Development and external validation of a novel modality for rapid recognition of aortic dissection based on peripheral pulse oximetry waveforms

BACKGROUND

Aortic dissection (AD) is a life-threatening cardiovascular emergency that is often misdiagnosed as other chest pain conditions. Physiologically, AD may cause abnormalities in peripheral blood flow, which can be detected using pulse oximetry waveforms.

PURPOSE

This study aimed to assess the feasibility of identifying AD based on pulse oximetry waveforms and to highlight the key waveform features that play a crucial role in this diagnostic method.

METHODS

This prospective study employed high-risk chest pain cohorts from two emergency departments. The initial cohort was enriched with AD patients (n = 258, 47% AD) for model development, while the second cohort consisted of chest pain patients awaiting angiography (n = 71, 25% AD) and was used for external validation. Pulse oximetry waveforms from the four extremities were collected for each patient. After data preprocessing, a recognition model based on the random forest algorithm was trained using patients' gender, age, and waveform difference features extracted from the pulse oximetry waveforms. The performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis and decision curve analysis (DCA). The importance of features was also assessed using Shapley Value and Gini importance.

RESULTS

The model demonstrated strong performance in identifying AD in both the training and external validation sets. In the training set, the model achieved an area under the ROC curve of 0.979 (95% CI: 0.961-0.990), sensitivity of 0.918 (95% CI: 0.873-0.955), specificity of 0.949 (95% CI: 0.912-0.985), and accuracy of 0.933 (95% CI: 0.904-0.959). In the external validation set, the model attained an area under the ROC curve of 0.855 (95% CI: 0.720-0.965), sensitivity of 0.889 (95% CI: 0.722-1.000), specificity of 0.698 (95% CI: 0.566-0.812), and accuracy of 0.794 (95% CI: 0.672-0.878). Decision curve analysis (DCA) further showed that the model provided a substantial net benefit for identifying AD. The median mean and median variance of the four limbs' signals were the most influential features in the recognition model.

CONCLUSIONS

This study demonstrated the feasibility and strong performance of identifying AD based on peripheral pulse oximetry waveforms in high-risk chest pain populations in the emergency setting. The findings also provided valuable insights for future human fluid dynamics simulations to elucidate the impact of AD on blood flow in greater detail.

Modified furosemide responsiveness index and biomarkers for AKI progression and prognosis: a prospective observational study

BACKGROUND

Modified furosemide responsiveness index (mFRI) is a novel biomarker for assessing diuretic response and AKI progression in patients with early AKI. However, the comparative predictive performance of mFRI and novel renal biomarkers for adverse renal outcomes remains unclear. In a single-center prospective study, we aimed to evaluate the discriminatory abilities of mFRI and other novel renal biomarkers in predicting AKI progression and prognosis in patients with initial mild and moderate AKI (KDIGO stage 1 to 2).

RESULTS

Patients with initial mild and moderate AKI within 48 h following cardiac surgery were included in this study. The mFRI, renal biomarkers (including serum or urinary neutrophil gelatinase-associated lipocalin [sNGAL or uNGAL], serum cystatin C, urinary N-acetyl-beta-D-glycosaminidase [uNAG], urinary albumin-to-creatinine ratio) and cytokines (TNF, IL-1β, IL-2R, IL-6, IL-8, and IL-10) were measured at AKI diagnosis. The mFRI was calculated for each patient, which was defined as 2-hour urine output divided by furosemide dose and body weight. Of 1013 included patients, 154 (15.2%) experienced AKI progression, with 59 (5.8%) progressing to stage 3 and 33 (3.3%) meeting the composite outcome of hospital mortality or receipt of renal replacement therapy (RRT). The mFRI showed non-inferiority or potential superiority to renal biomarkers and cytokines in predicting AKI progression (area under the curve [AUC] 0.80, 95% confidence interval [CI] 0.77-0.82), progression to stage 3 (AUC 0.87, 95% CI 0.85-0.89), and composite outcome of death and receipt of RRT (AUC 0.85, 95% CI 0.82-0.87). Furthermore, the combination of a functional biomarker (mFRI) and a urinary injury biomarker (uNAG or uNGAL) resulted in a significant improvement in the prediction of adverse renal outcomes than either individual biomarker (all P < 0.05). Moreover, incorporating these panels into clinical model significantly enhanced its predictive capacity for adverse renal outcomes, as demonstrated by the C index, integrated discrimination improvement, and net reclassification improvement (all P < 0.05).

