Thenar oxygen saturation (StO2) alterations during a spontaneous breathing trial predict extubation failure

In vivocharacterization of the optical and hemodynamic properties of the human sternocleidomastoid muscle through ultrasound-guided hybrid near-infrared spectroscopies

Objective.In this paper, we present a detailedin vivocharacterization of the optical and hemodynamic properties of the human sternocleidomastoid muscle (SCM), obtained through ultrasound-guided near-infrared time-domain and diffuse correlation spectroscopies.Approach.A total of sixty-five subjects (forty-nine females, sixteen males) among healthy volunteers and thyroid nodule patients have been recruited for the study. Their SCM hemodynamic (oxy-, deoxy- and total hemoglobin concentrations, blood flow, blood oxygen saturation and metabolic rate of oxygen extraction) and optical properties (wavelength dependent absorption and reduced scattering coefficients) have been measured by the use of a novel hybrid device combining in a single unit time-domain near-infrared spectroscopy, diffuse correlation spectroscopy and simultaneous ultrasound imaging.Main results.We provide detailed tables of the results related to SCM baseline (i.e. muscle at rest) properties, and reveal significant differences on the measured parameters due to variables such as side of the neck, sex, age, body mass index, depth and thickness of the muscle, allowing future clinical studies to take into account such dependencies.Significance.The non-invasive monitoring of the hemodynamics and metabolism of the sternocleidomastoid muscle during respiration became a topic of increased interest partially due to the increased use of mechanical ventilation during the COVID-19 pandemic. Near-infrared diffuse optical spectroscopies were proposed as potential practical monitors of increased recruitment of SCM during respiratory distress. They can provide clinically relevant information on the degree of the patient's respiratory effort that is needed to maintain an optimal minute ventilation, with potential clinical application ranging from evaluating chronic pulmonary diseases to more acute settings, such as acute respiratory failure, or to determine the readiness to wean from invasive mechanical ventilation.

Assessment of power spectral density of microvascular hemodynamics in skeletal muscles at very low and low-frequency via near-infrared diffuse optical spectroscopies

In this work, we used a hybrid time domain near-infrared spectroscopy (TD-NIRS) and diffuse correlation spectroscopy (DCS) device to retrieve hemoglobin and blood flow oscillations of skeletal muscle microvasculature. We focused on very low (VLF) and low-frequency (LF) oscillations (i.e., frequency lower than 0.145 Hz), that are related to myogenic, neurogenic and endothelial activities. We measured power spectral density (PSD) of blood flow and hemoglobin concentration in four muscles (thenar eminence, plantar fascia, sternocleidomastoid and forearm) of 14 healthy volunteers to highlight possible differences in microvascular hemodynamic oscillations. We observed larger PSDs for blood flow compared to hemoglobin concentration, in particular in case of distal muscles (i.e., thenar eminence and plantar fascia). Finally, we compared the PSDs measured on the thenar eminence of healthy subjects with the ones measured on a septic patient in the intensive care unit: lower power in the endothelial-dependent frequency band, and larger power in the myogenic ones were observed in the septic patient, in accordance with previous works based on laser doppler flowmetry.

Microvascular effects of oxygen and carbon dioxide measured by vascular occlusion test in healthy volunteers

BACKGROUND

Changes in near-infrared spectroscopy-derived regional tissue oxygen saturation (StO₂) during a vascular occlusion test (VOT; ischemic provocation of microcirculation by rapid inflation and deflation of a tourniquet) allow estimating peripheral tissue O₂ consumption (desaturation slope; DS), vascular reactivity (recovery slope; RS) and post-ischemic hyperperfusion (AUC-H). The effects of isolated alterations in the inspiratory fraction of O₂ (FiO₂) and changes in expiratory CO₂ remain to be elucidated. Therefore, in this secondary analysis we determined the effects of standardized isolated instances of hypoxia, hyperoxia, hypocapnia and hypercapnia on the VOT-induced StO₂ changes in healthy volunteers (n = 20) to establish reference values for future physiological studies.

METHODS

StO₂ was measured on the thenar muscle. Multiple VOTs were performed in a standardized manner: i.e. at room air (baseline), during hyperoxia (FiO₂ 1.0), mild hypoxia (FiO₂ ≈ 0.11), and after a second baseline, during hypocapnia (end-tidal CO₂ (etCO₂) 2.5-3.0 vol%) and hypercapnia (etCO₂ 7.0-7.5 vol%) at room air. Differences in DS, RS, and AUC-H were tested using repeated-measures ANOVA.

