Predictors of Prolonged Mechanical Ventilation in Paediatric Patients
After Anatomical Correction of Simple Congenital Heart Disease
Abstract
Abstract Background: We aimed to use preoperative clinical data from
paediatric patients with simple congenital heart disease to predict the
risk of prolonged mechanical ventilation after surgery. Methods: The
clinical data from paediatric patients with simple congenital heart
disease who underwent anatomical correction under cardiopulmonary bypass
in a single centre during a continuous period were retrospectively
collected. Univariate and multivariate logistic regression analyses were
performed to identify the risk factors for prolonged mechanical
ventilation (>24 h) after surgery, and a mathematical model
was established. Then, using data from another centre, we adopted an ROC
curve to verify the scalability of the model. Results: A total of 585
paediatric patients were eligible for inclusion in this study.
Multivariate logistic regression analysis showed that weight (kg), the
size of the ventricular septal defect, the size of the atrial septal
defect and the shunt direction of the defect site were significantly
correlated with prolonged mechanical ventilation (>24 h)
after surgery. The risk prediction model was established and the area
under the curve of the model was 0.853 (ROC curve). A set of data from
another heart centre, with equivalent inclusion criteria, was used to
validate the scalability of the model, and the area under the curve of
the accepted validated data was 0.841 (ROC curve). Conclusions: The risk
of prolonged mechanical ventilation (>24 h) after surgery
in paediatric patients with simple congenital heart disease with
anatomical correction assisted by cardiopulmonary bypass can be well
predicted by using preoperative clinical data.