CONCLUSIONS

As a rapid, cost-effective and easily accessible biomarker, mFRI, exhibited superior or comparable predictive capabilities for AKI progression and prognosis compared to renal biomarkers in cardiac surgical patients with mild to moderate AKI.

TRIAL REGISTRATION

Clinicaltrials.gov, NCT04962412. Registered July 15, 2021, https://clinicaltrials.gov/ct2/show/NCT04962412?cond=NCT04962412&draw=2&rank=1 .

Remifentanil vs. dexmedetomidine for cardiac surgery patients with noninvasive ventilation intolerance: a multicenter randomized controlled trial

BACKGROUND

The optimal sedative regime for noninvasive ventilation (NIV) intolerance remains uncertain. The present study aimed to assess the efficacy and safety of remifentanil (REM) compared to dexmedetomidine (DEX) in cardiac surgery patients with moderate-to-severe intolerance to NIV.

METHODS

In this multicenter, prospective, single-blind, randomized controlled study, adult cardiac surgery patients with moderate-to-severe intolerance to NIV were enrolled and randomly assigned to be treated with either REM or DEX for sedation. The status of NIV intolerance was evaluated using a four-point NIV intolerance score at different timepoints within a 72-h period. The primary outcome was the mitigation rate of NIV intolerance following sedation.

RESULTS

A total of 179 patients were enrolled, with 89 assigned to the REM group and 90 to the DEX group. Baseline characteristics were comparable between the two groups, including NIV intolerance score [3, interquartile range (IQR) 3-3 vs. 3, IQR 3-4, p = 0.180]. The chi-squared test showed that mitigation rate, defined as the proportion of patients who were relieved from their initial intolerance status, was not significant at most timepoints, except for the 15-min timepoint (42% vs. 20%, p = 0.002). However, after considering the time factor, generalized estimating equations showed that the difference was statistically significant, and REM outperformed DEX (odds ratio = 3.31, 95% confidence interval: 1.35-8.12, p = 0.009). Adverse effects, which were not reported in the REM group, were encountered by nine patients in the DEX group, with three instances of bradycardia and six cases of severe hypotension. Secondary outcomes, including NIV failure (5.6% vs. 7.8%, p = 0.564), tracheostomy (1.12% vs. 0%, p = 0.313), ICU LOS (7.7 days, IQR 5.8-12 days vs. 7.0 days, IQR 5-10.6 days, p = 0.219), and in-hospital mortality (1.12% vs. 2.22%, p = 0.567), demonstrated comparability between the two groups.

CONCLUSIONS

In summary, our study demonstrated no significant difference between REM and DEX in the percentage of patients who achieved mitigation among cardiac surgery patients with moderate-to-severe NIV intolerance. However, after considering the time factor, REM was significantly superior to DEX. Trial registration ClinicalTrials.gov (NCT04734418), registered on January 22, 2021. URL of the trial registry record: https://register.

CLINICALTRIALS

gov/prs/app/action/SelectProtocol?sid=S000AM4S&selectaction=Edit&uid=U00038YX&ts=3&cx=eqn1z0 .