RESULTS

DS and RS remained constant during all applied conditions. AUC-H after hypoxia was smaller compared to hyperoxia (963 %*sec vs hyperoxia 1702 %*sec, P = 0.005), while there was no difference in AUC-H duration between hypoxia and baseline. The StO₂ peak (after tourniquet deflation) during hypoxia was lower compared to baseline and hyperoxia (92 % vs 94 % and 98 %, P < 0.001).

CONCLUSION

We conclude that in healthy volunteers at rest, common situations observed during anesthesia and intensive care such as exposure to hypoxia, hyperoxia, hypocapnia, or hypercapnia, did not affect peripheral tissue O₂ consumption and vascular reactivity as assessed by VOT-induced changes in StO₂. These observations may serve as reference values for future physiological studies.

TRIAL REGISTRATION

This study represents a secondary analysis of an original study which has been registered at ClinicalTrials.gov nr: NCT02561052.

Alterations in tissue oxygen saturation measured by near-infrared spectroscopy in trauma patients after initial resuscitation are associated with occult shock

PURPOSE

Persistent occult hypoperfusion after initial resuscitation is strongly associated with increased morbidity and mortality after severe trauma. The objective of this study was to analyze regional tissue oxygenation, along with other global markers, as potential detectors of occult shock in otherwise hemodynamically stable trauma patients.

METHODS

Trauma patients undergoing active resuscitation were evaluated 8 h after hospital admission with the measurement of several global and local hemodynamic/metabolic parameters. Apparently hemodynamically stable (AHD) patients, defined as having SBP ≥ 90 mmHg, HR < 100 bpm and no vasopressor support, were followed for 48 h, and finally classified according to the need for further treatment for persistent bleeding (defined as requiring additional red blood cell transfusion), initiation of vasopressors and/or bleeding control with surgery and/or angioembolization. Patients were labeled as "Occult shock" (OS) if they required any intervention or "Truly hemodynamically stable" (THD) if they did not. Regional tissue oxygenation (rSO₂) was measured non-invasively by near-infrared spectroscopy (NIRS) on the forearm. A vascular occlusion test was performed, allowing a 3-min deoxygenation period and a reoxygenation period following occlusion release. Minimal rSO₂ (rSO₂min), Delta-down (rSO₂-rSO₂min), maximal rSO₂ following cuff-release (rSO₂max), and Delta-up (rSO₂max-rSO₂min) were computed. The NIRS response to the occlusion test was also measured in a control group of healthy volunteers.

RESULTS

Sixty-six consecutive trauma patients were included. After 8 h, 17 patients were classified as AHD, of whom five were finally considered to have OS and 12 THD. No hemodynamic, metabolic or coagulopathic differences were observed between the two groups, while NIRS-derived parameters showed statistically significant differences in Delta-down, rSO₂min, and Delta-up.

CONCLUSIONS

After 8 h of care, NIRS evaluation with an occlusion test is helpful for identifying occult shock in apparently hemodynamically stable patients.

LEVEL OF EVIDENCE

IV, descriptive observational study.

TRIAL REGISTRATION

ClinicalTrials.gov Registration Number: NCT02772653.

An artificial intelligence system to predict the optimal timing for mechanical ventilation weaning for intensive care unit patients: A two-stage prediction approach

BACKGROUND

For the intensivists, accurate assessment of the ideal timing for successful weaning from the mechanical ventilation (MV) in the intensive care unit (ICU) is very challenging.

PURPOSE

Using artificial intelligence (AI) approach to build two-stage predictive models, namely, the try-weaning stage and weaning MV stage to determine the optimal timing of weaning from MV for ICU intubated patients, and implement into practice for assisting clinical decision making.

