The effect of mechanical ventilation compared to high-flow nasal cannula on gastric residual volume and reflux events using novel automated technology.

BACKGROUND & AIM: High flow nasal cannula (HFNC) oxygen therapy is frequently used following extubation. A case report, utilizing an innovative medical technology (The smART + Platform, ART MEDICAL Ltd., Netanya, Israel) that enables the detection of gastric refluxes and gastric residual volumes (GRV), has suggested that HFNC may be associated with increased reflux events and GRV. This study measured reflux events and GRV using smART+ in mechanically ventilated patients before and after extubation while they were receiving HFNC therapy. We aim to show if there is a significant difference in reflux events and GRV between HFNC users and mechanically ventilated patients. METHODS: This is a post hoc analysis examines data of a randomized controlled trial (RCT) involving critically ill adult patients who received enteral nutrition through the smART + Platform. The study was approved by the local ethics committee. We compared the frequency and amplitude of reflux events and GRV in mechanically ventilated patients. These parameters were assessed both 3 h before extubation and subsequently after extubation when the patients were connected to HFNC. Patients served as their own controls. To evaluate the differences between the pre- and post-extubation measurements, we applied a parametric paired t-test. RESULTS: Ten patients (mean age of 58 years; mean APACHE II score 22; mean 3.9 days of mechanical ventilation) were included. Three hours prior extubation the mean GRV was 4.1 ml/h compared to 14.03 ml/h on HFNC (p = 0.004). The mean frequency of major reflux events was 2.33/h in ventilated patients versus 4.4/h in the HFNC patients (p = 0.73). The mean frequency of major reflux events was 9.17/h in ventilated patients versus 9.83/h in HFNC patients (p = 0.14). CONCLUSIONS: Leveraging the smART + Platform, we demonstrated that the use of HFNC significantly increases GRV compared with patients on mechanical ventilation and may increase the frequency of major reflux events, thereby increasing the risk of aspiration. Further studies are required to support our conclusions.

Urea to creatinine ratio as a predictor of persistent critical illness.

INTRODUCTION: Persistent critical illness (PCI) is a syndrome in which the acute presenting problem has been stabilized, but the patient's clinical state does not allow ICU discharge. The burden associated with PCI is substantial. The most obvious marker of PCI is prolonged ICU length of stay (LOS), usually greater than 10 days. Urea to Creatinine ratio (UCr) has been suggested as an early marker of PCI development. METHODS: A single-center retrospective study. Data of patients admitted to a general mixed medical-surgical ICU during Jan 1st 2018 till Dec 31st 2022 was extracted, including demographic data, baseline characteristics, daily urea and creatinine results, renal replacement therapy (RRT) provided, and outcome measures - length of stay, and mortality (ICU, and 90 days). Patients were defined as PCI patients if their LOS was >10 days. We used Fisher exact test or Chi-square to compare PCI and non-PCI patients. The association between UCr with PCI development was assessed by repeated measures linear model. Multivariate Cox regression was used for 1 year mortality assessment. RESULTS: 2098 patients were included in the analysis. Patients who suffered from PCI were older, with higher admission prognostic scores. Their 90-day mortality was significantly higher than non-PCI patients (34.58% vs 12.18%, p < 0.0001). A significant difference in UCr was found only on the first admission day among all patients. This was not found when examining separately surgical, trauma, or transplantation patients. We did not find a difference in UCr in different KDIGO (Kidney Disease Improving Global Outcomes) stages. Elevated UCr and PCI were found to be significantly associated with 1 year mortality. CONCLUSION: In this single center retrospective cohort study, UCr was not found to be associated with PCI development.

Predicting invasive mechanical ventilation in COVID 19 patients: A validation study.

