Chen-Yu Wang

Background/Objectives: The European Society for Clinical Nutrition and Metabolism recommends avoiding full feeding during the first 48–72 h of critical illness and gradually achieving 70–100% of energy requirements within the first week. However, the clinical significance of achieving approximately 70% of estimated energy requirements by day 7 in the intensive care unit (ICU) remains unclear in routine practice. This study investigated whether achieving ≥70% of estimated energy requirements by day 7 was associated with in-hospital mortality and explored clinical factors associated with achieving this target. Methods: This retrospective study included critically ill patients who remained in the ICU through day 7 and had complete nutritional data for the first 7 ICU days. Cox proportional hazards regression was performed in a day-7 landmark cohort to investigate the association between day-7 energy adequacy and subsequent in-hospital mortality. Logistic regression analysis was performed to identify the factors associated with achieving ≥70% of estimated energy requirements by day 7. Results: Among 507 patients in the day-7 landmark cohort, 355 achieved ≥70% of estimated energy requirements by ICU day 7, and 152 did not. Achievement of the day-7 target was associated with lower in-hospital mortality (adjusted hazard ratio [aHR] 0.50, 95% CI 0.30–0.84; p = 0.008), whereas achievement of the same target by day 3 was not significantly associated with in-hospital mortality. Older age and elevated serum albumin levels were independently associated with achieving the day-7 target. In exploratory subgroup analyses, the point estimates were directionally similar across clinically relevant subgroups, but no statistically significant interaction was observed. Conclusions: Achievement of ≥70% of estimated energy requirements by ICU day 7 was associated with lower in-hospital mortality, whereas achievement of the same target by day 3 was not significantly associated with in-hospital mortality. These findings should be interpreted as observational and hypothesis-generating.

Higher caloric intake may reduce hospital mortality in critically ill patients at high nutritional risk, but the optimal dose for short-term outcomes remains uncertain and evidence on long-term effects is limited. This study evaluated the association between caloric intake and one-year mortality and identified subgroups that may benefit from higher intake. We conducted a retrospective cohort study in a tertiary medical ICU (2015–2019) including adults receiving mechanical ventilation; ICU stays <48 h were excluded. The exposure was mean caloric intake during ICU days 1–7, defined as total energy from enteral and parenteral routes normalized to body weight. Outcomes included ICU length of stay, ventilator days, and one-year mortality from the Taiwan National Health Insurance Database. Multivariable Cox models adjusted for age, sex, albumin, hemoglobin, blood glucose, ICU admission etiology, APACHE II score, shock category, mNUTRIC score, renal replacement therapy, and cumulative day-1–7 fluid balance. Among 3,764 patients (mean age 67.1 years; mean Acute Physiology and Chronic Health Evaluation II score 26.5), older age, male sex, lower albumin and hemoglobin, shock requiring multiple vasopressors, greater positive fluid balance, and lower caloric intake were associated with higher one-year mortality. Subgroup analyses showed that patients younger than 65 years, those with an APACHE II score ≥26, and those with refractory shock derived greater benefit from higher caloric intake. Higher first-week caloric intake was associated with lower one-year mortality, particularly in younger patients, in those with greater illness severity, and in those requiring multiple vasopressors.

Critically ill patients require individualized nutrition support, with assessment tools like Nutrition Risk Screening 2002 and Nutrition Risk in the Critically Ill scores. Challenges in continu-ous nutrition care prompt the need for innovative solutions. This study develops an artificial intelligence assisted nutrition risk evaluation model using explainable machine learning to support intensive care unit dietitians. Ethical approval was obtained for a retrospective analysis of 2,122 pa-tients. Nutrition risk assessment involved six dietitians, with 1,994 patients assessed comprehensively. Artificial intelligence models and shapley additive explanations analysis were used to predict and understand nutrition risk. High nutrition risk (35.2%) correlated with elder age, lower body weight, BMI, albumin, and higher disease severity. The AUROC scores achieved by XGBoost (0.921), CatBoost (0.926), and LightGBM (0.923) were superior to those of Logistic Regression. Key features influencing nutrition risk included Acute Physiology and Chronic Health Evaluation II score, albumin, age, BMI, and haemoglobin. The study introduces an artificial intelligence assisted nutrition risk evaluation model, offering a promising avenue for continuous and timely nutrition support in critically ill patients. External validation and exploration of feature relationships are needed.