METHODS

AI and machine learning (ML) technologies were used to establish the predictive models in the stages. Each stage comprised 11 prediction time points with 11 prediction models. Twenty-five features were used for the first-stage models while 20 features were used for the second-stage models. The optimal models for each time point were selected for further practical implementation in a digital dashboard style. Seven machine learning algorithms including Logistic Regression (LR), Random Forest (RF), Support Vector Machines (SVM), K Nearest Neighbor (KNN), lightGBM, XGBoost, and Multilayer Perception (MLP) were used. The electronic medical records of the intubated ICU patients of Chi Mei Medical Center (CMMC) from 2016 to 2019 were included for modeling. Models with the highest area under the receiver operating characteristic curve (AUC) were regarded as optimal models and used to develop the prediction system accordingly.

RESULTS

A total of 5,873 cases were included in machine learning modeling for Stage 1 with the AUCs of optimal models ranging from 0.843 to 0.953. Further, 4,172 cases were included for Stage 2 with the AUCs of optimal models ranging from 0.889 to 0.944. A prediction system (dashboard) with the optimal models of the two stages was developed and deployed in the ICU setting. Respiratory care members expressed high recognition of the AI dashboard assisting ventilator weaning decisions. Also, the impact analysis of with- and without-AI assistance revealed that our AI models could shorten the patients' intubation time by 21 hours, besides gaining the benefit of substantial consistency between these two decision-making strategies.

CONCLUSION

We noticed that the two-stage AI prediction models could effectively and precisely predict the optimal timing to wean intubated patients in the ICU from ventilator use. This could reduce patient discomfort, improve medical quality, and lower medical costs. This AI-assisted prediction system is beneficial for clinicians to cope with a high demand for ventilators during the COVID-19 pandemic.

Incidence of extubation failure and its predictors among adult patients in intensive care unit of low-resource setting: A prospective observational study

BACKGROUND

Previous studies have found an association between various predictors and extubation failure (EF) in intensive care units (ICUs). However, this problem remains unexplored in low-resource settings, where predicting the extubation outcomes are more challenging. This study investigates the incidence of EF and its predictors among patients who received mechanical ventilation (MV).

METHODS

This is a prospective observational study of 123 patients' ≥ 18 years of age receiving MV for ≥ 48 hours and tolerated spontaneous breathing trials (SBTs) in the ICU of a low-resource setting. We collected data on the baseline characteristics and clinical profiles before and after SBTs. Patients were categorized into extubation failure (EF) and extubation success (ES) groups. Multivariate logistic regression analyses were performed to identify independent predictors for EF. A p-value < 0.05 is considered statistically significant.

RESULTS

We included 123 patients, and 42 (34.15%) had developed EF. The identified predictors for EF: Moderate to copious secretions (adjusted odds ratio [AOR]: 3.483 [95% confidence interval [CI] 1.10-11.4]), age > 60 years of age ([AOR]: 4.157 [95% CI 1.38-12.48]), and prolonged duration of MV ≥ 10 days ([AOR]: 4.77 [95% CI 1.55-14.66]).

CONCLUSION

Moderate to copious secretions, patients > 60 years of age, and prolonged duration of MV ≥ 10 days were the best predictors of EF. Based on our findings, we recommend that the identified predictors could help in the decision-making process of extubation from MV.

Early prediction of noninvasive ventilation failure after extubation: development and validation of a machine-learning model

Background

Noninvasive ventilation (NIV) has been widely used in critically ill patients after extubation. However, NIV failure is associated with poor outcomes. This study aimed to determine early predictors of NIV failure and to construct an accurate machine-learning model to identify patients at risks of NIV failure after extubation in intensive care units (ICUs).

Methods

Patients who underwent NIV after extubation in the eICU Collaborative Research Database (eICU-CRD) were included. NIV failure was defined as need for invasive ventilatory support (reintubation or tracheotomy) or death after NIV initiation. A total of 93 clinical and laboratory variables were assessed, and the recursive feature elimination algorithm was used to select key features. Hyperparameter optimization was conducted with an automated machine-learning toolkit called Neural Network Intelligence. A machine-learning model called Categorical Boosting (CatBoost) was developed and compared with nine other models. The model was then prospectively validated among patients enrolled in the Cardiac Surgical ICU of Zhongshan Hospital, Fudan University.