Structural basis of antagonist selectivity in endothelin receptors

Endothelins and their receptors, ETA and ETB, play vital roles in maintaining vascular homeostasis. Therapeutically targeting endothelin receptors, particularly through ETA antagonists, has shown efficacy in treating pulmonary arterial hypertension (PAH) and other cardiovascular- and renal-related diseases. Here we present cryo-electron microscopy structures of ETA in complex with two PAH drugs, macitentan and ambrisentan, along with zibotentan, a selective ETA antagonist, respectively. Notably, a specialized anti-ETA antibody facilitated the structural elucidation. These structures, together with the active-state structures of ET-1-bound ETA and ETB, and the agonist BQ3020-bound ETB, in complex with Gq, unveil the molecular basis of agonist/antagonist binding modes in endothelin receptors. Key residues that confer antagonist selectivity to endothelin receptors were identified along with the activation mechanism of ETA. Furthermore, our results suggest that ECL2 in ETA can serve as an epitope for antibody-mediated receptor antagonism. Collectively, these insights establish a robust theoretical framework for the rational design of small-molecule drugs and antibodies with selective activity against endothelin receptors.

Using a Novel Machine-Learning Algorithm as an Auxiliary Approach to Predict the Transfusion Volume in Mitral Valve Surgery

Background: Blood transfusion is an indispensable supportive therapy. It plays a pivotal role in the perioperative management of cardiac surgery. The aim of this study was to develop a model for predicting the transfusion volume in isolated mitral valve surgery. Methods: We gathered data from 677 patients undergoing isolated mitral valve surgery with and without simultaneous tricuspid valve operation. The dataset was partitioned into a training dataset (70%) and a testing dataset (30%). We evaluated 18 machine-learning algorithms, incorporating inputs from 36 demographic and perioperative features. Additionally, the performance of multiple linear regressions was compared with machine-learning algorithms. CatBoost was selected for further analysis, and Shapley additive explanation (SHAP) values were employed to evaluate feature importance. Finally, we explored the impact of various features on the accuracy of CatBoost by analyzing the reasons for misjudgment. Results: CatBoost outperformed all 18 machine learning algorithms with an R-squared value of 0.420, mean absolute error of 0.702, mean squared error of 1.208, and root mean squared error of 1.090, surpassing multiple linear regression. The analysis of the testing group achieved 72.5% accuracy. SHAP identified 20 pertinent features influencing transfusion volume. No significant differences were observed between correctly and incorrectly predicted groups in tricuspid valve repair, American Society of Anesthesiologists classification, or platelet count. Conclusion: CatBoost effectively predicts the intraoperative transfusion volume in mitral valve surgery, aiding clinicians in transfusion decision-making and enhancing patient care.

AKI-Pro score for predicting progression to severe acute kidney injury or death in patients with early acute kidney injury after cardiac surgery

BACKGROUND

No reliable clinical tools exist to predict acute kidney injury (AKI) progression. We aim to explore a scoring system for predicting the composite outcome of progression to severe AKI or death within seven days among early AKI patients after cardiac surgery.

METHODS

In this study, we used two independent cohorts, and patients who experienced mild/moderate AKI within 48 h after cardiac surgery were enrolled. Eventually, 3188 patients from the MIMIC-IV database were used as the derivation cohort, while 499 patients from the Zhongshan cohort were used as external validation. The primary outcome was defined by the composite outcome of progression to severe AKI or death within seven days after enrollment. The variables identified by LASSO regression analysis were entered into logistic regression models and were used to construct the risk score.

RESULTS

The composite outcome accounted for 3.7% (n = 119) and 7.6% (n = 38) of the derivation and validation cohorts, respectively. Six predictors were assembled into a risk score (AKI-Pro score), including female, baseline eGFR, aortic surgery, modified furosemide responsiveness index (mFRI), SOFA, and AKI stage. And we stratified the risk score into four groups: low, moderate, high, and very high risk. The risk score displayed satisfied predictive discrimination and calibration in the derivation and validation cohort. The AKI-Pro score discriminated the composite outcome better than CRATE score, Cleveland score, AKICS score, Simplified renal index, and SRI risk score (all P < 0.05).

CONCLUSIONS

The AKI-Pro score is a new clinical tool that could assist clinicians to identify early AKI patients at high risk for AKI progression or death.