INTRODUCTION: The decision to intubate and ventilate a patient is mainly clinical. Both delaying intubation (when needed) and unnecessarily invasively ventilating (when it can be avoided) are harmful. We recently developed an algorithm predicting respiratory failure and invasive mechanical ventilation in COVID-19 patients. This is an internal validation study of this model, which also suggests a categorized "time-weighted" model. METHODS: We used a dataset of COVID-19 patients who were admitted to Rabin Medical Center after the algorithm was developed. We evaluated model performance in predicting ventilation, regarding the actual endpoint of each patient. We further categorized each patient into one of four categories, based on the strength of the prediction of ventilation over time. We evaluated this categorized model performance regarding the actual endpoint of each patient. RESULTS: 881 patients were included in the study; 96 of them were ventilated. AUC of the original algorithm is 0.87-0.94. The AUC of the categorized model is 0.95. CONCLUSIONS: A minor degradation in the algorithm accuracy was noted in the internal validation, however, its accuracy remained high. The categorized model allows accurate prediction over time, with very high negative predictive value.

Controlled enteral nutrition in critical care patients - A randomized clinical trial of a novel management system.

PURPOSE: Nutritional therapy is essential to ICU care. Successful early enteral feeding is hindered by lack of protocols, gastrointestinal intolerance and feeding interruptions, leading to impaired nutritional intake. smART+ was developed as a nutrition management feeding platform controlling tube positioning, reflux, gastric pressure, and malnutrition. This study evaluated the potential of this new ICU care platform to deliver targeted nutrition and improve ICU outcomes. METHODS: Critically ill patients ≥18 years-old, mechanically ventilated and enterally fed, were randomized to receive ESPEN-guideline-based nutrition or smART+ -guided nutrition for 2-14 days. Primary endpoint was average deviation from daily targeted nutrition determined via calculation of energy targets per calorimetry. Secondary endpoints included gastric residual volumes, length of stay (LOS) and length of ventilation (LOV). RESULTS: smART+ achieved a mean deviation from daily targeted nutrition of 10.5% (n = 48) versus 34.3% for control (n = 50), p < 0.0001. LOS and LOV were decreased in the smART+ group versus control (mean LOS: 10.4 days versus 13.7; reduction 3.3 days, adjusted HR 1.71, 95% CI:1.13,2.60, p = 0.012; mean LOV: 9.5 days versus 12.8 days reduction of 3.3 days, adjusted HR 1.64, 95% CI:1.08-2.51, p = 0.021). Feeding goals were met (within ±10%) on 75.7% of days for smART+ versus 23.3% for control (p < 0.001). No treatment-related adverse events occurred in either group. The study was stopped due to success in a planned interim analysis of the first 100 patients. CONCLUSION: The smART+ Platform improved adherence to feeding goals and reduced LOS and LOV versus standard of care in critically ill patients. TRIAL REGISTRATION: NCT04098224; registered September 23, 2019.

Using Machine-Learning to Assess the Prognostic Value of Early Enteral Feeding Intolerance in Critically Ill Patients: A Retrospective Study.

BACKGROUND: The association between gastrointestinal intolerance during early enteral nutrition (EN) and adverse clinical outcomes in critically ill patients is controversial. We aimed to assess the prognostic value of enteral feeding intolerance (EFI) markers during early ICU stays and to predict early EN failure using a machine learning (ML) approach. METHODS: We performed a retrospective analysis of data from adult patients admitted to Beilinson Hospital ICU between January 2011 and December 2018 for more than 48 h and received EN. Clinical data, including demographics, severity scores, EFI markers, and medications, along with 72 h after admission, were analyzed by ML algorithms. Prediction performance was assessed by the area under the receiver operating characteristics (AUCROC) of a ten-fold cross-validation set. RESULTS: The datasets comprised 1584 patients. The means of the cross-validation AUCROCs for 90-day mortality and early EN failure were 0.73 (95% CI 0.71-0.75) and 0.71 (95% CI 0.67-0.74), respectively. Gastric residual volume above 250 mL on the second day was an important component of both prediction models. CONCLUSIONS: ML underlined the EFI markers that predict poor 90-day outcomes and early EN failure and supports early recognition of at-risk patients. Results have to be confirmed in further prospective and external validation studies.

