the outcome, we found that eosinophil counts might not be a predictor of ICU admission. In 99 ICU patients, 58 of whom survived and 41 of whom died, low eosinophil level was an indicator of death in severe COVID-19 patients with a cutoff value of 0.04 × 109/L, which had an area under the curve of 0.665 (95% CI = 1.089-17.839; P = .045) with sensitivity and specificity of 0.569 and 0.7317, respectively.ConclusionOur research revealed that a low eosinophil level is a predictor of death in ICU patients rather than a cause of ICU admission.

Prognostic Assessment of COVID-19 in the Intensive Care Unit by Machine Learning Methods: Model Development and Validation.

BackgroundPatients with COVID-19 in the intensive care unit (ICU) have a high mortality rate, and methods to assess patients’ prognosis early and administer precise treatment are of great significance.ObjectiveThe aim of this study was to use machine learning to construct a model for the analysis of risk factors and prediction of mortality among ICU patients with COVID-19.MethodsIn this study, 123 patients with COVID-19 in the ICU of Vulcan Hill Hospital were retrospectively selected from the database, and the data were randomly divided into a training data set (n=98) and test data set (n=25) with a 4:1 ratio. Significance tests, correlation analysis, and factor analysis were used to screen 100 potential risk factors individually. Conventional logistic regression methods and four machine learning algorithms were used to construct the risk prediction model for the prognosis of patients with COVID-19 in the ICU. The performance of these machine learning models was measured by the area under the receiver operating characteristic curve (AUC). Interpretation and evaluation of the risk prediction model were performed using calibration curves, SHapley Additive exPlanations (SHAP), Local Interpretable Model-Agnostic Explanations (LIME), etc, to ensure its stability and reliability. The outcome was based on the ICU deaths recorded from the database.ResultsLayer-by-layer screening of 100 potential risk factors finally revealed 8 important risk factors that were included in the risk prediction model: lymphocyte percentage, prothrombin time, lactate dehydrogenase, total bilirubin, eosinophil percentage, creatinine, neutrophil percentage, and albumin level. Finally, an eXtreme Gradient Boosting (XGBoost) model established with the 8 important risk factors showed the best recognition ability in the training set of 5-fold cross validation (AUC=0.86) and the verification queue (AUC=0.92). The calibration curve showed that the risk predicted by the model was in good agreement with the actual risk. In addition, using the SHAP and LIME algorithms, feature interpretation and sample prediction interpretation algorithms of the XGBoost black box model were implemented. Additionally, the model was translated into a web-based risk calculator that is freely available for public usage.ConclusionsThe 8-factor XGBoost model predicts risk of death in ICU patients with COVID-19 well; it initially demonstrates stability and can be used effectively to predict COVID-19 prognosis in ICU patients.