Lung Mechanics of Mechanically Ventilated Patients With COVID-19: Analytics With High-Granularity Ventilator Waveform Data

Machine learning modeling of lung mechanics: Assessing the variability and propagation of uncertainty in respiratory-system compliance and airway resistance

BACKGROUND AND OBJECTIVE

Traditional assessment of patient response in mechanical ventilation relies on respiratory-system compliance and airway resistance. Clinical evidence has shown high variability in these parameters, highlighting the difficulty of predicting them before the start of ventilation therapy. This motivates the creation of computational models that can connect structural and tissue features with lung mechanics. In this work, we leverage machine learning (ML) techniques to construct predictive lung function models informed by non-linear finite element simulations, and use them to investigate the propagation of uncertainty in the lung mechanical response.

METHODS

We revisit a continuum poromechanical formulation of the lungs suitable for determining patient response. Based on this framework, we create high-fidelity finite element models of human lungs from medical images. We also develop a low-fidelity model based on an idealized sphere geometry. We then use these models to train and validate three ML architectures: single-fidelity and multi-fidelity Gaussian process regression, and artificial neural networks. We use the best predictive ML model to further study the sensitivity of lung response to variations in tissue structural parameters and boundary conditions via sensitivity analysis and forward uncertainty quantification. Codes are available for download at https://github.com/comp-medicine-uc/ML-lung-mechanics-UQ RESULTS: The low-fidelity model delivers a lung response very close to that predicted by high-fidelity simulations and at a fraction of the computational time. Regarding the trained ML models, the multi-fidelity GP model consistently delivers better accuracy than the single-fidelity GP and neural network models in estimating respiratory-system compliance and resistance (R2∼0.99). In terms of computational efficiency, our ML model delivers a massive speed-up of ∼970,000× with respect to high-fidelity simulations. Regarding lung function, we observed an almost matched and non-linear behavior between specific structural parameters and chest wall stiffness with compliance. Also, we observed a strong modulation of airways resistance with tissue permeability.

CONCLUSIONS

Our findings unveil the relevance of specific lung tissue parameters and boundary conditions in the respiratory-system response. Furthermore, we highlight the advantages of adopting a multi-fidelity ML approach that combines data from different levels to yield accurate and efficient estimates of clinical mechanical markers. We envision that the methods presented here can open the way to the development of predictive ML models of the lung response that can inform clinical decisions.

Effect of patient-ventilator asynchrony on lung and diaphragmatic injury in experimental acute respiratory distress syndrome in a porcine model

BACKGROUND

Patient-ventilator asynchrony during mechanical ventilation may exacerbate lung and diaphragm injury in spontaneously breathing subjects. We investigated whether subject-ventilator asynchrony increases lung or diaphragmatic injury in a porcine model of acute respiratory distress syndrome (ARDS).

METHODS

ARDS was induced in adult female pigs by lung lavage and injurious ventilation before mechanical ventilation by pressure assist-control for 12 h. Mechanically ventilated pigs were randomised to breathe spontaneously with or without induced subject-ventilator asynchrony or neuromuscular block (n=7 per group). Subject-ventilator asynchrony was produced by ineffective, auto-, or double-triggering of spontaneous breaths. The primary outcome was mean alveolar septal thickness (where thickening of the alveolar wall indicates worse lung injury). Secondary outcomes included distribution of ventilation (electrical impedance tomography), lung morphometric analysis, inflammatory biomarkers (gene expression), lung wet-to-dry weight ratio, and diaphragmatic muscle fibre thickness.

RESULTS

Subject-ventilator asynchrony (median [interquartile range] 28.8% [10.4] asynchronous breaths of total breaths; n=7) did not increase mean alveolar septal thickness compared with synchronous spontaneous breathing (asynchronous breaths 1.0% [1.6] of total breaths; n=7). There was no difference in mean alveolar septal thickness throughout upper and lower lung lobes between pigs randomised to subject-ventilator asynchrony vs synchronous spontaneous breathing (87.3-92.2 μm after subject-ventilator asynchrony, compared with 84.1-95.0 μm in synchronised spontaneous breathing;). There were also no differences between groups in wet-to-dry weight ratio, diaphragmatic muscle fibre thickness, atelectasis, lung aeration, or mRNA expression levels for inflammatory cytokines pivotal in ARDS pathogenesis.

CONCLUSIONS

Subject-ventilator asynchrony during spontaneous breathing did not exacerbate lung injury and dysfunction in experimental porcine ARDS.

Early short course of neuromuscular blocking agents in patients with COVID-19 ARDS: a propensity score analysis

BACKGROUND

The role of neuromuscular blocking agents (NMBAs) in coronavirus disease 2019 (COVID-19) acute respiratory distress syndrome (ARDS) is not fully elucidated. Therefore, we aimed to investigate in COVID-19 patients with moderate-to-severe ARDS the impact of early use of NMBAs on 90-day mortality, through propensity score (PS) matching analysis.