Transient diabetes insipidus in critically ill COVID19 patients.

PURPOSE: Vasopressin has become an important vasopressor drug while treating a critically ill patient to maintain adequate mean arterial pressure. Diabetes insipidus (DI) is a rare syndrome characterized by the excretion of a large volume of diluted urine, inappropriate for water homeostasis. We noticed that several COVID19 patients developed excessive polyuria suggestive of DI, with a concomitant plasma sodium-level increase and/or low urine osmolality. We noticed a temporal relationship between vasopressin treatment cessation and polyuria periods. We reviewed those cases to better describe this phenomenon. METHODS: We retrospectively collected COVID19 ECMO patients' (from July 6, 2020, to November 30, 2021) data from the electronic medical records. By examining urine output, urine osmolality (if applicable), plasma sodium level, and plasma osmolality, we set DI diagnosis. We described the clinical course of DI episodes and compared baseline characteristics between patients who developed DI and those who did not. RESULTS: Out of 37 patients, 12 had 18 episodes of DI. These patients were 7 years younger and had lower severity scores (APACHE-II and SOFA). Mortality difference was not seen between groups. 17 episodes occurred after vasopressin discontinuation; 14 episodes were treated with vasopressin reinstitution. DI lasted for a median of 21 h, with a median increase of 14 mEq/L of sodium. CONCLUSIONS: Temporary DI prevalence after vasopressin discontinuation in COVID19 ECMO patients might be higher than previously described for vasopressin-treated patients.

Community optimized management for better eating after hospital sTay among geriatric patients of poor socio-economic status - The COMEAT study.

INTRODUCTION: In patients suffering from disease-related and socioeconomic malnutrition and being discharged from hospital, continuity of care is challenging. Lack of adequate nutrition may lead to increase in morbidity and mortality. The aim of this study was to overcome the handicap of limited nutrition access in this category of patients and to study its consequences on clinical outcome. METHODS: Hospitalized patients screened to be at risk of malnutrition were approached and if diagnosed as suffering from malnutrition and from limited financial resources, they were randomized to receive a delivered daily dinner tray for 6 months and an assistance during the meal by a philanthropic association, or to regular food. At entry to the study, patients were assessed by indirect calorimetry, bioimpedance, Hospital Anxiety and Depression Scale (HADS), Functional independence measure (FIM) and SF 36 quality of life questionnaire. The latest questionnaires were reproduced after 3 and 6 months. Survival was followed at 6 months. The student t-test, the paired t-test, ANOVA were used. 180 days survival curves were expressed using the Kaplan-Meier method. RESULTS: 631 patients were screened and 60 patients were randomized. There was no difference between groups. Survival at 6 months was improved significantly in the intervention group (87%) compared to the control group (65%, p<005). HADS did significantly improve at 3 months and other parameters (FIM, SF 36) were not changed significantly. CONCLUSIONS: In hospitalized patients at nutritional risk, lunch home delivery, supported by a physical company after hospital discharge was associated with significantly lower mortality rates and improved depression and anxiety scores in elderly patients suffering from socioeconomic related malnutrition.

A novel machine learning model to predict respiratory failure and invasive mechanical ventilation in critically ill patients suffering from COVID-19.

