A validated risk stratification tool for congenital diaphragmatic hernia is required for accurate outcomes analyses.
Existing mortality-predictive models include those of the congenital diaphragmatic hernia Study Group based on birth weight and the 5-minute Apgar score.
|The Congenital Diaphragmatic Hernia Study Group model showed superior discrimination|
|Journal of Pediatric Surgery|
The Canadian Neonatal Network score is based on gestational age and admission score in Score for Neonatal Acute Physiology version II.
The Wilford Hall/Santa Rosa clinical prediction formula is derived from blood gas measurements.
Dr Robert Bairdab and colleagues from Canada evaluated the calibration and discrimination of these predictive models using the Canadian Pediatric Surgical Network dataset.
The team collected neonatal risk variables and birth hospital survivorship prospectively in 11 perinatal centers, between 2005 and 2006.
Actual vs predicted outcomes were analyzed for each equation to measure the calibration and discrimination of each model.
The researchers reported that 20 of 94 infants with congenital diaphragmatic hernia died during birth hospitalization.
The Congenital Diaphragmatic Hernia Study Group model demonstrated superior discrimination.
Model calibration reflected by the Hosmer-Lemeshow P value was poorest with the Wilford Hall/Santa Rosa clinical prediction formula.
However, it was comparable between the Congenital Diaphragmatic Hernia Study Group and Canadian Neonatal Network.
Dr Bairdab's team concluded, "Predictive outcome models are essential for risk-adjusted outcome analysis of congenital diaphragmatic hernia."
"The ideal predictive equation should prove robust across congenital diaphragmatic hernia datasets."