Predicting hypoglycemia in critically Ill patients using machine learning and electronic health records.

Hypoglycemia is a common occurrence in critically ill patients and is associated with significant mortality and morbidity. We developed a machine learning model to predict hypoglycemia by using a multicenter intensive care unit (ICU) electronic health record dataset. Machine learning algorithms were trained and tested on patient data from the publicly available eICU Collaborative Research Database. Forty-four features including patient demographics, laboratory test results, medications, and vitals sign recordings were considered. The outcome of interest was the occurrence of a hypoglycemic event (blood glucose < 72 mg/dL) during a patient's ICU stay. Machine learning models used data prior to the second hour of the ICU stay to predict hypoglycemic outcome. Data from 61,575 patients who underwent 82,479 admissions at 199 hospitals were considered in the study. The best-performing predictive model was the eXtreme gradient boosting model (XGBoost), which achieved an area under the received operating curve (AUROC) of 0.85, a sensitivity of 0.76, and a specificity of 0.76. The machine learning model developed has strong discrimination and calibration for the prediction of hypoglycemia in ICU patients. Prospective trials of these models are required to evaluate their clinical utility in averting hypoglycemia within critically ill patient populations.

Evidence-Based Management of the Critically Ill Adult With SARS-CoV-2 Infection.

Human infection by the novel viral pathogen SARS-CoV-2 results in a clinical syndrome termed Coronavirus Disease 2019 (COVID-19). Although the majority of COVID-19 cases are self-limiting, a substantial minority of patients develop disease severe enough to require intensive care. Features of critical illness associated with COVID-19 include hypoxemic respiratory failure, acute respiratory distress syndrome (ARDS), shock, and multiple organ dysfunction syndrome (MODS). In most (but not all) respects critically ill patients with COVID-19 resemble critically ill patients with ARDS due to other causes and are optimally managed with standard, evidence-based critical care protocols. However, there is naturally an intense interest in developing specific therapies for severe COVID-19. Here we synthesize the rapidly expanding literature around the pathophysiology, clinical presentation, and management of COVID-19 with a focus on those points most relevant for intensivists tasked with caring for these patients. We specifically highlight evidence-based approaches that we believe should guide the identification, triage, respiratory support, and general ICU care of critically ill patients infected with SARS-CoV-2. In addition, in light of the pressing need and growing enthusiasm for targeted COVID-19 therapies, we review the biological basis, plausibility, and clinical evidence underlying these novel treatment approaches.

A novel multi-modal platform to image molecular and elemental alterations in ischemic stroke.

Stroke is a major global health problem, with the prevalence and economic burden predicted to increase due to aging populations in western society. Following stroke, numerous biochemical alterations occur and damage can spread to nearby tissue. This zone of "at risk" tissue is termed the peri-infarct zone (PIZ). As the PIZ contains tissue not initially damaged by the stroke, it is considered by many as salvageable tissue. For this reason, much research effort has been undertaken to improve the identification of the PIZ and to elucidate the biochemical mechanisms that drive tissue damage in the PIZ in the hope of identify new therapeutic targets. Despite this effort, few therapies have evolved, attributed in part, to an incomplete understanding of the biochemical mechanisms driving tissue damage in the PIZ. Magnetic resonance imaging (MRI) has long been the gold standard to study alterations in gross brain structure, and is frequently used to study the PIZ following stroke. Unfortunately, MRI does not have sufficient spatial resolution to study individual cells within the brain, and reveals little information on the biochemical mechanisms driving tissue damage. MRI results may be complemented with histology or immuno-histochemistry to provide information at the cellular or sub-cellular level, but are limited to studying biochemical markers that can be successfully "tagged" with a stain or antigen. However, many important biochemical markers cannot be studied with traditional MRI or histology/histochemical methods. Therefore, we have developed and applied a multi-modal imaging platform to reveal elemental and molecular alterations that could not previously be imaged by other traditional methods. Our imaging platform incorporates a suite of spectroscopic imaging techniques; Fourier transform infrared imaging, Raman spectroscopic imaging, Coherent anti-stoke Raman spectroscopic imaging and X-ray fluorescence imaging. This approach does not preclude the use of traditional imaging techniques, and rather it should be use to complement traditional methods such as MRI or histology and immunohistochemistry, to gain a greater insight into disease mechanisms. We demonstrate the potential of this approach by characterizing biochemical alterations within the PIZ 24h after the induction of photothrombotic stroke in mice. Substantial molecular and elemental alterations were identified in the PIZ 24h after stroke that are consistent with tissue swelling and edema, but not oxidative stress. This reveals important mechanistic information, that could not previously be obtained, which should be considered in future studies aimed at developing therapeutic intervention from this model.