In hypoxemic patients at risk for developing respiratory failure, the decision to initiate invasive mechanical ventilation (IMV) may be extremely difficult, even more so among patients suffering from COVID-19. Delayed recognition of respiratory failure may translate into poor outcomes, emphasizing the need for stronger predictive models for IMV necessity. We developed a two-step model; the first step was to train a machine learning predictive model on a large dataset of non-COVID-19 critically ill hypoxemic patients from the United States (MIMIC-III). The second step was to apply transfer learning and adapt the model to a smaller COVID-19 cohort. An XGBoost algorithm was trained on data from the MIMIC-III database to predict if a patient would require IMV within the next 6, 12, 18 or 24 h. Patients' datasets were used to construct the model as time series of dynamic measurements and laboratory results obtained during the previous 6 h with additional static variables, applying a sliding time-window once every hour. We validated the adaptation algorithm on a cohort of 1061 COVID-19 patients from a single center in Israel, of whom 160 later deteriorated and required IMV. The new XGBoost model for the prediction of the IMV onset was trained and tested on MIMIC-III data and proved to be predictive, with an AUC of 0.83 on a shortened set of features, excluding the clinician's settings, and an AUC of 0.91 when the clinician settings were included. Applying these models "as is" (no adaptation applied) on the dataset of COVID-19 patients degraded the prediction results to AUCs of 0.78 and 0.80, without and with the clinician's settings, respectively. Applying the adaptation on the COVID-19 dataset increased the prediction power to an AUC of 0.94 and 0.97, respectively. Good AUC results get worse with low overall precision. We show that precision of the prediction increased as prediction probability was higher. Our model was successfully trained on a specific dataset, and after adaptation it showed promise in predicting outcome on a completely different dataset. This two-step model successfully predicted the need for invasive mechanical ventilation 6, 12, 18 or 24 h in advance in both general ICU population and COVID-19 patients. Using the prediction probability as an indicator of the precision carries the potential to aid the decision-making process in patients with hypoxemic respiratory failure despite the low overall precision.

Effect of Combined Protein-Enriched Enteral Nutrition and Early Cycle Ergometry in Mechanically Ventilated Critically Ill Patients-A Pilot Study.

BACKGROUND: Cycle ergometry (CE) is a method of exercise used in clinical practice. Limited data demonstrate its effectiveness in critically ill patients. We aimed to evaluate the combination of CE and a high-protein diet in critically ill patients. METHODS: This was an open label pilot trial comparing conventional physiotherapy with enteral nutrition (EN) (control, Group 1), CE with EN (Group 2), and CE with protein-enriched EN (Group 3). The primary outcome was length of ventilation (LOV). Secondary outcomes were intensive care unit (ICU) mortality, length of ICU stay (ICU LOS), length of hospital stay (Hospital LOS), and rate of re-intubation. RESULTS: Per protocol, 41 ICU patients were enrolled. Thirteen patients were randomized to Group 1 (control), fourteen patients to Group 2, and fourteen patients to Group 3 (study groups). We found no statistically significant difference in LOV between the study arms (14.2 ± 9.6 days, 15.8 ± 7.1 days, and 14.9 ± 9.4 days, respectively, p = 0.89). Secondary outcomes did not demonstrate any significant differences between arms. CONCLUSIONS: In this pilot trial, CE combined with either standard EN or protein-enriched EN was not associated with better clinical outcomes, as compared to conventional physiotherapy with standard EN. Larger trials are needed in order to further evaluate this combination.

Correlations between First 72 h Hypophosphatemia, Energy Deficit, Length of Ventilation, and Mortality-A Retrospective Cohort Study.

INTRODUCTION: Hypophosphatemia may prolong ventilation and induce weaning failure. Some studies have associated hypophosphatemia with increased mortality. Starting or restarting nutrition in a critically ill patient may be associated with refeeding syndrome and hypophosphatemia. The correlation between nutrition, mechanical ventilation, and hypophosphatemia has not yet been fully elucidated. METHODS: A retrospective cohort study of 825 admissions during two consecutive years was conducted. Using the electronic medical chart, demographic and clinical data were obtained. Hypophosphatemia was defined as a phosphate level below 2.5 mg/dL (0.81 mmol/L) in the first 72 h of ICU admission. Comparisons between baseline characteristics and outcomes and multivariate analysis were performed. RESULTS: A total of 324 (39.27%) patients had hypophosphatemia during the first 72 h of ICU admission. Patients with hypophosphatemia tended to be younger, with lower APACHE-II, SOFA24, and ΔSOFA scores. They had a longer length of stay and length of ventilation, more prevalent prolonged ventilation, and decreased mortality. Their energy deficit was lower. There was no effect of hypophosphatemia severity on these results. In multivariate analysis, hypophosphatemia was not found to be statistically significant either with respect to mortality or survivor's length of ventilation, but lower average daily energy deficit and SOFA24 were found to be statistically significant with respect to survivor's length of ventilation. CONCLUSION: Hypophosphatemia had no effect on mortality or length of ventilation. Lower average daily energy deficit is associated with a longer survivor's length of ventilation.

