Trends in the use of non-invasive respiratory support for term infants in tertiary neonatal units in Australia and New Zealand

Abstract

OBJECTIVE: To determine whether the use of non-invasive respiratory support, such as continuous positive airway pressure and nasal high flow, to treat term infants in Australian and New Zealand tertiary neonatal intensive care units (NICUs) has changed over time, and if so, whether there are parallel changes in short-term respiratory morbidities. DESIGN: Retrospective database review of patient-level data from the Australian and New Zealand Neonatal Network (ANZNN) from 2010 to 2018. Denominator data on the number of term inborn livebirths in each facility was only available as annual totals. PATIENTS AND SETTING: Term, inborn infants cared for in NICUs within the ANZNN. MAIN OUTCOME MEASURES: The primary outcome was the annual change in hospital-specific rates of non-invasive respiratory support per 1000 inborn livebirths, expressed as a percentage change. Secondary outcomes were the change in rates of mechanical ventilation, pneumothorax requiring drainage, exogenous surfactant treatment and death before hospital discharge. RESULTS: A total of 14 656 term infants from 21 NICUs were included from 2010 to 2018, of whom 12 719 received non-invasive respiratory support. Non-invasive respiratory support use increased on average by 8.7% per year (95% CI: 7.9% to 9.4% per year); the number of term infants receiving non-invasive respiratory support almost doubled from 980 in 2010 (10.8/1000 livebirths) to 1913 in 2018 (20.8/1000). There was no change over time in rate of mechanical ventilation or death. The rate of pneumothorax requiring drainage increased over time, as did surfactant treatment. CONCLUSIONS: Non-invasive respiratory support use to treat term infants cared for in NICUs within the ANZNN is increasing over time. Clinicians should be diligent in selecting infants most likely to benefit from treatment with non-invasive respiratory support in this relatively low-risk population of term newborn infants. Analysis of patient-level data by individual NICUs is recommended to control for potential confounding due to changes in population over time.