Preparing to take the USMLE Step 1: a survey on medical students' self-reported study habits.

BACKGROUND: The USA Medical Licensing Examination Step 1 is a computerised multiple-choice examination that tests the basic biomedical sciences. It is administered after the second year in a traditional four-year MD programme. Most Step 1 scores fall between 140 and 260, with a mean (SD) of 227 (22). Step 1 scores are an important selection criterion for residency choice. Little is known about which study habits are associated with a higher score. OBJECTIVE: To identify which self-reported study habits correlate with a higher Step 1 score. METHODS: A survey regarding Step 1 study habits was sent to third year medical students at Tulane University School of Medicine every year between 2009 and 2011. The survey was sent approximately 3 months after the examination. RESULTS: 256 out of 475 students (54%) responded. The mean (SD) Step 1 score was 229.5 (22.1). Students who estimated studying more than 8-11 h per day had higher scores (p<0.05), but there was no added benefit with additional study time. Those who reported studying <40 days achieved higher scores (p<0.05). Those who estimated completing >2000 practice questions also obtained higher scores (p<0.01). Students who reported studying in a group, spending the majority of study time on practice questions or taking >40 preparation days did not achieve higher scores. CONCLUSIONS: Certain self-reported study habits may correlate with a higher Step 1 score compared with others. Given the importance of achieving a high Step 1 score on residency choice, it is important to further identify which characteristics may lead to a higher score.

Sleep disorder syndromes of post-acute sequelae of SARS-CoV-2 (PASC) / Long Covid.

INTRODUCTION: COVID-19 infection has resulted in a high prevalence of a post-infectious syndrome, known as post-acute sequelae of SARS-CoV-2 (PASC) or "Long COVID". PASC is a heterogeneous disease with a high prevalence of sleep disturbances, varying from an insomnia disorder to excessive daytime sleepiness. METHODS: Patients seen in the Covid Survivorship Program at the Beth Israel Deaconess Medical Center Boston, USA, were screened for sleep disorders as part of a comprehensive multi-system evaluation. Those who screened positive were referred for a comprehensive sleep evaluation in a dedicated COVID-19-Sleep clinic, followed by diagnostic sleep testing and treatment. This report summarizes patients who completed an American Academy of Sleep Medicine (AASM) accredited facility-based diagnostic evaluation. International Classification of Sleep Disorders 3rd Edition-Revised criteria were met for all diagnoses. RESULTS: In 42 patients with PASC, five categories of sleep disorder syndromes were observed following a sleep clinic evaluation, including obstructive sleep apnea, chronic insomnia disorder, primary hypersomnia, REM behavior disorder (RBD), and new onset circadian phase delay. Seven patients met criteria for idiopathic hypersomnia, and two had narcolepsy type 2. RBD patients were infected in three different waves; circadian disturbance patients were all infected in the winter wave of 2020/21, and the primary hypersomnolence group occurred during all waves, predominantly the initial wave of 2020. A peculiar form of insomnia was a persistent loss of sleep regularity. CONCLUSIONS: Specific sleep symptoms/syndromes are reported in this select group of patients with PASC/Long Covid. As new onset sleep complaints are prevalent in PASC, we recommend a complete clinical and investigative sleep evaluation for persistent severe sleep symptoms following COVID-19 infection.

Resilience, Survival, and Functional Independence in Older Adults Facing Critical Illness.