The effect of indirect calorimetry guided isocaloric nutrition on mortality in critically ill patients-a systematic review and meta-analysis.

Indirect calorimetry (IC)-guided nutrition might positively affect the clinical outcome of critically ill patients. In this systematic review and meta-analysis, our objective was to assess the benefit of isocaloric nutrition guided by IC, compared to hypocaloric nutrition, for critically ill patients admitted to the intensive care unit (ICU). We performed a systematic review of all randomized controlled trials published through January 2021, assessing the benefit of isocaloric nutrition guided by IC. The primary outcome was 28-day all-cause mortality. Secondary outcomes were ICU and 90-day all-cause mortality, rate of nosocomial infections, and adverse events. Four trials evaluating 1052 patients were included. Patients treated with isocaloric nutrition had a lower 28-day mortality rate (risk ratio (RR) 0.79, 95% confidence interval (CI) 0.63-0.99, P = 0.04). No between-group difference was found in ICU and 90-day mortality (RR 0.92, 95% CI 0.68-1.23, P = 0.56 and RR 0.88, 95% CI 0.72-1.07; P = 0.2, respectively) and in the rate of nosocomial infections (RR 1.15, 95% CI 0.77-1.72, P = 0.51). A pooled analysis of studies that evaluated the benefit of isocaloric nutrition guided by IC, for critically ill patients in the ICU, has shown reduced 28-day mortality. However, there was no difference in 90-day mortality and nosocomial infection rate.

Feasibility of achieving different protein targets using a hypocaloric high-protein enteral formula in critically ill patients.

BACKGROUND AND AIMS: Combining energy and protein targets during the acute phase of critical illness is challenging. Energy should be provided progressively to reach targets while avoiding overfeeding and ensuring sufficient protein provision. This prospective observational study evaluated the feasibility of achieving protein targets guided by 24-h urinary nitrogen excretion while avoiding overfeeding when administering a high protein-to-energy ratio enteral nutrition (EN) formula. METHODS: Critically ill adult mechanically ventilated patients with an APACHE II score > 15, SOFA > 4 and without gastrointestinal dysfunction received EN with hypocaloric content for 7 days. Protein need was determined by 24-h urinary nitrogen excretion, up to 1.2 g/kg (Group A, N = 10) or up to 1.5 g/kg (Group B, N = 22). Variables assessed included nitrogen intake, excretion, balance; resting energy expenditure (REE); phase angle (PhA); gastrointestinal tolerance of EN. RESULTS: Demographic characteristics of groups were similar. Protein target was achieved using urinary nitrogen excretion measurements. Nitrogen balance worsened in Group A but improved in Group B. Daily protein and calorie intake and balance were significantly increased in Group B compared to Group A. REE was correlated to PhA measurements. Gastric tolerance of EN was good. CONCLUSIONS: Achieving the protein target using urinary nitrogen loss up to 1.5 g/kg/day was feasible in this hypercatabolic population. Reaching a higher protein and calorie target did not induce higher nitrogen excretion and was associated with improved nitrogen balance and a better energy intake without overfeeding. PhA appears to be related to REE and may reflect metabolism level, suggestive of a new phenotype for nutritional status. Trial registration 0795-18-RMC.

Enteral and supplemental parenteral nutrition enriched with omega-3 polyunsaturated fatty acids in intensive care patients - A randomized, controlled, double-blind clinical trial.

