The Value of Combining Carbon Dioxide Gap and Oxygen-Derived Variables with Lactate Clearance in Predicting Mortality after Resuscitation of Septic Shock Patients

Background. Achieving hemodynamic stabilization does not prevent progressive tissue hypoperfusion and organ dysfunction during resuscitation of septic shock patients. Many indicators have been proposed to judge the optimization of oxygen delivery to meet tissue oxygen consumption. Methods. A prospective observational study was conducted to evaluate and validate combining CO2 gap and oxygen-derived variables with lactate clearance during early hours of resuscitation of adults presenting with septic shock. Results. Our study included 456 adults with a mean age of 63.2 ± 6.9 years, with 71.9% being males. Respiratory and urinary infections were the origin of about 75% of sepsis. Mortality occurred in 164 (35.9%) patients. The APACHE II score was 18.2 ± 3.7 versus 34.3 ± 6.8 ( p < 0.001 ), the initial SOFA score was 5.8 ± 3.1 versus 7.3 ± 1.4 ( p = 0.001 ), while the SOFA score after 48 hours was 4.2 ± 1.8 versus 9.4 ± 3.1 ( p < 0.001 ) in the survivors and nonsurvivors, respectively. Hospital mortality was independently predicted by hyperlactatemia (OR: 2.47; 95% CI: 1.63–6.82, p = 0.004 ), PvaCO2 gap (OR: 2.62; 95% CI: 1.28–6.74, p = 0.026 ), PvaCO2/CavO2 ratio (OR: 2.16; 95% CI: 1.49–5.74, p = 0.006 ), and increased SOFA score after 48 hours of admission (OR: 1.86; 95% CI: 1.36–8.13, p = 0.02 ). A blood lactate cutoff of 40 mg/dl at the 6th hour of resuscitation (T6) had a 92.7% sensitivity and 75.3% specificity for predicting hospital mortality (AUROC = 0.902) with 81.6% accuracy. Combining the lactate cutoff of 40 mg/dl and PvaCO2/CavO2 ratio cutoff of 1.4 increased the specificity to 93.2% with a sensitivity of 75.6% in predicting mortality and with 86.8% accuracy. Combining the lactate cutoff of 40 mg/dl and PvaCO2 gap of 6 mmHg increased the sensitivity to 93% and increased the specificity to 98% in predicting mortality with 91% accuracy. Conclusion. Combining the carbon dioxide gap and arteriovenous oxygen difference with lactate clearance during early hours of resuscitation of septic shock patients helps to predict hospital mortality more accurately.

Machine-learning-based COVID-19 mortality prediction model and identification of patients at low and high risk of dying

Background The coronavirus disease 2019 (COVID-19) pandemic caused by the SARS-Cov2 virus has become the greatest health and controversial issue for worldwide nations. It is associated with different clinical manifestations and a high mortality rate. Predicting mortality and identifying outcome predictors are crucial for COVID patients who are critically ill. Multivariate and machine learning methods may be used for developing prediction models and reduce the complexity of clinical phenotypes. Methods Multivariate predictive analysis was applied to 108 out of 250 clinical features, comorbidities, and blood markers captured at the admission time from a hospitalized cohort of patients (N = 250) with COVID-19. Inspired modification of partial least square (SIMPLS)-based model was developed to predict hospital mortality. Prediction accuracy was randomly assigned to training and validation sets. Predictive partition analysis was performed to obtain cutting value for either continuous or categorical variables. Latent class analysis (LCA) was carried to cluster the patients with COVID-19 to identify low- and high-risk patients. Principal component analysis and LCA were used to find a subgroup of survivors that tends to die. Results SIMPLS-based model was able to predict hospital mortality in patients with COVID-19 with moderate predictive power (Q2 = 0.24) and high accuracy (AUC > 0.85) through separating non-survivors from survivors developed using training and validation sets. This model was obtained by the 18 clinical and comorbidities predictors and 3 blood biochemical markers. Coronary artery disease, diabetes, Altered Mental Status, age > 65, and dementia were the topmost differentiating mortality predictors. CRP, prothrombin, and lactate were the most differentiating biochemical markers in the mortality prediction model. Clustering analysis identified high- and low-risk patients among COVID-19 survivors. Conclusions An accurate COVID-19 mortality prediction model among hospitalized patients based on the clinical features and comorbidities may play a beneficial role in the clinical setting to better management of patients with COVID-19. The current study revealed the application of machine-learning-based approaches to predict hospital mortality in patients with COVID-19 and identification of most important predictors from clinical, comorbidities and blood biochemical variables as well as recognizing high- and low-risk COVID-19 survivors.

