Automated prediction of extubation success in extremely preterm infants: the APEX multicenter study

Neonatal respiratory care practice among level III and IV NICUs in New England

OBJECTIVES

To assess respiratory care guidelines and explore variations in management of very low birth weight (VLBW) infants within a collaborative care framework. Additionally, to gather clinical leaders' perspectives on guidelines and preferences for ventilation modalities.

STUDY DESIGN

Leaders from each NICU participated in a practice survey regarding the prevalence of unit clinical guidelines, and management, at many stages of care.

RESULTS

Units have an average of 4.3 (±2.1) guidelines, of 9 topics queried. Guideline prevalence was not associated with practice or outcomes. An FiO2 requirement of 0.3-0.4 and a CPAP of 6-7 cmH2O, are the most common thresholds for surfactant administration, which is most often done after intubation, and followed by weaning from ventilatory support. Volume targeted ventilation is commonly used. Extubation criteria vary widely.

CONCLUSIONS

Results identify trends and areas of variation and suggest that the presence of guidelines alone is not predictive of outcome.

Using machine-learning models to predict extubation failure in neonates with bronchopulmonary dysplasia

AIM

To develop a decision-support tool for predicting extubation failure (EF) in neonates with bronchopulmonary dysplasia (BPD) using a set of machine-learning algorithms.

METHODS

A dataset of 284 BPD neonates on mechanical ventilation was used to develop predictive models via machine-learning algorithms, including extreme gradient boosting (XGBoost), random forest, support vector machine, naïve Bayes, logistic regression, and k-nearest neighbor. The top three models were assessed by the area under the receiver operating characteristic curve (AUC), and their performance was tested by decision curve analysis (DCA). Confusion matrix was used to show the high performance of the best model. The importance matrix plot and SHapley Additive exPlanations values were calculated to evaluate the feature importance and visualize the results. The nomogram and clinical impact curves were used to validate the final model.

RESULTS

According to the AUC values and DCA results, the XGboost model performed best (AUC = 0.873, sensitivity = 0.896, specificity = 0.838). The nomogram and clinical impact curve verified that the XGBoost model possessed a significant predictive value. The following were predictive factors for EF: pO2, hemoglobin, mechanical ventilation (MV) rate, pH, Apgar score at 5 min, FiO2, C-reactive protein, Apgar score at 1 min, red blood cell count, PIP, gestational age, highest FiO2 at the first 24 h, heart rate, birth weight, pCO2. Further, pO2, hemoglobin, and MV rate were the three most important factors for predicting EF.

CONCLUSIONS

The present study indicated that the XGBoost model was significant in predicting EF in BPD neonates with mechanical ventilation, which is helpful in determining the right extubation time among neonates with BPD to reduce the occurrence of complications.

Weaning from mechanical ventilation and assessment of extubation readiness

Tremendous advancements in neonatal respiratory care have contributed to the improved survival of extremely preterm infants (gestational age ≤ 28 weeks). While mechanical ventilation is often considered one of the most important breakthroughs in neonatology, it is also associated with numerous short and long-term complications. For those reasons, clinical research has focused on strategies to avoid or reduce exposure to mechanical ventilation. Nonetheless, in the extreme preterm population, 70-100% of infants born 22-28 weeks of gestation are exposed to mechanical ventilation, with nearly 50% being ventilated for ≥ 3 weeks. As contemporary practices have shifted towards selectively reserving mechanical ventilation for those patients, mechanical ventilation weaning and extubation remain a priority yet offer a heightened challenge for clinicians. In this review, we will summarize the evidence for different strategies to expedite weaning and assess extubation readiness in preterm infants, with a particular focus on extremely preterm infants.

Artificial intelligence in the NICU to predict extubation success in prematurely born infants

OBJECTIVES

Mechanical ventilation in prematurely born infants, particularly if prolonged, can cause long term complications including bronchopulmonary dysplasia. Timely extubation then is essential, yet predicting its success remains challenging. Artificial intelligence (AI) may provide a potential solution.