BACKGROUND & AIMS: Enteral nutrition (EN) and parenteral nutrition (PN) enriched with omega-3 polyunsaturated fatty acids (PUFA) have beneficial effects in critical illness. This study aimed to assess the combined effect of EN and supplemental PN enriched with omega-3 PUFA on blood oxygenation in intensive care unit (ICU) patients. METHODS: Single-center, prospective, randomized, controlled, double-blind, phase III trial conducted from 10/2013 to 11/2017. A total of 100 ICU patients (18-85 years, APACHE II score > 15) requiring mechanical ventilation were randomly assigned to received combined EN and PN either with omega-3 PUFA (omega-3 group) or without (control group) for up to 28 days. Primary endpoint: 'change of PaO2/FiO2 from day (D) 1 to D4'. Secondary endpoints: lung function parameters, ICU complications, length of hospital stay, days free of ICU care/ventilation/sedation/catecholamine treatment, mortality, erythrocyte fatty acid composition, inflammatory parameters. Safety parameters: standard laboratory assessment, vital signs, physical examination, SOFA score, adverse events. RESULTS: Combined EN and PN covered energy requirements to more than 80%. Blood oxygenation (ΔPaO2/FiO2 from D1 to D4: -1.3 ± 83.7, n = 42, and 13.3 ± 86.1, n = 39, in omega-3 and control group, respectively, p = 0.7795) and other lung function parameters did not differ between groups but days free of catecholamine treatment were significantly higher in the omega-3 group (~4 days, p = 0.0481). On D6, significantly more patients in the omega-3 group tolerated EN alone (51.0% vs. 29.8%, p = 0.0342). Eicosapentaenoic acid (EPA) content in erythrocytes was significantly increased in the omega-3 group at last observation compared with the control group (ΔEPA: 0.928 ± 0.808% vs. -0.024 ± 0.190%, p < 0.0001). No further significant group differences were detected. CONCLUSIONS: Enteral and supplemental PN both enriched with omega-3 PUFA did not improve lung function but allowed earlier weaning from catecholamine treatment and PN. Supplemental PN succeeded to adequately cover energy requirements in critically ill patients. TRIAL REGISTRATION: www.clinicaltrials.gov, registration number: NCT01162928.

Validation of GLIM malnutrition criteria for diagnosis of malnutrition in ICU patients: An observational study.

BACKGROUND & AIMS: Patients in the Intensive Care Unit (ICU) are at high risk of malnutrition. The only validated malnutrition assessment tool is the Subjective Global Assessment (SGA). The Global Leadership Initiative on Malnutrition (GLIM) is a new malnutrition assessment tool. The present study compares the nutrition-related parameters of the following tools: GLIM tool, SGA, Phase Angle (PA), Low Fat-Free Mass Index (FFMI), and Patient- and Nutrition-Derived Outcome Risk Assessment score (PANDORA), in an attempt to validate an objective tool. METHODS: Eighty-four ICU patients were included. The tools mentioned above were assessed for their validity in diagnosing malnutrition. All patients were defined as suffering from acute disease and received medical nutrition therapy. To evaluate whether there is a correlation between the GLIM criteria, SGA, PA, and low FFMI, we compared the SGA, PA, and low FFMI to the GLIM criteria using Spearman correlation coefficients and a Chi-square test. Also, a Mann-Whitney U test was used to test the mean differences between the GLIM criteria and the PANDORA. The area under the curve (AUC) of the proposed parameters was evaluated for diagnosis of malnutrition to seek cutoff points that yield good sensitivity and specificity. RESULTS: Mean age was 50 ± 20 years, BMI 25.3 ± 5.1 kg/m2, APACHE II 20.5 ± 7.7, PANDORA score 32 ± 8.5. GLIM malnutrition criteria were significantly correlated with the gold standard SGA assessment and with low FFMI, with PA (Phase Angle), and with the PANDORA score. The area under the curve, by using the ROC curve analysis for GLIM criteria stratified by the SGA results, was 0.85 (P < 0.001). Sensitivity was 85%, and specificity 79%. However, when comparing the low FFMI, PA, and PANDORA to the GLIM criteria, the ROC curve analysis results were considered poor rank. CONCLUSIONS: The SGA malnutrition assessment highly validated the GLIM criteria framework combined with the two-criteria diagnosis of malnutrition with a high level of precision. The GLIM malnutrition assessment seems to be acceptable in the ICU setting.