Skin mottling score assesses peripheral tissue hypoperfusion in critically ill patients following cardiac surgery

BACKGROUND

Skin mottling is a common manifestation of peripheral tissue hypoperfusion, and its severity can be described using the skin mottling score (SMS). This study aims to evaluate the value of the SMS in detecting peripheral tissue hypoperfusion in critically ill patients following cardiac surgery.

METHODS

Critically ill patients following cardiac surgery with risk factors for tissue hypoperfusion were enrolled (n = 373). Among these overall patients, we further defined a hypotension population (n = 178) and a shock population (n = 51). Hemodynamic and perfusion parameters were recorded. The primary outcome was peripheral hypoperfusion, defined as significant prolonged capillary refill time (CRT, > 3.0 s). The characteristics and hospital mortality of patients with and without skin mottling were compared. The area under receiver operating characteristic curves (AUROC) were used to assess the accuracy of SMS in detecting peripheral hypoperfusion. Besides, the relationships between SMS and conventional hemodynamic and perfusion parameters were investigated, and the factors most associated with the presence of skin mottling were identified.

RESULTS

Of the 373-case overall population, 13 (3.5%) patients exhibited skin mottling, with SMS ranging from 1 to 5 (5, 1, 2, 2, and 3 cases, respectively). Patients with mottling had lower mean arterial pressure, higher vasopressor dose, less urine output (UO), higher CRT, lactate levels and hospital mortality (84.6% vs. 12.2%, p < 0.001). The occurrences of skin mottling were higher in hypotension population and shock population, reaching 5.6% and 15.7%, respectively. The AUROC for SMS to identify peripheral hypoperfusion was 0.64, 0.68, and 0.81 in the overall, hypotension, and shock populations, respectively. The optimal SMS threshold was 1, which corresponded to specificities of 98, 97 and 91 and sensitivities of 29, 38 and 67 in the three populations (overall, hypotension and shock). The correlation of UO, lactate, CRT and vasopressor dose with SMS was significant, among them, UO and CRT were identified as two major factors associated with the presence of skin mottling.

CONCLUSION

In critically ill patients following cardiac surgery, SMS is a very specific yet less sensitive parameter for detecting peripheral tissue hypoperfusion.

Furosemide Responsiveness Predicts Acute Kidney Injury Progression After Cardiac Surgery

BACKGROUND

As patients with acute kidney injury (AKI) progress to a higher stage, the risk for poor outcomes dramatically rises. Early identification of patients at high risk for AKI progression remains a major challenge. This study aimed to evaluate the value of furosemide responsiveness (FR) for predicting AKI progression in patients with initial mild and moderate AKI after cardiac surgery.

METHODS

We performed 2 separate exploratory analyses. The Zhongshan cohort was a single-center, prospective, observational cohort, whereas the Beth Israel Deaconess Medical Center cohort was a single-center, retrospective cohort. We calculated 2 FR parameters for each patient, namely the FR index and modified FR index, defined as 2-hour urine output divided by furosemide dose (FR index, mL/mg/2 h) and by furosemide dose and body weight (modified FR index, mL/[mg·kg]/2 h), respectively. The primary outcome was AKI progression within 7 days.

RESULTS

AKI progression occurred in 80 (16.0%) and 359 (11.3%) patients in the Zhongshan and Beth Israel Deaconess Medical Center cohorts, respectively. All FR parameters (considered continuously or in quartiles) were inversely associated with risk of AKI progression in both cohorts (all adjusted P < .01). The addition of FR parameters significantly improved prediction for AKI progression based on baseline clinical models involving C-index, net reclassification improvement, and integrated discrimination improvement index in both cohorts (all P < .01).

CONCLUSIONS

FR parameters were inversely associated with risk of AKI progression in patients with mild and moderate AKI after cardiac surgery. The addition of FR parameters significantly improved prediction for AKI progression based on baseline clinical models.

