Comparison between single compartment model and recruitment basis function model on NICU patients

Abstract

Respiratory distress syndrome (RDS) is commonly diagnosed in preterm infants in neonatal intensive care units (NICUs) due to prematurity at birth resulting in surfactant deficiency. Mechanical ventilation (MV) is used to support breathing of infants with RDS. In this study, respiratory mechanics of 10 invasively ventilated infants from Christchurch Women’s Hospital under standard care are compared with two lung mechanics models validated in adult MV patients. A single compartment model is compared with a parabolic basis function model with dynamic elastance (Edrs) used to capture patient-specific effort. This latter model applies parabolic and linear shapes to identify lung recruitment and distension. The model was fit to 25,657 breaths. The median [interquartile range (IQR)] of elastance from single compartment model (Elung) was 1.51 [0.72 - 2.76] cmH2O/ml, and elastance from recruitment basis function (E1) was 3.42[1.88 - 5.97] cmH2O/ml. Relative breath-to-breath variability (%Δe) was also compared, with median IQR %ΔElung of 0.22 [-9.73 - 12.06] and %Δe1 of -0.48[-9.28 - 10.34]. Elung is less sensitive than e1 to differences across infants where e1 was also less variable breath-to-breath. The parabolic model thus captured patient condition and the use of Edrs captured patient effort.