Introduction This study aimed to develop and validate an interpretable machine-learning model based on clinical features for early predicting in-hospital mortality in critically ill patients with sepsis. Methods We enrolled all patients with sepsis in the Medical Information Mart for Intensive Care IV (MIMIC-IV, v.1.0) database from 2008 to 2019. Lasso regression was used for feature selection. Seven machine-learning methods were applied to develop the models. The best model was selected based on its accuracy and area under curve (AUC) in the validation cohort. Furthermore, we employed the SHapley Additive exPlanations (SHAP) method to illustrate the effects of the features attributed to the model, and to analyze how the individual features affect the output of the model, and to visualize the Shapley value for a single individual. Results In total, 8,817 patients with sepsis were eligible for participation, the median age was 66.8 years (IQR, 55.9–77.1 years), and 3361 of 8817 participants (38.1%) were women. After selection, 25 of a total 57 clinical parameters collected on day 1 after ICU admission remained associated with prognosis and were used for developing the machine-learning models. Among seven constructed models, the eXtreme Gradient Boosting (XGBoost) model achieved the best performance with an AUC of 0.884 and an accuracy of 89.5% in the validation cohort. Feature importance analysis showed that Glasgow Coma Scale (GCS) score, blood urea nitrogen, respiratory rate, urine output, and age were the top 5 features of the XGBoost model with the greatest impact. Furthermore, SHAP force analysis illustrated how the constructed model visualized the individualized prediction of death. Conclusions We have demonstrated the potential of machine-learning approaches for predicting outcome early in patients with sepsis. The SHAP method could improve the interpretability of machine-learning models and help clinicians better understand the reasoning behind the outcome. Graphical abstract

Background: Malnutrition poses a significant public health challenge, particularly as the global population ages. However, there is a notable lack of comprehensive literature analyzing the global burden of malnutrition among the elderly.Methods: Data on nutritional deficiencies indicators—prevalence, incidence, and disability-adjusted life years (DALYs)—for individuals aged 70 and older were extracted from the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) database from 1990 to 2021. The estimated annual percentage change (EAPC) was used to assess trends in malnutrition burden. Predictions for new cases over the next 25 years were also made.Results: In 2021, over 97.60 million cases of malnutrition among the elderly were reported globally, a 1.2-fold increase from 44.36 million cases in 1990. The global prevalence rate of malnutrition has decreased with an EAPC of −0.32%. Women experienced more cases than men but had a lower prevalence rate. Furthermore, in high socio-demographic index (SDI) regions, both prevalence rate and DALYs rates increased from 1990 to 2021, with EAPCs of 0.33% and 1.34%, respectively. The burden of malnutrition in the elderly was negatively correlated with SDI. Predictions from 2022 to 2046 estimated approximately 29.64 million new cases of malnutrition by 2046, despite a declining incidence rate.Conclusion: Despite a decline in malnutrition prevalence rate and DALYs rate since 1990, the burden remains high, particularly in high SDI regions where rates have increased. With a projected rise in new cases, effective prevention and management strategies are urgently needed to support the health of older adults.