Studying the Influence of Finite Element Mesh Size on the Accuracy of Ventricular Tachycardia Simulation

Background

Ventricular tachycardia (VT) is a life-threatening heart condition commonly seen in patients with myocardial infarction (MI). Although personalized computational modeling has been used to understand VT and its treatment noninvasively, this approach can be computationally intensive and time consuming. Therefore, finding a balance between mesh size and computational efficiency is important. This study aimed to find an optimal mesh resolution that minimizes the need for computational resources while maintaining numerical accuracy and to investigate the effect of mesh resolution variation on the simulation results.

Methods

We constructed ventricular models from contrast-enhanced magnetic resonance imaging data from six patients with MI. We created seven different models for each patient, with average edge lengths ranging from 315 to 645 µm using commercial software, Mimics. Programmed electrical stimulation was used to assess VT inducibility from 19 sites in each heart model.

Results

The simulation results in the slab model with adaptive tetrahedral mesh (same as in the patient-specific model) showed that the absolute and relative differences in conduction velocity (CV) were 6.1 cm/s and 7.8% between average mesh sizes of 142 and 600 µm, respectively. However, the simulation results in the six patient-specific models showed that average mesh sizes with 350 µm yielded over 85% accuracy for clinically relevant VT. Although average mesh sizes of 417 and 478 µm could also achieve approximately 80% accuracy for clinically relevant VT, the percentage of incorrectly predicted VTs increases. When conductivity was modified to match the CV in the model with the finest mesh size, the overall ratio of positively predicted VT increased.

Conclusions

The proposed personalized heart model could achieve an optimal balance between simulation time and VT prediction accuracy when discretized with adaptive tetrahedral meshes with an average edge length about 350 µm.

Impact of body mass index on perioperative mortality of acute stanford type A aortic dissection: a systematic review and meta-analysis

BACKGROUND

Obesity may increase perioperative mortality of acute Stanford type A aortic dissection (ATAAD). However, the available evidence was limited. This study aimed to systematically review published literatures about body mass index (BMI) and perioperative mortality of ATAAD.

METHODS

Electronic literature search was conducted in PubMed, Medline, Embase and Cochrane Library databases. All observational studies that investigated BMI and perioperative mortality of ATAAD were included. Pooled odds ratio (OR) and 95% confidence interval (CI) were calculated using a random-effects model. Meta-regression analysis was performed to assess the effects of different clinical variables on BMI and perioperative mortality of ATAAD. Sensitivity analysis was performed to determine the sources of heterogeneity. Egger's linear regression method and funnel plot were used to determine the publication bias.

RESULTS

A total of 12 studies with 5,522 patients were eligible and included in this meta-analysis. Pooled analysis showed that perioperative mortality of ATAAD increased by 22% for each 1 kg/m2 increase in BMI (OR = 1.22, 95% CI: 1.10-1.35). Univariable meta-regression analysis indicated that age and female gender significantly modified the association between BMI and perioperative mortality of ATAAD in a positive manner (meta-regression on age: coefficient = 0.04, P = 0.04; meta-regression on female gender: coefficient = 0.02, P = 0.03). Neither significant heterogeneity nor publication bias were found among included studies.

CONCLUSIONS

BMI is closely associated with perioperative mortality of ATAAD. Optimal perioperative management needs to be further explored and individualized for obese patient with ATAAD, especially in elderly and female populations.

TRIAL REGISTRATION

PROSPERO (CRD42022358619). BMI and perioperative mortality of ATAAD.

Effect of glucocorticoid for patients with type A aortic dissection undergoing surgical repair with deep hypothermic circulatory arrest: A single-center, retrospective study

BACKGROUND

Postoperative patients with Type A aortic dissection (TAAD) often experience severe inflammatory responses caused by multiple factors perioperatively. However, the effect of postoperative glucocorticoid (GC) use, which is a potent anti-inflammatory agent, on complications or all-cause mortality is unclear.

