More judicious use of cephalosporins, especially 3rd-generation cephalosporins, may decrease ESBL-producing E. coli or K. pneumoniae bacteremia, and also improve patient outcome.

No abstract

No abstract

Background Heparin is one of the most commonly used medications in intensive care units. In clinical practice, the use of a weight-based heparin dosing nomogram is standard practice for the treatment of thrombosis. Recently, machine learning techniques have dramatically improved the ability of computers to provide clinical decision support and have allowed for the possibility of computer generated, algorithm-based heparin dosing recommendations. Objective The objective of this study was to predict the effects of heparin treatment using machine learning methods to optimize heparin dosing in intensive care units based on the predictions. Patient state predictions were based upon activated partial thromboplastin time in 3 different ranges: subtherapeutic, normal therapeutic, and supratherapeutic, respectively. Methods Retrospective data from 2 intensive care unit research databases (Multiparameter Intelligent Monitoring in Intensive Care III, MIMIC-III; e–Intensive Care Unit Collaborative Research Database, eICU) were used for the analysis. Candidate machine learning models (random forest, support vector machine, adaptive boosting, extreme gradient boosting, and shallow neural network) were compared in 3 patient groups to evaluate the classification performance for predicting the subtherapeutic, normal therapeutic, and supratherapeutic patient states. The model results were evaluated using precision, recall, F1 score, and accuracy. Results Data from the MIMIC-III database (n=2789 patients) and from the eICU database (n=575 patients) were used. In 3-class classification, the shallow neural network algorithm performed the best (F1 scores of 87.26%, 85.98%, and 87.55% for data set 1, 2, and 3, respectively). The shallow neural network algorithm achieved the highest F1 scores within the patient therapeutic state groups: subtherapeutic (data set 1: 79.35%; data set 2: 83.67%; data set 3: 83.33%), normal therapeutic (data set 1: 93.15%; data set 2: 87.76%; data set 3: 84.62%), and supratherapeutic (data set 1: 88.00%; data set 2: 86.54%; data set 3: 95.45%) therapeutic ranges, respectively. Conclusions The most appropriate model for predicting the effects of heparin treatment was found by comparing multiple machine learning models and can be used to further guide optimal heparin dosing. Using multicenter intensive care unit data, our study demonstrates the feasibility of predicting the outcomes of heparin treatment using data-driven methods, and thus, how machine learning–based models can be used to optimize and personalize heparin dosing to improve patient safety. Manual analysis and validation suggested that the model outperformed standard practice heparin treatment dosing.

IntroductionThe peripheral perfusion index (PI) is a noninvasive numerical value of peripheral perfusion, and the transcutaneous oxygen challenge test (OCT) is defined as the degree of transcutaneous partial pressure of oxygen (PtcO2) response to 1.0 FiO2. The value of noninvasive monitoring peripheral perfusion to predict outcome remains to be established in septic patients after resuscitation. Moreover, the prognostic value of PI has not been investigated in septic patients.MethodsForty-six septic patients, who were receiving PiCCO-Plus cardiac output monitoring, were included in the study group. Twenty stable postoperative patients were studied as a control group. All the patients inspired 1.0 of FiO2 for 10 minutes during the OCT. Global hemodynamic variables, traditional metabolic variables, PI and OCT related-variables were measured simultaneously at 24 hours after PiCCO catheter insertion. We obtained the 10min-OCT ((PtcO2 after 10 minutes on inspired 1.0 oxygen) - (baseline PtcO2)), and the oxygen challenge index ((10min-OCT)/(PaO2 on inspired 1.0 oxygen - baseline PaO2)) during the OCT.ResultsThe PI was significantly correlated with baseline PtcO2, 10min-OCT and oxygen challenge index (OCI) in all the patients. The control group had a higher baseline PtcO2, 10min-OCT and PI than the septic shock group. In the sepsis group, the macro hemodynamic parameters and ScvO2 showed no differences between survivors and nonsurvivors. The nonsurvivors had a significantly lower PI, 10min-OCT and OCI, and higher arterial lactate level. The PI, 10min-OCT and OCI predicted the ICU mortality with an accuracy that was similar to arterial lactate level. A PI <0.2 and a 10min-OCT <66mmHg were related to poor outcome after resuscitation.ConclusionsThe PI and OCT are predictive of mortality for septic patients after resuscitation. Further investigations are required to determine whether the correction of an impaired level of peripheral perfusion may improve the outcome of septic shock patients.

