Gestational Age and Risk of Mortality in Term-Born Critically Ill Neonates Admitted to PICUs in Australia and New Zealand

Potential benefits of prenatal diagnosis of TGA in Australia may be outweighed by the adverse effects of earlier delivery: likely causation and potential solutions

OBJECTIVE

Prenatal diagnosis of transposition of great arteries (TGA) is expected to improve postoperative outcomes after neonatal arterial switch operation (ASO); however, published reports give conflicting results. We aimed to determine the association between prenatal diagnosis and early postoperative outcomes after neonatal ASO.

METHODS

Cohort study involving 243 newborns who underwent ASO (70% prenatally diagnosed) between 2010 and 2019. Multivariable regression was used to determine the association between prenatal diagnosis and (a) birth characteristics and (b) postoperative outcomes.

RESULTS

Gestational age and birthweight centile were lower and small-for-gestational-age more common (11.8% vs 1.4%) in those diagnosed prenatally. Among births which followed labour induction or prelabour caesarean, prenatal diagnosis was associated with earlier gestation at birth (mean (SD), 38.5 (1.6) vs 39.2 (1.4), p=0.01). Among births which followed spontaneous labour, prenatal diagnosis was associated with earlier gestation at labour onset (38.2 (1.8) vs 39.2 (1.4), p=0.01). Prenatal diagnosis was associated with longer postoperative mechanical ventilation (incidence rate ratio 1.74, 95% CI 1.37 to 2.21), intensive care (1.70, 1.31 to 2.21) and hospital length of stay (1.37, 1.14 to 1.66) after ASO. Gestational age mediated up to 60% of the effect of prenatal diagnosis on postoperative outcomes.

CONCLUSION

Among newborns undergoing ASO for TGA, prenatal diagnosis is associated with poorer early postoperative outcomes. In addition to minimising iatrogenic factors (such as planned births) resulting in earlier births, evaluation of other dynamics following a prenatal diagnosis which may result in poor fetal growth and earlier onset of spontaneous labour is important.

Machine Learning Algorithms to Predict Mortality of Neonates on Mechanical Intubation for Respiratory Failure

BACKGROUND

Early identification of critically ill neonates with poor outcomes can optimize therapeutic strategies. We aimed to examine whether machine learning (ML) methods can improve mortality prediction for neonatal intensive care unit (NICU) patients on intubation for respiratory failure.

METHODS

A total of 1734 neonates with respiratory failure were randomly divided into training (70%, n = 1214) and test (30%, n = 520) sets. The primary outcome was the probability of NICU mortality. The areas under the receiver operating characteristic curves (AUCs) of several ML algorithms were compared with those of the conventional neonatal illness severity scoring systems including the NTISS and SNAPPE-II.

RESULTS

For NICU mortality, the random forest (RF) model showed the highest AUC (0.939 (0.921-0.958)) for the prediction of neonates with respiratory failure, and the bagged classification and regression tree model demonstrated the next best results (0.915 (0.891-0.939)). The AUCs of both models were significantly better than the traditional NTISS (0.836 (0.800-0.871)) and SNAPPE-II scores (0.805 (0.766-0.843)). The superior performances were confirmed by higher accuracy and F1 score and better calibration, and the superior and net benefit was confirmed by decision curve analysis. In addition, Shapley additive explanation (SHAP) values were utilized to explain the RF prediction model.

CONCLUSIONS

Machine learning algorithms increase the accuracy and predictive ability for mortality of neonates with respiratory failure compared with conventional neonatal illness severity scores. The RF model is suitable for clinical use in the NICU, and clinicians can gain insights and have better communication with families in advance.