METHODS

Patients with TAAD who underwent surgical repair requiring deep hypothermic circulatory arrest between January 2020 and December 2021 were included in the study. Characteristics of patients treated with and without GCs were compared. The primary outcome was in-hospital mortality, and a composite secondary outcome was defined as in-hospital death or any major complications. Propensity score matching was used to balance baseline differences between groups. Kaplan-Meier curves were used to compare survival probability.

RESULTS

A total of 393 postoperative patients with TAAD were included in the study. Forty of them (10.2%) received GC treatment at a median daily methylprednisolone-equivalent dose of 0.6 mg/kg (0.4-0.7) for a median period of 2 (1-3) days. Patients on GCs had more intraoperative blood transfusions, higher postoperative APACHE II (12 vs 9, p = .004) and SOFA (9 vs 6, p < .001) scores, worse perioperative hepatic, renal and cardiac function. The in-hospital mortality in the matched cohort did not differ between groups [GC n = 11/40 (27.5%) versus Non-GC n = 19/80 (23.8%); p = .661].

CONCLUSIONS

The propensity to use GCs correlated with the critical status of the patient. However, low dose and short-term postoperative GC treatment did not reduce in-hospital mortality rates among patients with TAAD. A more appropriate regimen should be further investigated.

Postoperative glucocorticoids in patients with acute type A aortic dissection (GLAD): study protocol for a prospective, single-center, randomized controlled trial

BACKGROUND

Patients receiving surgical treatment of acute type A Aortic Dissection (aTAAD) are common to suffer organ dysfunction in the intensive care unit due to overwhelming inflammation. Previous studies have revealed that glucocorticoids may reduce complications in certain patient groups, but evidence between postoperative glucocorticoids administration and improvement in organ dysfunction after aTAAD surgery are lacking.

METHODS

This study will be an investigator-initiated, prospective, single-blind, randomized, single-center study. Subjects with confirmed diagnosis of aTAAD undergoing surgical treatment will be enrolled and 1:1 randomly assigned to receive either glucocorticoids or normal treatment. All patients in the glucocorticoids group will be given methylprednisolone intravenously for 3 days after enrollment. The primary endpoint will be the amplitude of variation of Sequential Organ Failure Assessment score on post-operative day 4 compared to baseline.

DISCUSSION

The trial will explore the rationale for postoperative application of glucocorticoids in patients after aTAAD surgery.

TRIAL REGISTRATION

This study has been registered on ClinicalTrials.gov (NCT04734418).

2022 CMICS Expert Consensus on the Management of Isolated Tricuspid Regurgitation after Left-Sided Valve Surgery

Tricuspid regurgitation (TR) may occur late after left-sided valve surgery (LSVS). Isolated tricuspid regurgitation after left-sided valve surgery (iTR-LSVS) refers to isolated TR without significant lesions in the mitral and/or aortic position late after mitral and/or aortic replacement or repair. Severe TR has a negative impact on long-term prognosis and requires surgical or transcatheter treatment. However, there is no clear recommendation on when and how intervention should be performed for patients with iTR-LSVS in the current guidelines for the management of valvular heart disease. The historically high operative mortality may be reduced by current minimally invasive techniques and transcatheter therapy. To further understand iTR-LSVS, standardize the treatment, improve the prognosis, and promote the collaboration, the Chinese Minimally Invasive Cardiovascular Surgery Committee (CMICS) wrote this expert consensus on the management of iTR-LSVS from the aspects of etiology, preoperative evaluation, indications for intervention, surgical treatment, transcatheter therapy, and postoperative management.

Extracellular vesicles derived from CD4+ T cells carry DGKK to promote sepsis-induced lung injury by regulating oxidative stress and inflammation

BACKGROUND

Sepsis is an abnormal immune response after infection, wherein the lung is the most susceptible organ to fail, leading to acute lung injury. To overcome the limitations of current therapeutic strategies and develop more specific treatment, the inflammatory process, in which T cell-derived extracellular vesicles (EVs) play a central role, should be explored deeply.