METHODS

We analyzed a convenience sample of patients with COVID-19 and moderate-to-severe ARDS, admitted to 244 intensive care units within the COVID-19 Critical Care Consortium, from February 1, 2020, through October 31, 2021. Patients undergoing at least 2 days and up to 3 consecutive days of NMBAs (NMBA treatment), within 48 h from commencement of IMV were compared with subjects who did not receive NMBAs or only upon commencement of IMV (control). The primary objective in the PS-matched cohort was comparison between groups in 90-day in-hospital mortality, assessed through Cox proportional hazard modeling. Secondary objectives were comparisons in the numbers of ventilator-free days (VFD) between day 1 and day 28 and between day 1 and 90 through competing risk regression.

RESULTS

Data from 1953 patients were included. After propensity score matching, 210 cases from each group were well matched. In the PS-matched cohort, mean (± SD) age was 60.3 ± 13.2 years and 296 (70.5%) were male and the most common comorbidities were hypertension (56.9%), obesity (41.1%), and diabetes (30.0%). The unadjusted hazard ratio (HR) for death at 90 days in the NMBA treatment vs control group was 1.12 (95% CI 0.79, 1.59, p = 0.534). After adjustment for smoking habit and critical therapeutic covariates, the HR was 1.07 (95% CI 0.72, 1.61, p = 0.729). At 28 days, VFD were 16 (IQR 0-25) and 25 (IQR 7-26) in the NMBA treatment and control groups, respectively (sub-hazard ratio 0.82, 95% CI 0.67, 1.00, p = 0.055). At 90 days, VFD were 77 (IQR 0-87) and 87 (IQR 0-88) (sub-hazard ratio 0.86 (95% CI 0.69, 1.07; p = 0.177).

CONCLUSIONS

In patients with COVID-19 and moderate-to-severe ARDS, short course of NMBA treatment, applied early, did not significantly improve 90-day mortality and VFD. In the absence of definitive data from clinical trials, NMBAs should be indicated cautiously in this setting.

An archetypal model of a breathable air-circuit in an electro-pneumatic ventilator device

Mechanical ventilator is a machine that is mechanically designed to deliver breathable air in and out of the lungs to provide a breathing mechanism for a patient who is physically unable to breathe, it is an indispensable life-support device in critical care medicine and medical emergencies such as scenarios during the COVID-19 pandemic. This research presents a model design of the pneumatic circuit that is electronically controlled, by using computer-aided pneumatic rig over selected 5/3, 5/2, 3/2 solenoid gating valves, the performance of these valves must be investigated to ascertain the most appropriate valve to be used for the electro-pneumatic mechanical ventilator. An elaborate parametric investigation reported for volume-controlled ventilators illustrate the influences of key parameters on the dynamics of the ventilated respiratory system. This study presents the linearity of tidal volume, peak pressure and lung compliance for the parameters considered. However, the maximum pressure of the ventilation device increases slowly when the tidal volumes exceed 600 ml. In addition, influence of evacuation time of the ventilator predicted over high throughput in time regimes of 1 s; 1.2 s; 1.4 s; 1.6 s, and 1.8 s showed that the pressure platform in the pipe might not appear if the exhaust time of the ventilator is less than 1.6 s. The 5/2 solenoid valve was considered the best with consistent flowrate. The archetypal model of the pneumatic circuit developed in this research could find vital application in the design of patient-interfacing devices particularly in ventilators and neonatal incubator.

Putative Role of the Lung-Brain Axis in the Pathogenesis of COVID-19-Associated Respiratory Failure: A Systematic Review

The emergence of SARS-CoV-2 and its related disease caused by coronavirus (COVID-19) has posed a huge threat to the global population, with millions of deaths and the creation of enormous social and healthcare pressure. Several studies have shown that besides respiratory illness, other organs may be damaged as well, including the heart, kidneys, and brain. Current evidence reports a high frequency of neurological manifestations in COVID-19, with significant prognostic implications. Importantly, emerging literature is showing that the virus may spread to the central nervous system through neuronal routes, hitting the brainstem and cardiorespiratory centers, potentially exacerbating the respiratory illness. In this systematic review, we searched public databases for all available evidence and discuss current clinical and pre-clinical data on the relationship between the lung and brain during COVID-19. Acknowledging the involvement of these primordial brain areas in the pathogenesis of the disease may fuel research on the topic and allow the development of new therapeutic strategies.

Lower Driving Pressure and Neuromuscular Blocker Use Are Associated With Decreased Mortality in Patients With COVID-19 ARDS

BACKGROUND

The impact of mechanical ventilation parameters and management on outcomes of patients with coronavirus disease 2019 (COVID-19) ARDS is unclear.