The centenary of the Harris-Benedict equations: How to assess energy requirements best? Recommendations from the ESPEN expert group.

BACKGROUND & AIMS: The year 2019 marked the centenary of the publication of the Harris and Benedict equations for estimation of energy expenditure. In October 2019 a Scientific Symposium was organized by the European Society for Clinical Nutrition and Metabolism (ESPEN) in Vienna, Austria, to celebrate this historical landmark, looking at what is currently known about the estimation and measurement of energy expenditure. METHODS: Current evidence was discussed during the symposium, including the scientific basis and clinical knowledge, and is summarized here to assist with the estimation and measurement of energy requirements that later translate into energy prescription. RESULTS: In most clinical settings, the majority of predictive equations have low to moderate performance, with the best generally reaching an accuracy of no more than 70%, and often lead to large errors in estimating the true needs of patients. Generally speaking, the addition of body composition measurements did not add to the accuracy of predictive equations. Indirect calorimetry is the most reliable method to measure energy expenditure and guide energy prescription, but carries inherent limitations, greatly restricting its use in real life clinical practice. CONCLUSIONS: While the limitations of predictive equations are clear, their use is still the mainstay in clinical practice. It is imperative to recognize specific patient populations for whom a specific equation should be preferred. When available, the use of indirect calorimetry is advised in a variety of clinical settings, aiming to avoid under-as well as overfeeding.

Improved ICU mortality prediction based on SOFA scores and gastrointestinal parameters.

BACKGROUND: The Sequential Organ Failure Assessment (SOFA) score is commonly used in ICUs around the world, designed to assess the severity of the patient's clinical state based on function/dysfunction of six major organ systems. The goal of this work is to build a computational model to predict mortality based on a series of SOFA scores. In addition, we examined the possibility of improving the prediction by incorporating a new component designed to measure the performance of the gastrointestinal system, added to the other six components. METHODS: In this retrospective study, we used patients' three latest SOFA scores recorded during an individual ICU stay as input to different machine learning models and ensemble learning models. We added three validated parameters representing gastrointestinal failure. Among others, we used classification models such as Support Vector Machines (SVMs), Neural Networks, Logistic Regression and a penalty function used to increase model robustness in regard to certain extreme cases, which may be found in ICU population. We used the Area under Curve (AUC) performance metric to examine performance. RESULTS: We found an ensemble model of linear and logistic regression achieves a higher AUC compared related works in past years. After incorporating the gastrointestinal failure score along with the penalty function, our best performing ensemble model resulted in an additional improvement in terms of AUC metrics. We implemented and compared 36 different models that were built using both the information from the SOFA score as well as that of the gastrointestinal system. All compared models have approximately similar and relatively large AUC (between 0.8645 and 0.9146) with the best results are achieved by incorporating the gastrointestinal parameters into the prediction models. CONCLUSIONS: Our findings indicate that gastrointestinal parameters carry significant information as a mortality predictor in addition to the conventional SOFA score. This information improves the predictive power of machine learning models by extending the SOFA to include information related to gastrointestinal organ system. The described method improves mortality prediction by considering the dynamics of the extended SOFA score. Although tested on a limited data set, the results' stability across different models suggests robustness in real-time use.

Early Administration of Protein in Critically Ill Patients: A Retrospective Cohort Study.