METHODS

Liquid chromatography-tandem mass spectrometry was performed for serum EV protein profiling. The serum diacylglycerol kinase kappa (DGKK) and endotoxin contents of patients with sepsis-induced lung injury were measured. Apoptosis, oxidative stress, and inflammation in A549 cells, bronchoalveolar lavage fluid, and lung tissues of mice were measured by flow cytometry, biochemical analysis, enzyme-linked immunosorbent assay, quantitative real-time polymerase chain reaction, and western blot.

RESULTS

DGKK, the key regulator of the diacylglycerol (DAG)/protein kinase C (PKC) pathway, exhibited elevated expression in serum EVs of patients with sepsis-induced lung injury and showed strong correlation with sepsis severity and disease progression. DGKK was expressed in CD4+ T cells under regulation of the NF-κB pathway and delivered by EVs to target cells, including alveolar epithelial cells. EVs produced by CD4+ T lymphocytes exerted toxic effects on A549 cells to induce apoptotic cell death, oxidative cell damage, and inflammation. In mice with sepsis induced by cecal ligation and puncture, EVs derived from CD4+ T cells also promoted tissue damage, oxidative stress, and inflammation in the lungs. These toxic effects of T cell-derived EVs were attenuated by the inhibition of PKC and NOX4, the downstream effectors of DGKK and DAG.

CONCLUSIONS

This approach established the mechanism that T-cell-derived EVs carrying DGKK triggered alveolar epithelial cell apoptosis, oxidative stress, inflammation, and tissue damage in sepsis-induced lung injury through the DAG/PKC/NOX4 pathway. Thus, T-cell-derived EVs and the elevated distribution of DGKK should be further investigated to develop therapeutic strategies for sepsis-induced lung injury.

Interpreting Infrared Thermography with Deep Learning to Assess the Mortality Risk of Critically Ill Patients at Risk of Hypoperfusion

Background

Hypoperfusion, a common manifestation of many critical illnesses, could lead to abnormalities in body surface thermal distribution. However, the interpretation of thermal images is difficult. Our aim was to assess the mortality risk of critically ill patients at risk of hypoperfusion in a prospective cohort by infrared thermography combined with deep learning methods.

Methods

This post-hoc study was based on a cohort at high-risk of hypoperfusion. Patients' legs were selected as the region of interest. Thermal images and conventional hypoperfusion parameters were collected. Six deep learning models were attempted to derive the risk of mortality (range: 0 to 100%) for each patient. The area under the receiver operating characteristic curve (AUROC) was used to evaluate predictive accuracy.

Results

Fifty-five hospital deaths occurred in a cohort consisting of 373 patients. The conventional hypoperfusion (capillary refill time and diastolic blood pressure) and thermal (low temperature area rate and standard deviation) parameters demonstrated similar predictive accuracies for hospital mortality (AUROC 0.73 and 0.77). The deep learning methods, especially the ResNet (18), could further improve the accuracy. The AUROC of ResNet (18) was 0.94 with a sensitivity of 84% and a specificity of 91% when using a cutoff of 36%. ResNet (18) presented a significantly increasing trend in the risk of mortality in patients with normotension (13 [7 to 26]), hypotension (18 [8 to 32]) and shock (28 [14 to 62]).

Conclusions

Interpreting infrared thermography with deep learning enables accurate and non-invasive assessment of the severity of patients at risk of hypoperfusion.

A cross-sectional survey on the lung ultrasound training and practice of respiratory therapists in mainland China

Purpose

This national study aimed to investigate the lung ultrasound (LUS) training and practice of respiratory therapists (RTs) in mainland China.

Methods

A cross-sectional multicenter survey was conducted from May 22, 2021 to August 12, 2021, through online platforms. This survey included RTs in mainland China. The survey was divided into four sections: (1) demographic characteristics and basic information; (2) basic information about LUS training and practice; (3) LUS practice details; and (4) Other ultrasound training and practice.

