NEONATAL SOCIETY ABSTRACTS
Hypercarbia during the first 72 hours of life is associated with intraventricular haemorrhage
Presented at the Neonatal Society 2007 Summer Meeting (programme).
Luyt K1, Baumer JH2
1 Peter Dunn NICU, St Michael's Hospital, Bristol, UK
2 Child Health Department, Derriford Hospital, Plymouth, UK
Background: Permissive hypercarbia is widely used to reduce lung injury in ventilated preterm infants. Most intraventricular haemorrhages (IVH) occur during the first 72 hours of life. Cerebral autoregulation is lost with progressive hypercarbia, thereby potentially increasing the risk of IVH (1). Hypercarbia, although associated with IVH in several published studies, could be a marker of severity of illness or of lung disease (2).
Objective: To determine whether hypercarbia during the first 72 hours of life is independently associated with IVH in preterm infants.
Design/Methods: 142 infants of ≤32 weeks gestation, enrolled at one centre in the International Trigger Ventilation Trial (Baumer JH. Arch Dis Child 2000;82:F5), were randomised to conventional or patient triggered ventilation. The study was undertaken in accordance with UK ethical guidelines. Baseline data collected included gestation, birth weight, gender, antenatal steroids, multiple gestation, delivery mode, surfactant and ethamsylate administration, and ventilation mode. During the first 72 hours the following data were collected: CRIB score (illness severity score), 4 hourly PaCO2, ventilator settings, coagulation studies and occurrence of pneumothorax. A ventilation index (PaCO2 x ventilation rate x peak inspiratory pressure/1000) was used as a composite proxy measurement for severity of lung disease. Cumulative hypercarbia above 60 mmHg (mmHg.hr) was calculated for each PaCO2 measured ((PaCO2-60mmHg) x measurement interval). Univariate and logistic regression analyses were performed. The p-value denoting statistical significance was adjusted to p≤0.003 to correct for multiple comparisons (Bonferroni correction).
Results: 133 infants had cranial ultrasounds and were included in the analysis. The dependent variable (IVH>1), was dichotomised as IVH>grade 1=1 and no IVH or IVH grade 1=0. Variables associated (p<0.05) with IVH in univariate analysis were: gestational age, birth weight, incomplete/no course of antenatal steroids, vaginal delivery, pneumothorax, abnormal coagulation, cumulative hypercarbia above 60mmHg and mean ventilation index. Significant variables were subjected to stepwise logistic regression analysis. In the multivariate model vaginal delivery and pneumothorax did not predict IVH. The model with best fit (Hosmer and Lemeshow, p=0.79) included gestation, incomplete/no course of antenatal steroids, CRIB score, cumulative hypercarbia above 60mmHg and mean ventilation index. The only significant predictors of IVH were gestational age (OR 0.60 per week of gestation (CI 0.41-0.88)), incomplete/no course of antenatal steroids (OR 3.81 (CI 1.11-13.03)) and cumulative hypercarbia above 60mmHg (OR 1.014 per mmHg.hour above 60mmHg (CI 1.005-1.023)).
Conclusions: Hypercarbia during the first 3 days of life is an independent predictor of moderate to severe IVH in ventilated preterm infants. Our logistic regression model adjusted for severity of illness and lung disease, factors not considered in previous publications.
1. Kaiser JR et al. Pediatr Res 2005;58:931-935
2. Kaiser JR et al J. Perinatol 2006;26(5):279-285