Airways management in SARS-COV-2 acute respiratory failure: A prospective observational multi-center study

Development and external validation of an interpretable machine learning model for the prediction of intubation in the intensive care unit

Given the limited capacity to accurately determine the necessity for intubation in intensive care unit settings, this study aimed to develop and externally validate an interpretable machine learning model capable of predicting the need for intubation among ICU patients. Seven widely used machine learning (ML) algorithms were employed to construct the prediction models. Adult patients from the Medical Information Mart for Intensive Care IV database who stayed in the ICU for longer than 24 h were included in the development and internal validation. The model was subsequently externally validated using the eICU-CRD database. In addition, the SHapley Additive exPlanations method was employed to interpret the influence of individual parameters on the predictions made by the model. A total of 11,988 patients were included in the final cohort for this study. The CatBoost model demonstrated the best performance (AUC: 0.881). In the external validation set, the efficacy of our model was also confirmed (AUC: 0.750), which suggests robust generalization capabilities. The Glasgow Coma Scale (GCS), body mass index (BMI), arterial partial pressure of oxygen (PaO2), respiratory rate (RR) and length of stay (LOS) before ICU were the top 5 features of the CatBoost model with the greatest impact. We developed an externally validated CatBoost model that accurately predicts the need for intubation in ICU patients within 24 to 96 h of admission, facilitating clinical decision-making and has the potential to improve patient outcomes. The prediction model utilizes readily obtainable monitoring parameters and integrates the SHAP method to enhance interpretability, providing clinicians with clear insights into the factors influencing predictions.

The use of vasopressors to reduce post-intubation hypotension in critically ill adult patients undergoing emergent endotracheal intubation: a scoping review

INTRODUCTION

Patients requiring emergent endotracheal intubation are at higher risk of post-intubation hypotension due to altered physiology in critical illness. Post-intubation hypotension increases mortality and hospital length of stay, however, the impact of vasopressors on its incidence and outcomes is not known. This scoping review identified studies reporting hemodynamic data in patients undergoing emergent intubation to provide a literature overview on post-intubation hypotension in cohorts that did and did not receive vasopressors.

METHODS

A systematic search of CINAHL, Cochrane, EMBASE and PubMed-Medline was performed from database inception until September 28, 2023. Two independent reviewers completed the title and abstract screen, full text review and data extraction per PRISMA guidelines. Studies including patients < 18 years or intubations during cardiac arrest were excluded. Primary outcome was the presence of hypotension within 30 min of emergent intubation. Secondary outcomes included mortality at 1 h and in-hospital.

RESULTS

The systematic search yielded 13,126 articles, with 61 selected for final inclusion. There were 24,547 patients with a mean age of 57.2 years and a slight male predominance (63.8%). Respiratory failure was the most common intubation indication. Across 18 studies reporting on vasopressor use prior to intubation, 1171/7085 patients received vasopressors pre-intubation. Post-intubation hypotension occurred in 22.2% of patients across all studies, and in 34.3% of patients in studies where vasopressor administration pre-intubation was specifically reported. One-hour mortality of patients across all studies and within the vasopressor use studies was 1.2% and 1.6%, respectively. In-hospital mortality across studies was 21.5%, and 13.1% in studies which reported on vasopressor use pre-intubation.

CONCLUSION

Patients requiring emergent intubation have a high rate of post-intubation hypotension and in-hospital mortality. While there is an intuitive rationale for the use of vasopressors during emergent intubation, current evidence is limited to support a definitive change in clinical practice at this time.

A Nationwide, Prospective Study of Tracheal Intubation in Critically Ill Adults in Spain: Management, Associated Complications, and Outcomes

OBJECTIVES

Our aims were to explore current intubation practices in Spanish ICUs to determine the incidence and risk factors of peri-intubation complications (primary outcome measure: major adverse events), the rate and factors associated with first-pass success, and their impact on mortality as well as the changes of the intubation procedure observed in the COVID-19 pandemic.

DESIGN

Prospective, observational, and cohort study.

SETTING

Forty-three Spanish ICU.

PATIENTS

A total of 1837 critically ill adult patients undergoing tracheal intubation. The enrollment period was six months (selected by each center from April 16, 2019, to October 31, 2020).

INTERVENTIONS

None.

MEASUREMENT AND MAIN RESULTS

At least one major adverse peri-intubation event occurred in 40.4 % of the patients (973 major adverse events were registered) the most frequent being hemodynamic instability (26.5%) and severe hypoxemia (20.3%). The multivariate analysis identified seven variables independently associated with a major adverse event whereas the use of neuromuscular blocking agents (NMBAs) was associated with reduced odds of major adverse events. Intubation on the first attempt was achieved in 70.8% of the patients. The use of videolaryngoscopy at the first attempt was the only protective factor (odds ratio 0.43; 95% CI, 0.28-0.66; p < 0.001) for first-attempt intubation failure. During the COVID-19 pandemic, the use of videolaryngoscopy and NMBAs increased significantly. The occurrence of a major peri-intubation event was an independent risk factor for 28-day mortality. Cardiovascular collapse also posed a serious threat, constituting an independent predictor of death.