The pathogenesis of septic myocardial depression is complicated. Mitochondrial dysfunction has been suggested to be one of the main reasons for the reduced cardiac function. As melatonin is an antioxidant with the potential to scavenge radicals in mitochondria, we therefore employed a sepsis model, that is, cecal ligation and double puncture (CLP) in rats, to study the melatonin effects on: (i), myocardial mitochondrial function; (ii), heart systolic function; and (iii), prognosis of septic rats. We demonstrate that melatonin treatment (30 mg/kg, 3, 6, 12, 18, 24 hr after CLP) (i) improved myocardial cytochrome c oxidase (CcOX) activity and blood lactate level, (ii) attenuated heart dysfunction with a higher left ventricular ejection fraction (EF), and (iii) promoted 48-h survival of the rats compared to CLP animals with no melatonin treatment. In conclusion, our results show that rat myocardial mitochondrial CcOX activity was depressed during severe sepsis accompanied by myocardial depression characterized by the decline of EF. In septic rats, melatonin increased the CcOX activity, improved heart systolic function, and lowered mortality rate. The clinical use of melatonin in septic myocardial depression should be tested in the future.

The aim of this study is to determine the mechanism of sepsis-induced vascular hyperpermeability and the beneficial effect of glucocorticoid in protecting vascular endothelium. Male Sprague-Dawley rats were given either a bolus intraperitoneal injection of a nonlethal dose of LPS (Escherichia coli 055:B5, 10 mg/kg, Sigma) or vehicle (pyrogen-free water). Animals of treatment groups were also given either dexamethasone (4 mg/kg, 30 min prior to LPS injection) or the matrix metalloproteinases (MMPs) inhibitor doxycycline (4 mg/kg, 30 min after LPS injection). Both activities and protein levels of MMP-2 (p < 0.001) and MMP-9 (p < 0.001) were significantly upregulated in aortic homogenates from LPS-treated rats, associated with decreased ZO-1 (p < 0.001) and syndecan-1 (p = 0.011) protein contents. Both dexamethasone and doxycycline could significantly inhibit MMPs activity and reserve the expressions of ZO-1 and syndecan-1. The inhibition of MMPs by dexamethasone was significantly lower than that by doxycycline, while the rescue of syndecan-1 expression from LPS-induced endotoxemic rat thoracic aorta was significantly higher in the dexamethasone-treated compared to the doxycycline-treated (p = 0.03). In conclusion, activation of MMPs plays important role in regulating ZO-1 and syndecan-1 protein levels in LPS mediated endothelial perturbation. Both dexamethasone and doxycycline inhibit activation of MMPs that may contribute to the rescue of ZO-1 and syndecan-1 expression.

New diagnostic biomarkers or therapeutic targets for sepsis have substantial significance for critical care medicine. In this study, 192 differentially expressed proteins were selected through iTRAQ. Based on cluster analysis of protein expression dynamics and protein-protein interactions, hemopexin, vimentin, and heat shock protein 90 were selected for further investigation. It was demonstrated that serum vimentin (VIM) levels were significantly increased in patients with sepsis and septic shock compared to controls and that VIM expression was significantly increased in lymphocytes isolated from septic shock and sepsis patients compared to controls. Moreover, a nonsurvivor group had higher serum VIM levels and VIM expression in lymphocytes. Caspase-3 was significantly upregulated in Jurkat T cells lacking VIM and when exposed to LPS compared to control cells. In contrast, caspase-3 was reduced nearly 40% in cells over-expressing VIM. IL-2, IL-10 and IFN-α levels were significantly decreased in cells lacking VIM compared to control cells, whereas they were not significantly altered in cells over-expressing VIM. These findings suggest that VIM modulates lymphocyte apoptosis and inflammatory responses and that VIM could be a new target for the diagnosis and prognostic prediction of patients with sepsis or septic shock.