Background: Diaphragmatic ultrasound has been increasingly used to evaluate diaphragm function. However, current diaphragmatic ultrasound parameters provide indirect estimates of diaphragmatic contractile function, and the predictive value is controversial. Two-dimensional (2D) speckle tracking is an effective technology for measuring tissue deformation and can be used to measure diaphragm longitudinal strain (DLS) to assess diaphragm function. The purpose of this study was to determine the feasibility and reproducibility of DLS quantification by 2D speckle tracking and to determine whether maximal DLS could be used to predict weaning outcomes. Methods: This study was performed in the intensive care unit of two teaching hospitals, and was divided into two studies. Study A was a prospective study to evaluate the feasibility, reliability, and repeatability of speckle tracking in assessing DLS in healthy subjects and mechanically ventilated patients. Study B was a multicentre retrospective study to assess the use of maximal DLS measured by speckle tracking in predicting weaning outcomes. Results: Twenty-five healthy subjects and twenty mechanically ventilated patients were enrolled in Study A. Diaphragmatic speckle tracking was easily accessible. The intra- and interoperator reliability were good to excellent under conditions of eupnoea, deep breathing, and mechanical ventilation. The intraclass correlation coefficient (ICC) ranged from 0.78 to 0.95. Ninety-six patients (fifty-nine patients were successfully weaned) were included in Study B. DLS exhibited a fair linear relationship with both the diaphragmatic thickening fraction (DTF) (R² = 0.73, p < 0.0001) and diaphragmatic excursion (DE) (R² = 0.61, p < 0.0001). For the prediction of successful weaning, the areas under the ROC curves of DLS, diaphragmatic thickening fraction DTF, RSBI, and DE were 0.794, 0.794, 0.723, and 0.728, respectively. The best cut-off value for predicting the weaning success of DLS was less than -21%, which had the highest sensitivity of 89.19% and specificity of 64.41%. Conclusions: Diaphragmatic strain quantification using speckle tracking is easy to obtain in healthy subjects and mechanically ventilated patients and has a high predictive value for mechanical weaning. However, this method offers no advantage over RSBI. Future research should assess its value as a predictor of weaning. Trial registration: This study was registered in the Chinese Clinical Trial Register (ChiCTR), ChiCTR2100049816. Registered 10 August 2021. http://www.chictr.org.cn/showproj.aspx?proj=131790

Introduction The association between thiamine use and clinical outcomes among patients with sepsis and alcohol use disorder (AUD) is unclear. Methods In this retrospective cohort study of patients from Medical Information Mart for Intensive Care III (MIMIC-III, version 1.4), we evaluated the association of thiamine use with clinical outcomes in patients with AUD and sepsis. The primary outcome was 28-day survival, and secondary outcomes included ICU, in-hospital, and 90-day mortality, ICU and hospital length of stay, duration of vasopressor use, need and duration of continuous renal replacement therapy (CRRT), and dynamic changes for variables up to day 7 after ICU admission. Results A total of 944 patients with sepsis and AUD were included in this cohort [median age, 53.1 years; women, 26.0% (245 of 944)]. Among all patients, 24.6% (233 of 944) received thiamine with a dose of 200 mg (IQR 100–345 mg). The 28-day mortality was 11.2% (26 of 233) in the thiamine use group compared with 18.6% (132 of 711) in the no thiamine use group (P = 0.009). After adjustment for a series of confounders, the mixed-effects Cox proportional hazards models showed that administration of thiamine was associated with a lower risk of 28-day mortality compared with no administration of thiamine. Conclusions In critically ill patients with alcohol use disorder admitted for sepsis, treatment with thiamine may be associated with a decreased risk of death. However, the present results should be interpreted with caution due to the limitations of retrospective design. Additional larger, multicenter randomized controlled trials are needed to confirm our findings.

Background: Sepsis-associated AKI is a life-threatening condition associated with profound metabolic dysfunction, including impaired mitochondrial fatty acid handling. Therapies capable of reversing this metabolic dysfunction remain lacking. Using a clinically relevant porcine model, we investigated whether kidney-targeted fibroblastic reticular cell-derived and CD5L-enriched exosomes could ameliorate sepsis-associated AKI. Methods: We generated engineered exosomes derived from fibroblastic reticular cells or CD5L-overexpressing cells and functionalized them with the kidney-targeting LTH peptide. Therapeutic efficacy and safety were evaluated in a porcine model of lipopolysaccharide-induced sepsis-associated AKI. Integrated multiomic and cellular analyses were performed to examine the associated molecular pathways. Results: In septic pigs, both fibroblastic reticular cell-derived and CD5L-enriched exosomes improved systemic hemodynamics, renal microvascular perfusion, and kidney injury indices. Exosome treatment reduced renal lipid-droplet and acylcarnitine accumulation. CD5L interacted with fatty acid synthase, and treatment was associated with changes in short-chain fatty acids and AMPK-ACC-CPT1A signaling. These lipid-metabolic changes were accompanied by improved mitochondrial bioenergetic indices, preserved mitochondrial morphology, and reduced markers of tubular cell pyroptosis. Conclusions: Kidney-targeted fibroblastic reticular cell-derived and CD5L-enriched exosomes produced comparable improvements in kidney injury indices and were associated with renal lipid-metabolic and mitochondrial remodeling in a large-animal model of sepsis-associated AKI.