Background: Long-term ventilator-dependent patients often face problems such as decreased quality of life, increased mortality, and increased medical costs. Respiratory therapists must perform complex and time-consuming ventilator weaning assessments, which typically take 48-72 hours. Traditional disengagement methods rely on manual evaluation and are susceptible to subjectivity, human errors, and low efficiency. Objective: This study aims to develop an artificial intelligence–based prediction model to predict whether a patient can successfully pass a spontaneous breathing trial (SBT) using the patient’s clinical data collected before SBT initiation. Instead of comparing different SBT strategies or analyzing their impact on extubation success, this study focused on establishing a data-driven approach under a fixed SBT strategy to provide an objective and efficient assessment tool. Through this model, we aim to enhance the accuracy and efficiency of ventilator weaning assessments, reduce unnecessary SBT attempts, optimize intensive care unit resource usage, and ultimately improve the quality of care for ventilator-dependent patients. Methods: This study used a retrospective cohort study and developed a novel deep learning architecture, hybrid CNN-MLP (convolutional neural network–multilayer perceptron), for analysis. Unlike the traditional CNN-MLP classification method, hybrid CNN-MLP performs feature learning and fusion by interleaving CNN and MLP layers so that data features can be extracted and integrated at different levels, thereby improving the flexibility and prediction accuracy of the model. The study participants were patients aged 20 years or older hospitalized in the intensive care unit of a medical center in central Taiwan between January 1, 2016, and December 31, 2022. A total of 3686 patients were included in the study, and 6536 pre-SBT clinical records were collected before each SBT of these patients, of which 3268 passed the SBT and 3268 failed. Results: The model performed well in predicting SBT outcomes. The training dataset’s precision is 99.3% (2443/2460 records), recall is 93.5% (2443/2614 records), specificity is 99.3% (2597/2614 records), and F1-score is 0.963. In the test dataset, the model maintains accuracy with a precision of 89.2% (561/629 records), a recall of 85.8% (561/654 records), a specificity of 89.6% (586/654 records), and an F1-score of 0.875. These results confirm the reliability of the model and its potential for clinical application. Conclusions: This study successfully developed a deep learning–based SBT prediction model that can be used as an objective and efficient ventilator weaning assessment tool. The model's performance shows that it can be integrated into clinical workflow, improve the quality of patient care, and reduce ventilator dependence, which is an important step in improving the effectiveness of respiratory therapy.

Background: Candidemia is associated with markedly high intensive care unit (ICU) mortality rates. Although the Impact of Early Enteral vs. Parenteral Nutrition on Mortality in Patients Requiring Mechanical Ventilation and Catecholamines (NUTRIREA-2) trial indicated that early enteral nutrition (EN) did not reduce 28-day mortality rates among critically ill patients with shock, the European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines recommend avoiding EN in cases of uncontrolled shock. Whether increased caloric intake from EN positively impacts clinical outcomes in patients with candidemia and shock remains unclear.Methods: We retrospectively collected data from a tertiary medical center between January 2015 and December 2018. We enrolled patients who developed shock within the first 7 days following ICU admission and received a diagnosis of candidemia during their ICU stay. Patients with an ICU stay shorter than 48 hours were excluded.Results: The study included 106 patients, among whom the hospital mortality rate was 77.4% (82 patients). The median age of the patients was 71 years, and the median Acute Physiology and Chronic Health Evaluation II score was 29. The Cox regression model revealed that a higher 7-day average caloric intake through EN (hazard ratio, 0.61; 95% CI, 0.44–0.83) was significantly associated with lower hospital mortality rates. Our findings suggest EN as the preferred feeding route for critically ill patients with shock.Conclusions: Increased caloric intake through EN may be associated with lower hospital mortality rates in patients with candidemia and shock.

BACKGROUND Long-term ventilator-dependent patients often face problems such as decreased quality of life, increased mortality, and increased medical costs. Respiratory therapists must perform complex and time-consuming ventilator weaning assessments, which typically take 48-72 hours. Traditional disengagement methods rely on manual evaluation and are susceptible to subjectivity, human errors, and low efficiency. OBJECTIVE This study aims to develop an artificial intelligence–based prediction model to predict whether a patient can successfully pass a spontaneous breathing trial (SBT) using the patient’s clinical data collected before SBT initiation. Instead of comparing different SBT strategies or analyzing their impact on extubation success, this study focused on establishing a data-driven approach under a fixed SBT strategy to provide an objective and efficient assessment tool. Through this model, we aim to enhance the accuracy and efficiency of ventilator weaning assessments, reduce unnecessary SBT attempts, optimize intensive care unit resource usage, and ultimately improve the quality of care for ventilator-dependent patients. METHODS This study used a retrospective cohort study and developed a novel deep learning architecture, hybrid CNN-MLP (convolutional neural network–multilayer perceptron), for analysis. Unlike the traditional CNN-MLP classification method, hybrid CNN-MLP performs feature learning and fusion by interleaving CNN and MLP layers so that data features can be extracted and integrated at different levels, thereby improving the flexibility and prediction accuracy of the model. The study participants were patients aged 20 years or older hospitalized in the intensive care unit of a medical center in central Taiwan between January 1, 2016, and December 31, 2022. A total of 3686 patients were included in the study, and 6536 pre-SBT clinical records were collected before each SBT of these patients, of which 3268 passed the SBT and 3268 failed. RESULTS The model performed well in predicting SBT outcomes. The training dataset’s precision is 99.3% (2443/2460 records), recall is 93.5% (2443/2614 records), specificity is 99.3% (2597/2614 records), and F1-score is 0.963. In the test dataset, the model maintains accuracy with a precision of 89.2% (561/629 records), a recall of 85.8% (561/654 records), a specificity of 89.6% (586/654 records), and an F1-score of 0.875. These results confirm the reliability of the model and its potential for clinical application. CONCLUSIONS This study successfully developed a deep learning–based SBT prediction model that can be used as an objective and efficient ventilator weaning assessment tool. The model's performance shows that it can be integrated into clinical workflow, improve the quality of patient care, and reduce ventilator dependence, which is an important step in improving the effectiveness of respiratory therapy. CLINICALTRIAL