METHODS

This multi-center observational study enrolled consecutive mechanically ventilated patients with COVID-19 ARDS admitted to one of 7 Korean ICUs between February 1, 2020-February 28, 2021. Patients who were age < 17 y or had missing ventilation parameters for the first 4 d of mechanical ventilation were excluded. Multivariate logistic regression was used to identify which strategies or ventilation parameters that were independently associated with ICU mortality.

RESULTS

Overall, 129 subjects (males, 60%) with a median (interquartile range) age of 69 (62-78) y were included. Neuromuscular blocker (NMB) use and prone positioning were applied to 76% and 16% of subjects, respectively. The ICU mortality rate was 37%. In the multivariate analysis, higher dynamic driving pressure (ΔP) values during the first 4 d of mechanical ventilation were associated with increased mortality (adjusted odds ratio 1.16 [95% CI 1.00-1.33], P = .046). NMB use was associated with decreased mortality (adjusted odds ratio 0.27 [95% CI 0.09-0.81], P = .02). The median tidal volume values during the first 4 d of mechanical ventilation and the ICU mortality rate were significantly lower in the NMB group than in the no NMB group. However, subjects who received NMB for ≥ 6 d (vs < 6 d) had higher ICU mortality rate.

CONCLUSIONS

In subjects with COVID-19 ARDS receiving mechanical ventilation, ΔP during the first 4 d of mechanical ventilation was independently associated with mortality. The short-term use of NMB facilitated lung-protective ventilation and was independently associated with decreased mortality.

Secondary bloodstream infection in critically ill patients with COVID-19

Objective

Secondary infection, especially bloodstream infection, is an important cause of death in critically ill patients with COVID-19. We aimed to describe secondary bloodstream infection (SBI) in critically ill adults with COVID-19 in the intensive care unit (ICU) and to explore risk factors related to SBI.

Methods

We reviewed all SBI cases among critically ill patients with COVID-19 from 12 February 2020 to 24 March 2020 in the COVID-19 ICU of Jingmen First People's Hospital. We compared risk factors associated with bloodstream infection in this study. All SBIs were confirmed by blood culture.

Results

We identified five cases of SBI among the 32 patients: three with Enterococcus faecium, one mixed septicemia (E. faecium and Candida albicans), and one C. parapsilosis. There were no significant differences between the SBI group and non-SBI group. Significant risk factors for SBI were extracorporeal membrane oxygenation, central venous catheter, indwelling urethral catheter, and nasogastric tube.

Conclusions

Our findings confirmed that the incidence of secondary infection, particularly SBI, and mortality are high among critically ill patients with COVID-19. We showed that long-term hospitalization and invasive procedures such as tracheotomy, central venous catheter, indwelling urethral catheter, and nasogastric tube are risk factors for SBI and other complications.

A Machine-Learning Approach for Dynamic Prediction of Sepsis-Induced Coagulopathy in Critically Ill Patients With Sepsis

Background: Sepsis-induced coagulopathy (SIC) denotes an increased mortality rate and poorer prognosis in septic patients. Objectives: Our study aimed to develop and validate machine-learning models to dynamically predict the risk of SIC in critically ill patients with sepsis. Methods: Machine-learning models were developed and validated based on two public databases named Medical Information Mart for Intensive Care (MIMIC)-IV and the eICU Collaborative Research Database (eICU-CRD). Dynamic prediction of SIC involved an evaluation of the risk of SIC each day after the diagnosis of sepsis using 15 predictive models. The best model was selected based on its accuracy and area under the receiver operating characteristic curve (AUC), followed by fine-grained hyperparameter adjustment using the Bayesian Optimization Algorithm. A compact model was developed, based on 15 features selected according to their importance and clinical availability. These two models were compared with Logistic Regression and SIC scores in terms of SIC prediction. Results: Of 11,362 patients in MIMIC-IV included in the final cohort, a total of 6,744 (59%) patients developed SIC during sepsis. The model named Categorical Boosting (CatBoost) had the greatest AUC in our study (0.869; 95% CI: 0.850-0.886). Coagulation profile and renal function indicators were the most important features for predicting SIC. A compact model was developed with an AUC of 0.854 (95% CI: 0.832-0.872), while the AUCs of Logistic Regression and SIC scores were 0.746 (95% CI: 0.735-0.755) and 0.709 (95% CI: 0.687-0.733), respectively. A cohort of 35,252 septic patients in eICU-CRD was analyzed. The AUCs of the full and the compact models in the external validation were 0.842 (95% CI: 0.837-0.846) and 0.803 (95% CI: 0.798-0.809), respectively, which were still larger than those of Logistic Regression (0.660; 95% CI: 0.653-0.667) and SIC scores (0.752; 95% CI: 0.747-0.757). Prediction results were illustrated by SHapley Additive exPlanations (SHAP) values, which made our models clinically interpretable. Conclusions: We developed two models which were able to dynamically predict the risk of SIC in septic patients better than conventional Logistic Regression and SIC scores.