A cost-effectiveness analysis of the application of nitric oxide versus oxygen gas for near-term newborns with respiratory failure: results from a Canadian randomized clinical trial.

OBJECTIVE

To conduct a cost-effectiveness analysis of the use of inhaled nitric oxide (NO) vs. oxygen administered to near-term (gestational age > or =34 wks) newborns with severe respiratory illness that were referred for consideration of extracorporeal membrane oxygenation (ECMO).

DESIGN

The cost-effectiveness analysis is based on outcome and utilization data from two multicentered randomized clinical trials conducted by the Canadian Inhaled Nitric Oxide Study group, one for patients with congenital diaphragmatic hernia (CDH) and one for patients without CDH. Data from the western Canadian ECMO center were used to establish costs.

SETTING

Patients were cared for in Canadian regional neonatal intensive care units, including two ECMO centers. Air transport was used for transporting patients between centers.

PATIENTS: Term and near-term newborns with severe respiratory illness who were receiving maximum conventional therapy and whose oxygenation index was >40.

INTERVENTIONS: Patients randomly received NO or oxygen. If their conditions deteriorated, they qualified for ECMO. Not all that qualified for ECMO received it because of individual parent/ physician preferences.

MEASUREMENTS AND MAIN

RESULTS

The cost-effectiveness ratio was the ratio of net cost (including neonatal intensive care, ECMO, and transport) to net outcome (survival) for the two interventions. For non-CDH cases, the cost-effectiveness ratio was $36,613 (Canadian) per life saved; the confidence intervals were wide and the results were not statistically significant. For CDH patients, the death rate was lower for oxygen and the oxygen patients cost less; the results were not statistically significant.

CONCLUSIONS

The small numbers of patients in the trials precluded significant results. Further, our results have a short-term time horizon (discharge to home or death). Thus, for non-CDH patients, the favorable ratio provides very qualified evidence in favor of NO.