It is currently uncertain whether early administration of protein improves patient outcomes. We examined mortality rates of critically ill patients receiving early compared to late protein administration. This was a retrospective cohort study of mixed ICU patients receiving enteral or parenteral nutritional support. Patients receiving >0.7 g/kg/d protein within the first 3 days were considered the early protein group and those receiving less were considered the late protein group. The latter were subdivided into late-low group (LL) who received a low protein intake (<0.7 g/kg/d) throughout their stay and the late-high group (LH) who received higher doses (>0.7 g/kg/d) of protein following their first 3 days of admission. The outcome measure was all-cause mortality 60 days after admission. Of the 2253 patients included in the study, 371 (36%) in the early group, and 517 (43%) in the late-high group had died (p < 0.001 for difference). In multivariable Cox regression analysis, while controlling for confounders, early protein administration was associated with increased survival (HR 0.83, 95% CI 0.71⁻0.97, p = 0.017). Administration of protein early in the course of critical illness appears to be associated with improved survival in a mixed ICU population, even after adjusting for confounding variables.

Predictive equations versus measured energy expenditure by indirect calorimetry: A retrospective validation.

BACKGROUND & AIMS: Measuring resting energy expenditure (REE) via indirect calorimetry (IC) in intensive care unit (ICU) patient is the gold standard recommended by guidelines. However technical difficulties hinder its use and predictive equations are largely used instead. We sought to validate commonly used equations using a large cohort of patients. METHODS: Patients hospitalized from 2003 to 2015 in a 16-bed ICU at a university-affiliated, tertiary care hospital who had IC measurement to assess caloric targets were included. Data was drawn from a computerized system and included REE and other variables required by equations. Measurements were restricted to 5 REE per patient to avoid bias. Equation performance was assessed by comparing means, standard deviations, correlation, concordance and agreement, which was defined as a measurement within 85-115% of measured REE. A total of 8 equations were examined. RESULTS: A total of 3573 REE measurements in 1440 patients were included. Mean patient age was 58 years and 65% were male. A total of 562 (39%) patients had >2 REE measurements. Standard deviation of REE ranged from 430 to 570 kcal. The Faisy equation had the least mean difference (90 Kcal); Harris-Benedict had the highest correlation (52%) and agreement (50%) and Jolliet the highest concordance (62%). Agreement within 10% of caloric needs was met only in a third of patients. CONCLUSIONS: Predictive equations have low performance when compared to REE in ICU patients. We therefore suggest that predictive equations cannot wholly replace indirect calorimetry for the accurate estimation of REE in this population.

Prevention of Exposure Keratopathy in Critically Ill Patients: A Single-Center, Randomized, Pilot Trial Comparing Ocular Lubrication With Bandage Contact Lenses and Punctal Plugs.

OBJECTIVES: To compare the effectiveness of bandage contact lenses and punctal plugs with ocular lubricants in preventing corneal damage in mechanically ventilated and sedated critically ill patients. DESIGN: Single-center, prospective, randomized, pilot study. SETTING: Sixteen-bed, general ICU at a tertiary academic medical center. PATIENTS: Adults admitted to the ICU and anticipated to require mechanical ventilation and continuous sedation for greater than or equal to 4 days. INTERVENTIONS: Patients were randomized to receive eye care with ocular lubricants (n = 38), bandage contact lenses (n = 33), or punctal plugs (n = 33). The bandage contact lenses were changed every 4 days, whereas the punctal plugs remained in situ for the entire study. MEASUREMENTS AND MAIN RESULTS: The primary endpoint was the presence or absence of corneal damage as assessed by the grade of keratopathy. Patients were examined by an ophthalmologist blinded to the study group every 4 days and at the time of withdrawal from the study, due to cessation of sedation, discharge from the ICU, or death. The mean duration of the study was 8.6 ± 6.2 days. The grade of keratopathy in the ocular lubricant group increased significantly in both eyes (p = 0.01 for both eyes) while no worsening was noted in either the lens or punctal plugs groups. In a post hoc analysis of patients with an initially abnormal ophthalmic examination, significant healing of keratopathy was noted in the lens group (p = 0.02 and 0.018 for left and right eyes, respectively) and in the right eye of the plugs group (p = 0.005); no improvement was noted in the ocular lubricant group. CONCLUSIONS: Compared with ocular lubrication, bandage contact lenses and punctal plugs were more effective in limiting keratopathy, and their use, particularly of bandage contact lenses, was associated with significant healing of existing lesions.