BACKGROUND: Older adults surviving critical illness often experience new or worsening functional impairments. Modifiable positive psychological constructs such as resilience may mitigate post-intensive care morbidity. RESEARCH QUESTION: Is pre-ICU resilience associated with: (1) post-ICU survival; (2) the drop in functional independence during the ICU stay; or (3) the trend in predicted independence before vs after the ICU stay? STUDY DESIGN AND METHODS: This retrospective cohort study was performed by using Medicare-linked Health and Retirement Study surveys from 2006 to 2018. Older adults aged ≥ 65 years admitted to an ICU were included. Resilience was calculated prior to ICU admission. The resilience measure was defined from the Simplified Resilience Score, which was previously adapted and validated for the Health and Retirement Study. Resilience was scored by using the Leave-Behind survey normalized to a scale from 0 (lowest resilience) to 12 (highest resilience). Outcomes were survival and probability of functional independence. Survival was modeled by using Gompertz models and independence using joint survival models adjusting for sociodemographic and clinical variables. Average marginal effects were estimated to determine independence probabilities. RESULTS: Across 3,409 patients aged ≥ 65 years old admitted to ICUs, preexisting frailty (30.5%) and cognitive impairment (24.3%) were common. Most patients were previously independent (82.7%). Mechanical ventilation occurred in 14.8% and sepsis in 43.2%. Those in the highest resilience group (vs lowest resilience) had a lower risk of post-ICU mortality (adjusted hazard ratio, 0.81; 95% CI, 0.70-0.94). Higher resilience was associated with greater likelihood of post-ICU functional independence (estimated probability of functional independence 5 years after ICU discharge in highest-to-lowest resilience groups (adjusted hazard ratio [95% CI]): 0.53 (0.33-0.74), 0.47 (0.26-0.68), 0.49 (0.28-0.70), and 0.36 (0.17-0.55); P < .01. Resilience was not associated with a difference in the drop in independence during the ICU stay or a difference in the pre-ICU vs post-ICU trend in predicted independence . INTERPRETATION: ICU survivors with higher resilience had increased rates of survival and functional independence, although the slope of functional decline before vs after the ICU stay did not differ according to resilience group.

Nirmatrelvir-Ritonavir for Acute COVID-19 in Patients With Cardiovascular Disease and Postacute Sequelae of SARS-CoV-2 Infection.

Background: There is limited evidence of association of nirmatrelvir-ritonavir (NMV-r) and incidence of postacute sequelae of SARS-CoV-2 infection (PASC) in patients with pre-existing cardiovascular disease (CVD). Objectives: The objective of this study was to assess the association of NMV-r in nonhospitalized, vaccinated patients with pre-existing CVD and occurrence of PASC. Methods: We conducted a retrospective cohort study utilizing the TriNetX research network, including vaccinated patients with pre-existing CVD who developed COVID-19 between December 2021 and December 2022. Two cohorts were created based on NMV-r administration within 5 days of diagnosis: NMV-r and non-NMV-r cohort. The main outcome was presence of PASC, assessed between 30 to 90 days and 90 to 180 days after index COVID-19 infection. After propensity score matching, both cohorts were compared using t-test and chi-square test for continuous and categorical variables, respectively. Results: A total of 26,953 patients remained in each cohort after propensity score matching. Broadly defined PASC occurred in 6,925 patients (26%) in the NMV-r cohort vs 8,150 patients (30.6%) in the non-NMV-r cohort (OR: 0.80; 95% CI: 0.76-0.82; P < 0.001) from 30 to 90 days and in 6,692 patients (25.1%) as compared to 8,910 patients (33.5%) (OR: 0.25, 95% CI: 0.23-0.29; P < 0.001) from 90 to 180 days. Similarly, narrowly defined PASC occurred in 5,335 patients (20%) in the NMV-r cohort vs 6,271 patients (23.6%) in the non-NMV-r cohort between 30 and 90 days (OR: 0.81, 95% CI: 0.78-0.84, P < 0.001) and in 5,121 patients (19.2%) as compared to 6,964 patients (26.1%) (OR: 0.67, 95% CI: 0.64-0.70, P < 0.001) between 90 and 180 days. Conclusions: NMV-r in nonhospitalized vaccinated patients with pre-existing CVD with COVID-19 was associated with a reduction in PASC and health care utilization.