(1) Background: This study aimed to investigate the associations between the Global Initiative for Chronic Obstructive Lung Disease (GOLD) staging systems, medical costs, and mortality among patients with chronic obstructive lung disease (COPD). Predictions of the effectiveness of the two versions of the staging systems were also compared. (2) Purpose: this study investigated the associations between the Global Initiative for Chronic Obstructive Lung Disease (GOLD) staging systems, medical costs, and mortality among patients with COPD. Predicting effectiveness between the two versions of the staging systems was also compared. (3) Procedure: This study used a secondary clinical database of a medical center in central Taiwan to examine records between 2011 and 2017. A total of 613 patients with COPD were identified. The independent variables comprised the COPD GOLD Guideline staging of the 2007 and 2011 versions, demographic characteristics, health status, and physician seniority. The dependent variables included total medical cost, average length of hospital stay, and mortality. The statistical methods included binomial logistic regression and the general linear model (GLM). (4) Discussion: The total medical cost during the observation period for patients with COPD averaged TWD 292,455.6. The average length of hospital stay was 9.7 days. The mortality rate was 9.6%, compared with that of patients in Grade 1 of the 2007 version; patients in Grade 4 of the 2007 version had significantly higher odds of death (OR = 4.07, p = 0.02). The accuracy of mortality prediction for both the 2007 and 2011 versions of the staging was equal, at 90.4%. The adjusted GLM analysis revealed that patients in Group D of the 2011 version had a significantly longer length of hospital stay than those in Group A of the 2011 version (p = 0.04). No difference between the 2007 and 2011 versions was found regarding the total medical cost. Complications were significantly associated with the total medical cost and average length of hospital stay. (5) Conclusions: The COPD staging 2011 version was associated with an average length of hospital stay, whereas the COPD staging 2007 version was related to mortality risk. Therefore, the 2011 version can estimate the length of hospital stay. However, in predicting prognosis and mortality, the 2007 version is recommended.

Few large-scale studies have focused on tracheostomy in patients with prolonged mechanical ventilation. This retrospective population-based study extracted data from the longitudinal National Health Insurance Research Database in Taiwan to compare long-term mortality between patients on prolonged mechanical ventilation with and without tracheostomy and their related medical expenditures. Data on newly developed respiratory failure in patients on ventilator support were extracted from 1 January 2002 to 31 December 2008. Of 10,705 patients included, 1372 underwent tracheostomy (n = 563) or translaryngeal intubation (n = 779). Overall survival of the patients with tracheostomy was followed for 5 years. Average survival was 4.98 years for the patients with tracheostomy and 5.48 years for the patients with translaryngeal intubation (not significant). Sex, age, premium-based monthly salary difference, occupation, urbanization level, chronic obstructive pulmonary disease, chronic heart failure, chronic renal disease, and cerebrovascular diseases were significantly associated with mortality for endotracheal intubation. Male sex, chronic heart failure, chronic renal disease, age ≥45 years, and low income were associated with significantly higher mortality. Although total medical expenditures were higher for the patients with tracheostomy, annual medical expenditures were not significantly different. There were no differences in long-term mortality between the two groups.

Positioning patients in the prone position leads to reduced hospital mortality rates for those with severe acute respiratory distress syndrome (ARDS). What constitutes the optimal feeding strategy for prone patients with ARDS is controversial. We conducted a retrospective study that enrolled 110 prone patients with ARDS in two medical intensive care units (ICUs) from September 2015 to November 2018. Inclusion criteria were as follows: age ≥20 years, diagnosis of respiratory failure requiring mechanical ventilation, diagnosis of ARDS within 72 h of ICU admission, placement in a prone position within the first 7 days of ICU admission, and ICU stay of more than 7 days. Exclusion criteria were as follows: nil per os orders because of gastrointestinal bleeding or hemodynamic instability, and ventilator dependency because of chronic respiratory failure. The consecutive daily enteral nutrition(EN)/EN + parenteral nutrition(PN) ratio could predict hospital mortality rates within the first 7 days of admission when using generalized estimating equations (p = 0.013). A higher average EN/EN + PN ratio within the first 7 days predicted (hazard ratio: 0.97, confidence interval: 0.96–0.99) lower hospital mortality rates. To reduce hospital mortality rates, caloric intake with a higher EN ratio may be considered for patients in prone positions with ARDS.