CONCLUSIONS

A major adverse event occurred in up to 40% of the adults intubated in the ICU. Peri-intubation hemodynamic instability but not severe hypoxemia was identified as an independent predictor of death. The use of NMBAs was a protective factor for major adverse events, whereas the use of videolaringoscopy increases the first-pass success rate of intubation. Intubation practices changed during the COVID-19 pandemic.

Improvement in intubation success during COVID-19 pandemic with a simple and low-cost intervention: A quasi-experimental study

OBJECTIVES

primary objective: to improve the FPS rates after an educational intervention.

SECONDARY OBJECTIVE

to describe variables related to FPS in an ED and determine which ones were related to the highest number of attempts.

DESIGN

it was a prospective quasi-experimental study.

SETTING

done in an ED in a public Hospital in Argentina.

PATIENTS

there were patients of all ages with intubation in ED.

INTERVENTIONS

in the middle of the study, an educational intervention was done to improve FPS. Cognitive aids and pre- intubation Checklists were implemented.

MAIN VARIABLES OF INTEREST

the operator experience, the number of intubation attempts, intubation judgment, predictors of a difficult airway, Cormack score, assist devices, complications, blood pressure, heart rate, and pulse oximetry before and after intubation All the intubations were done by direct laryngoscopy (DL).

RESULTS

data from 266 patients were included of which 123 belonged to the basal period and 143 belonged to the post-intervention period. FPS percentage of the pre-intervention group was 69.9% (IC95%: 60.89-77.68) whereas the post-intervention group was 85.3% (IC95%: 78.20-90.48). The difference between these groups was statistically significant (p=0.002). Factors related to the highest number of attempts were low operator experience, Cormack-Lehane 3 score and no training.

CONCLUSIONS

a low-cost and simple educational intervention in airway management was significantly associated with improvement in FPS, reaching the same rate of FPS than in high income countries.

Dendrogram of transparent feature importance machine learning statistics to classify associations for heart failure: A reanalysis of a retrospective cohort study of the Medical Information Mart for Intensive Care III (MIMIC-III) database

BACKGROUND

There is a continual push for developing accurate predictors for Intensive Care Unit (ICU) admitted heart failure (HF) patients and in-hospital mortality.

OBJECTIVE

The study aimed to utilize transparent machine learning and create hierarchical clustering of key predictors based off of model importance statistics gain, cover, and frequency.

METHODS

Inclusion criteria of complete patient information for in-hospital mortality in the ICU with HF from the MIMIC-III database were randomly divided into a training (n = 941, 80%) and test (n = 235, 20%). A grid search was set to find hyperparameters. Machine Learning with XGBoost were used to predict mortality followed by feature importance with Shapely Additive Explanations (SHAP) and hierarchical clustering of model metrics with a dendrogram and heat map.

RESULTS

Of the 1,176 heart failure ICU patients that met inclusion criteria for the study, 558 (47.5%) were males. The mean age was 74.05 (SD = 12.85). XGBoost model had an area under the receiver operator curve of 0.662. The highest overall SHAP explanations were urine output, leukocytes, bicarbonate, and platelets. Average urine output was 1899.28 (SD = 1272.36) mL/day with the hospital mortality group having 1345.97 (SD = 1136.58) mL/day and the group without hospital mortality having 1986.91 (SD = 1271.16) mL/day. The average leukocyte count in the cohort was 10.72 (SD = 5.23) cells per microliter. For the hospital mortality group the leukocyte count was 13.47 (SD = 7.42) cells per microliter and for the group without hospital mortality the leukocyte count was 10.28 (SD = 4.66) cells per microliter. The average bicarbonate value was 26.91 (SD = 5.17) mEq/L. Amongst the group with hospital mortality the average bicarbonate value was 24.00 (SD = 5.42) mEq/L. Amongst the group without hospital mortality the average bicarbonate value was 27.37 (SD = 4.98) mEq/L. The average platelet value was 241.52 platelets per microliter. For the group with hospital mortality the average platelet value was 216.21 platelets per microliter. For the group without hospital mortality the average platelet value was 245.47 platelets per microliter. Cluster 1 of the dendrogram grouped the temperature, platelets, urine output, Saturation of partial pressure of Oxygen (SPO2), Leukocyte count, lymphocyte count, bicarbonate, anion gap, respiratory rate, PCO2, BMI, and age as most similar in having the highest aggregate gain, cover, and frequency metrics.

CONCLUSION

Machine Learning models that incorporate dendrograms and heat maps can offer additional summaries of model statistics in differentiating factors between in patient ICU mortality in heart failure patients.