CONTENT

A narrative review was undertaken to explore AI's role in predicting extubation success in prematurely born infants. Across the 11 studies analysed, the range of reported area under the receiver operator characteristic curve (AUC) for the selected prediction models was between 0.7 and 0.87. Only two studies implemented an external validation procedure. Comparison to the results of clinical predictors was made in two studies. One group reported a logistic regression model that outperformed clinical predictors on decision tree analysis, while another group reported clinical predictors outperformed their artificial neural network model (AUCs: ANN 0.68 vs. clinical predictors 0.86). Amongst the studies there was an heterogenous selection of variables for inclusion in prediction models, as well as variations in definitions of extubation failure.

SUMMARY

Although there is potential for AI to enhance extubation success, no model's performance has yet surpassed that of clinical predictors.

OUTLOOK

Future studies should incorporate external validation to increase the applicability of the models to clinical settings.

A clinical informatics approach to bronchopulmonary dysplasia: current barriers and future possibilities

Bronchopulmonary dysplasia (BPD) is a complex, multifactorial lung disease affecting preterm neonates that can result in long-term pulmonary and non-pulmonary complications. Current therapies mainly focus on symptom management after the development of BPD, indicating a need for innovative approaches to predict and identify neonates who would benefit most from targeted or earlier interventions. Clinical informatics, a subfield of biomedical informatics, is transforming healthcare by integrating computational methods with patient data to improve patient outcomes. The application of clinical informatics to develop and enhance clinical therapies for BPD presents opportunities by leveraging electronic health record data, applying machine learning algorithms, and implementing clinical decision support systems. This review highlights the current barriers and the future potential of clinical informatics in identifying clinically relevant BPD phenotypes and developing clinical decision support tools to improve the management of extremely preterm neonates developing or with established BPD. However, the full potential of clinical informatics in advancing our understanding of BPD with the goal of improving patient outcomes cannot be achieved unless we address current challenges such as data collection, storage, privacy, and inherent data bias.

Predicting CPAP failure after less invasive surfactant administration (LISA) in preterm infants by machine learning model on vital parameter data: a pilot study

Objective. Less invasive surfactant administration (LISA) has been introduced to preterm infants with respiratory distress syndrome on continuous positive airway pressure (CPAP) support in order to avoid intubation and mechanical ventilation. However, after this LISA procedure, a significant part of infants fails CPAP treatment (CPAP-F) and requires intubation in the first 72 h of life, which is associated with worse complication free survival chances. The aim of this study was to predict CPAP-F after LISA, based on machine learning (ML) analysis of high resolution vital parameter monitoring data surrounding the LISA procedure.Approach. Patients with a gestational age (GA) <32 weeks receiving LISA were included. Vital parameter data was obtained from a data warehouse. Physiological features (HR, RR, peripheral oxygen saturation (SpO2) and body temperature) were calculated in eight 0.5 h windows throughout a period 1.5 h before to 2.5 h after LISA. First, physiological data was analyzed to investigate differences between the CPAP-F and CPAP-Success (CPAP-S) groups. Next, the performance of two types of ML models (logistic regression: LR, support vector machine: SVM) for the prediction of CPAP-F were evaluated.Main results. Of 51 included patients, 18 (35%) had CPAP-F. Univariate analysis showed lower SpO2, temperature and heart rate variability (HRV) before and after the LISA procedure. The best performing ML model showed an area under the curve of 0.90 and 0.93 for LR and SVM respectively in the 0.5 h window directly after LISA, with GA, HRV, respiration rate and SpO2as most important features. Excluding GA decreased performance in both models.Significance. In this pilot study we were able to predict CPAP-F with a ML model of patient monitor signals, with best performance in the first 0.5 h after LISA. Using ML to predict CPAP-F based on vital signals gains insight in (possibly modifiable) factors that are associated with LISA failure and can help to guide personalized clinical decisions in early respiratory management.

Clinical usefulness of reintubation criteria in extremely preterm infants: a cohort study

OBJECTIVE

To describe the thresholds of instability used by clinicians at reintubation and evaluate the accuracy of different combinations of criteria in predicting reintubation decisions.

DESIGN

Secondary analysis using data obtained from the prospective observational Automated Prediction of Extubation Readiness study (NCT01909947) between 2013 and 2018.