Results

A total of 514 responses were received, and 494 valid responses were included in the analysis. 81.2% (401/494) participants’ highest degree of education was a bachelor’s degree, and 43.1% (213/494) participants were at level II in terms of job ranking. 99.2%(490/494) participants agreed that the RTs needed to learn lung ultrasound, but only 12.3% (61/494) participants had received a LUS training course. Further, 66.2% (327/494) experienced participants responded to Sect. 3. Most of RTs used LUS when the patient had hypoxia (265/327, 81%) or dyspnea (260/317, 79.5%); they also used it during spontaneous breathing trial(SBT) (191/327, 58.4%) or in prone position (177/327, 54.1%). The A-line (302/327, 92.4%), B-line (299/327, 91.4%), lung slide (263/327, 80.4%), and bat sign (259/327, 79.2%) were well known as LUS signs. Also, 30.6% (100/327) participants did not use the LUS protocol in their clinical practice, and only 25.4%(83/327) participants said they had used LUS scores. Moreover, 55.7% (182/327) participants frequently changed the respiratory therapy strategy according to LUS results.

Conclusions

We should improve the number and workplace of RTs in mainland China in the future. We should also standardize the application of LUS practice and training for RTs in mainland China and establish corresponding certification pathways.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12890-022-02213-6.

INFRARED THERMOGRAPHY-BASED BODY-SURFACE THERMAL INHOMOGENEITY MONITORING TO ASSESS THE SEVERITY OF HYPOPERFUSION IN CRITICALLY ILL PATIENTS

ABSTRACT

Background: Uneven body-surface thermal distribution is a manifestation of hypoperfusion and can be quantified by infrared thermography. Our aim was to investigate whether body-surface thermal inhomogeneity could accurately evaluate the severity of patients at risk of hypoperfusion. Methods: This was a prospective cohort study in which infrared thermography images were taken from unilateral legs of critically ill patients at high risk of hypoperfusion in a cardiac surgical intensive care unit. For each patient, five body-surface thermal inhomogeneity parameters, including standard deviation (SD), kurtosis, skewness, entropy, and low-temperature area rate (LTAR), were calculated. Demographic, clinical, and thermal characteristics of deceased and living patients were compared. The risk of mortality and capillary refill time (CRT) were chosen as the primary outcome and benchmarking parameter for hypoperfusion, respectively. The area under the receiver operating characteristic curve (AUROC) was used to evaluate predictive accuracy. Results: Three hundred seventy-three patients were included, and 55 (14.7%) died during hospital stay. Of inhomogeneity parameters, SD (0.738) and LTAR (0.768) had similar AUROC to CRT (0.757) for assessing mortality risk. Besides, there was a tendency for LTAR (1%-3%-7%) and SD (0.81°C-0.88°C-0.94°C) to increase in normotensive, hypotensive, and shock patients. These thermal parameters are associated with CRT, lactate, and blood pressure. The AUROC of a combined prediction incorporating three thermal inhomogeneity parameters (SD, kurtosis, and entropy) was considerably higher at 0.866. Conclusions: Body-surface thermal inhomogeneity provided a noninvasive and accurate assessment of the severity of critically ill patients at high risk of hypoperfusion.

Data science in the intensive care unit

In this editorial, we comment on the current development and deployment of data science in intensive care units (ICUs). Data in ICUs can be classified into qualitative and quantitative data with different technologies needed to translate and interpret them. Data science, in the form of artificial intelligence (AI), should find the right interaction between physicians, data and algorithm. For individual patients and physicians, sepsis and mechanical ventilation have been two important aspects where AI has been extensively studied. However, major risks of bias, lack of generalizability and poor clinical values remain. AI deployment in the ICUs should be emphasized more to facilitate AI development. For ICU management, AI has a huge potential in transforming resource allocation. The coronavirus disease 2019 pandemic has given opportunities to establish such systems which should be investigated further. Ethical concerns must be addressed when designing such AI.