Objective: To apply the technique of respiratory deadspace measurement to consecutive infants with congenital diaphragmatic hernia, who were referred to our institution, in order to assess the efficiency of gas exchange.
Design: A cohort study evaluating the utility of deadspace measurements in neonates with congenital diaphragmatic hernia.
Setting: Tertiary care pediatric intensive care unit in a university hospital.
Patients: Thirty infants with congenital diaphragmatic hernia were studied on presentation to our institution, either before the institution of extracorporeal membrane oxygenation (ECMO) or after stabilization on ECMO.
Methods: The CO2 concentration of expired gas sampled at the exhaust port of the test ventilator was continuously measured and transformed to mixed expired CO2 by the following formula that corrects for compressible volume measured by the ventilatory pneumotachometer: mixed expired CO2 = (PCO2 in exhaust gas) x (ventilatory pneumotachometer minute volume)/(minute volume at the proximal airway). We then utilized the Bohr-Enghoff method to calculate the deadspace/tidal volume ratio: deadspace/tidal volume ratio = (PaCO2 - mixed expired PCO2)/PaCO2.
Measurements and main results: Deadspace/tidal volume ratio was calculated either before the institution of ECMO or during temporary separation from ECMO support as the patients demonstrated improvements in gas exchange and lung compliance. One hundred two measurements were made in 30 patients, with a mean of four measurements per patient (range 1 to 10). There was a significant (p = .005) difference between the first deadspace/tidal volume ratio measured, in survivors vs. nonsurvivors. The mean of the highest deadspace/tidal volume ratio in survivors was 0.47 compared with 0.62 in nonsurvivors (p = .003). A deadspace/tidal volume ratio of > or = 0.60 predicted mortality, with a positive predictive value of 80%, a negative predictive value of 79%, and an odds ratio of 15. The mean pre-ECMO deadspace/tidal volume ratio in those infants who ultimately required ECMO was significantly greater than the mean value for infants not requiring ECMO (0.65 vs. 0.43; p = .004). In patients who were treated with ECMO, survivors demonstrated a significant decrease in deadspace/tidal volume ratio during the course of ECMO. This decrease was not seen in the ECMO-treated patients who did not survive.
Conclusions: Predictors of outcome in infants with congenital diaphragmatic hernia have been complicated and contradictory, particularly in the ECMO era. We demonstrated that the respiratory deadspace can be easily quantified in these infants, and that a physiologic deadspace of > 0.60 is associated with a 15-fold increase in mortality rate. We also demonstrated that in those infants treated with ECMO, the survivors manifested a significant decrease in their deadspace/tidal volume ratio before ECMO decannulation.