SETTING

Multicentre (three neonatal intensive care units).

PATIENTS

Infants with birth weight ≤1250 g, mechanically ventilated and undergoing their first planned extubation were included.

INTERVENTIONS

After extubation, hourly O2 requirements, blood gas values and occurrence of cardiorespiratory events requiring intervention were recorded for 14 days or until reintubation, whichever came first.

MAIN OUTCOME MEASURES

Thresholds at reintubation were described and grouped into four categories: increased O2, respiratory acidosis, frequent cardiorespiratory events and severe cardiorespiratory events (requiring positive pressure ventilation). An automated algorithm was used to generate multiple combinations of criteria from the four categories and compute their accuracies in capturing reintubated infants (sensitivity) without including non-reintubated infants (specificity).

RESULTS

55 infants were reintubated (median gestational age 25.2 weeks (IQR 24.5-26.1 weeks), birth weight 750 g (IQR 640-880 g)), with highly variable thresholds at reintubation. After extubation, reintubated infants had significantly greater O2 needs, lower pH, higher pCO2 and more frequent and severe cardiorespiratory events compared with non-reintubated infants. After evaluating 123 374 combinations of reintubation criteria, Youden indices ranged from 0 to 0.46, suggesting low accuracy. This was primarily attributable to the poor agreement between clinicians on the number of cardiorespiratory events at which to reintubate.

CONCLUSIONS

Criteria used for reintubation in clinical practice are highly variable, with no combination accurately predicting the decision to reintubate.

Optimal timing of extubation in preterm infants

In neonatal intensive care, endotracheal intubation is usually performed as an urgent or semi-urgent procedure in infants with critical or unstable conditions related to progressive respiratory failure. Extubation is not. Patients undergoing extubation are typically stable, with improved respiratory function. The key elements to facilitating extubation are to recognize improvement in respiratory status, promote weaning of mechanical ventilation, and accurately identify readiness for removal of the endotracheal tube. Therefore, extubation should be a planned and well-organized procedure. In this review, we will appraise the evidence for existing predictors of extubation readiness and provide patient-specific, pathophysiology-derived strategies to optimize the timing and success of extubation in neonates, with a focus on extremely preterm infants.

Accuracy of lung ultrasound in predicting extubation failure in neonates: A systematic review and meta-analysis

OBJECTIVE

To systematically review and meta-analyze the diagnostic accuracy of lung ultrasound score (LUS) in predicting extubation failure in neonates.

STUDY DESIGN

MEDLINE, COCHRANE, EMBASE, CINAHL, and clinicaltrials.gov were searched up to 30 November 2022, for studies evaluating the diagnostic accuracy of LUS in predicting extubation outcome in mechanically ventilated neonates.

METHODOLOGY

Two investigators independently assessed study eligibility, extracted data, and assessed study quality using the Quality Assessment for Studies of Diagnostic Accuracy 2 tool. We conducted a meta-analysis of pooled diagnostic accuracy data using random-effect models. Data were reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We calculated pooled sensitivity and specificity, pooled diagnostic odds ratios with 95% confidence intervals (CI), and area under the curve (AUC).

RESULTS

Eight observational studies involving 564 neonates were included, and the risk of bias was low in seven studies. The pooled sensitivity and specificity for LUS in predicting extubation failure in neonates were 0.82 (95% CI: 0.75-0.88) and 0.83 (95% CI: 0.78-0.86), respectively. The pooled diagnostic odds ratio was 21.24 (95% CI: 10.45-43.19), and the AUC for LUS predicting extubation failure was 0.87 (95% CI: 0.80-0.95). Heterogeneity among included studies was low, both graphically and by statistical criteria (I2  = 7.35%, p = 0.37).

CONCLUSIONS

The predictive value of LUS in neonatal extubation failure may hold promise. However, given the current level of evidence and the methodological heterogeneity observed, there is a clear need for large-scale, well-designed prospective studies that establish standardized protocols for lung ultrasound performance and scoring.

REGISTRATION

The protocol was registered in OSF (https://doi.org/10.17605/OSF.IO/ZXQUT).