Deep Learning Methods to Predict Mortality in COVID-19 Patients: A Rapid Scoping Review

The ongoing COVID-19 pandemic has become the most impactful pandemic of the past century. The SARS-CoV-2 virus has spread rapidly across the globe affecting and straining global health systems. More than 2 million people have died from COVID-19 (as of 30 January 2021). To lessen the pandemic’s impact, advanced methods such as Artificial Intelligence models are proposed to predict mortality, morbidity, disease severity, and other outcomes and sequelae. We performed a rapid scoping literature review to identify the deep learning techniques that have been applied to predict hospital mortality in COVID-19 patients. Our review findings provide insights on the important deep learning models, data types, and features that have been reported in the literature. These summary findings will help scientists build reliable and accurate models for better intervention strategies for predicting mortality in current and future pandemic situations.

External validation of Leipzig-Halifax scores for aortic dissection in Armenia

Background Few prognostic tools are currently available to predict hospital mortality in patients with acute type A aortic dissection. The aim of this study was to validate the performance of two existing risk-assessment tools, the original and the adjusted Leipzig-Halifax scorecards, to predict hospital mortality among Armenian patients with acute type A aortic dissection. Methods This retrospective cohort study included all consecutive patients with acute type A aortic dissection who were admitted to two tertiary cardiac centers in Armenia and underwent surgery from January 2008 to April 2018. We evaluated the predictive power of the original and adjusted Leipzig-Halifax scorecards using logistic regression analysis. Results Overall, 211 patients (76% males, mean age 57 ± 9 years) were included in the study, of whom 37 (17.5%) died during hospitalization. The adjusted Leipzig-Halifax score, but not the original Leipzig-Halifax score, was a significant predictor of hospital mortality. Patients with medium and high adjusted Leipzig-Halifax scores had a significantly higher odds of death compared to patients with low scores (odds ratio = 3.0 vs. 3.9, 95% confidence interval: 1.3–6.9 vs. 1.0–14.9, respectively). The areas under the receiver operating characteristic curves were 0.58 and 0.63, respectively, p > 0.05. Conclusion The adjusted Leipzig-Halifax score performed slightly better than the original Leipzig-Halifax score in the Armenian acute type A aortic dissection population. The adjusted Leipzig-Halifax score should now be applied prospectively to generate more data for further validation and potential improvement.

SOFAMONIA: Comparison of the original SOFA score with the proposed new score including serum ammonia

Introduction: Hyperammonemia can represent organic dysfunction of the brain, kidney, or liver. Evaluation of serum ammonia concentrations as a parameter for organ dysfunction may be justified. Objective: To evaluate the performance of serum ammonia as an additional or substitute variable for organ systems in the Sequential Organ Failure Assessment (SOFA) score. Methods: A prospective cohort study including 173 patients admitted to the intensive care unit between March 2015 and February 2016. SOFAMONIA scores were defined as follows: SOFAMONIA1 (Glasgow coma scale replaced by serum ammonia), SOFAMONIA2 (serum bilirubin replaced by serum ammonia), SOFAMONIA3 (renal system score replaced by serum ammonia), and SOFAMONIA4 (addition of serum ammonia to the original SOFA as the seventh variable, changing the maximum score from 24 to 28). Results: The original SOFA presented an area under the curve–receiver operating characteristic of 0.697 to predict hospital mortality. There was a positive correlation between SOFA and SOFAMONIA scores. SOFAMONIA1 presented a cut-off point of 5 for area under the curve 0.684 (0.610–0.753, 95% confidence interval), SOFAMONIA2 presented a cut-off point of 9 for area under the curve 0.701 (0.626–0.768, 95% confidence interval), SOFAMONIA3 presented a cut-off point of 8 for area under the curve 0.674 (0.598–0.743, 95% confidence interval), and SOFAMONIA4 presented a cut-off point of 8 for area under the curve 0.702 (0.628–0.769, 95% confidence interval). Conclusions: The addition of ammonia as the seventh parameter of the SOFA score showed the best performance to predict hospital mortality. The addition of ammonia as a representative of metabolic dysfunction may be useful in the follow-up of critically ill patients.