Results

Of 929 patients included in the eICU-CRD cohort, 248 (26.7%) had NIV failure. The time from extubation to NIV, age, Glasgow Coma Scale (GCS) score, heart rate, respiratory rate, mean blood pressure (MBP), saturation of pulse oxygen (SpO₂), temperature, glucose, pH, pressure of oxygen in blood (PaO₂), urine output, input volume, ventilation duration, and mean airway pressure were selected. After hyperparameter optimization, our model showed the greatest accuracy in predicting NIV failure (AUROC: 0.872 [95% CI 0.82–0.92]) among all predictive methods in an internal validation. In the prospective validation cohort, our model was also superior (AUROC: 0.846 [95% CI 0.80–0.89]). The sensitivity and specificity in the prediction group is 89% and 75%, while in the validation group they are 90% and 70%. MV duration and respiratory rate were the most important features. Additionally, we developed a web-based tool to help clinicians use our model.

Conclusions

This study developed and prospectively validated the CatBoost model, which can be used to identify patients who are at risk of NIV failure. Thus, those patients might benefit from early triage and more intensive monitoring.

Trial registration: NCT03704324. Registered 1 September 2018, https://register.clinicaltrials.gov.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12890-022-02096-7.

Usefulness of a finger-mounted tissue oximeter with near-infrared spectroscopy for evaluating the intestinal oxygenation and viability in rats

PURPOSE

To investigate the utility of the device for evaluating intestinal oxygenation and viability using an animal model.

METHODS

Sprague-Dawley rats underwent laparotomy under general anesthesia, and the blood vessels in the terminal ileum were clamped to create ischemia. We measured the regional tissue oxygenation saturation (rSO2) using an oximeter after 1, 3, and 6 h of vessel clamping. Ischemic tissue damage was assessed using a histological score. The intestine was reperfused after each clamping period, and intestinal rSO2 and survival rate were evaluated.

RESULTS

When reperfusion was performed at 1 and 3 h after ischemia, rSO2 increased after 10 min, and it improved to the same level as for normal intestine after 1 h; all rats survived for 1 week. In contrast, after 6 h of ischemia, rSO2 did not increase after reperfusion, and all animals died within 2 days. The histological scores increased after 1 h of reperfusion, with longer clamping periods.

CONCLUSION

A finger-mounted tissue oximeter could evaluate intestinal ischemia and the viability, which is thus considered to be a promising result for future clinical application.

Development and Validation of a Machine-Learning Model for Prediction of Extubation Failure in Intensive Care Units

Background: Extubation failure (EF) can lead to an increased chance of ventilator-associated pneumonia, longer hospital stays, and a higher mortality rate. This study aimed to develop and validate an accurate machine-learning model to predict EF in intensive care units (ICUs). Methods: Patients who underwent extubation in the Medical Information Mart for Intensive Care (MIMIC)-IV database were included. EF was defined as the need for ventilatory support (non-invasive ventilation or reintubation) or death within 48 h following extubation. A machine-learning model called Categorical Boosting (CatBoost) was developed based on 89 clinical and laboratory variables. SHapley Additive exPlanations (SHAP) values were calculated to evaluate feature importance and the recursive feature elimination (RFE) algorithm was used to select key features. Hyperparameter optimization was conducted using an automated machine-learning toolkit (Neural Network Intelligence). The final model was trained based on key features and compared with 10 other models. The model was then prospectively validated in patients enrolled in the Cardiac Surgical ICU of Zhongshan Hospital, Fudan University. In addition, a web-based tool was developed to help clinicians use our model. Results: Of 16,189 patients included in the MIMIC-IV cohort, 2,756 (17.0%) had EF. Nineteen key features were selected using the RFE algorithm, including age, body mass index, stroke, heart rate, respiratory rate, mean arterial pressure, peripheral oxygen saturation, temperature, pH, central venous pressure, tidal volume, positive end-expiratory pressure, mean airway pressure, pressure support ventilation (PSV) level, mechanical ventilation (MV) durations, spontaneous breathing trial success times, urine output, crystalloid amount, and antibiotic types. After hyperparameter optimization, our model had the greatest area under the receiver operating characteristic (AUROC: 0.835) in internal validation. Significant differences in mortality, reintubation rates, and NIV rates were shown between patients with a high predicted risk and those with a low predicted risk. In the prospective validation, the superiority of our model was also observed (AUROC: 0.803). According to the SHAP values, MV duration and PSV level were the most important features for prediction. Conclusions: In conclusion, this study developed and prospectively validated a CatBoost model, which better predicted EF in ICUs than other models.