Spectroscopic understanding of ultra-high rate performance for LiMn(0.75)Fe(0.25)PO4 nanorods-graphene hybrid in lithium ion battery.

Comprehensive X-ray absorption near-edge structure spectroscopy at the C, O and Li K-edges and the Mn, Fe, and P L-edges of LiMn(0.75)Fe(0.25)PO(4) nanorods-graphene has been reported in great detail. Compared to that of free standing graphene and LiMn(0.75)Fe(0.25)PO(4), the intimate interaction between the nanorods and graphene via charge redistribution has been unambiguously confirmed. This interaction not only anchors the nanorods onto the graphene but also modifies its surface chemistry, both of which afford the nanorods-graphene hybrid an ultra-high rate performance in lithium ion batteries. Such knowledge is important for the understanding of hybrid nanomaterials for lithium ion batteries and allows rational design for further improvements in performance.

Beyond the ICU Rotation: The Importance of Trainee Involvement in Post-Intensive Care Unit Clinics.

Medical learners are vital to the care of critically ill patients in the intensive care unit (ICU). Although these learners are exposed to the challenges and stresses of acute ICU management, they do not typically experience the benefits of following ICU patients and families longitudinally after their ICU rotation. Post-ICU clinics and recovery programs may fill this crucial gap in trainee education. These clinics have emerged as an appealing approach to potentially support patient recovery, enhance provider satisfaction, and provide feedback on vital lessons learned in long-term follow-up to improve the quality of ICU care. Notably, the effect of such a program on trainee education has not been explored. In this article, we propose a framework for medical learner participation in post-ICU follow-up based on the Accreditation Council for Graduate Medical Education milestones and discuss the potential benefits, including: education about post-ICU recovery, including post-intensive care syndrome and post-intensive care syndrome-family; experience in quality improvement to enhance ICU care by understanding long-term outcomes; engagement in reflection; and mitigation of compassion fatigue and burnout.

Structure - Functionality of lentil protein-polyphenol conjugates.

Lentil protein isolate (LPI) was conjugated with plant polyphenols (quercetin, rutin, ellagic acid), and the structural and functional characteristics of the conjugates were determined in comparison with the proteins and pure polyphenols. The interaction between polyphenols and protein was achieved by a grafting method at pH 9.0 in the presence of atmospheric oxygen. Surface plasmon resonance measurements showed polyphenols' direct interaction with LPI, with the order of binding strength quercetin > ellagic acid > rutin. The degree of conjugation also followed the same order. Structural analysis of the conjugates was performed using FTIR, intrinsic fluorescence, and surface hydrophobicity. A significant improvement in DPPH radical scavenging and ferric reducing antioxidant power of the conjugates was observed compared to the polyphenols. However, there was a decrease in the surface activity of the conjugates compared to LPI. Such conjugation provides a novel way to combine the advantages of using plant protein and polyphenols in developing a novel food ingredient.

Surface chemical heterogeneous distribution in over-lithiated Li1+xCoO2 electrodes.

In commercial Li-ion batteries, the internal short circuits or over-lithiation often cause structural transformation in electrodes and may lead to safety risks. Herein, we investigate the over-discharged mechanism of LiCoO2/graphite pouch cells, especially spatially resolving the morphological, surface phase, and local electronic structure of LiCoO2 electrode. With synchrotron-based X-ray techniques and Raman mapping, together with spectroscopy simulations, we demonstrate that over-lithiation reaction is a surface effect, accompanied by Co reduction and surface structure transformation to Li2CoO2/Co3O4/CoO/Li2O-like phases. This surface chemical distribution variation is relevant to the depth and exposed crystalline planes of LiCoO2 particles, and the distribution of binder/conductive additives. Theoretical calculations confirm that Li2CoO2-phase has lower electronic/ionic conductivity than LiCoO2-phase, further revealing the critical effect of distribution of conductive additives on the surface chemical heterogeneity evolution. Our findings on such surface phenomena are non-trivial and highlight the capability of synchrotron-based X-ray techniques for studying the spatial chemical phase heterogeneity.