24 – hour oxygen saturation recordings at discharge in preterm infants

Abstract

Preterm infants have an immature respiratory system and therefore experience an increased number of respiratory pauses and oxygen desaturations. There is now advanced pulse oximetry technology that can record oxygen saturations every two seconds for extended periods of time. There has been insufficient literature that reports the incidence of intermittent hypoxia at time of discharge home from the neonatal unit in preterm infants using new generation oximeters. These respiratory events preterm infants experience have been shown to have some effect on neurodevelopment, but their effect on growth has not been investigated before. The primary aim of this study was to determine the prevalence of intermittent hypoxia in preterm infants at time of discharge home from the neonatal unit. The study also addresses the issue of artefact within oximetry recordings and compares results from automatically edited, manually edited and unedited oximetry data, as well as determining whether overnight 12-hour recordings are of equal value to full 24-hour recordings. The secondary aim of this study was to determine whether intermittent hypoxia at discharge has any effect on post discharge growth and to determine changes in amount of intermittent hypoxia from discharge oximetry to oximetry one-month post discharge.

We recruited preterm infants from the Wellington neonatal intensive care unit. A 24-hour pulse oximetry recording was performed immediately prior to the infant’s discharge home. These oximetry recordings were analysed and median values for measures reported from oximetry recordings were determined. Rules to manually edit oximetry data were created and applied to oximetry recordings. These manually edited reports were then compared with automatically edited and unedited reports. Each recording was also edited to resemble a 12-hour overnight recording and this was compared to the full 24-hour recording. Infants born less than 32 weeks gestational age were further followed up with weekly growth measurements for one-month. A repeat 24-hour oximetry recording was performed at one-month post discharge for these infants and compared to their discharge recording.

We report high rates of intermittent hypoxia in preterm infants at time of discharge home from the neonatal unit. For example the median 4% oxygen desaturation index (DSI 4%). was 57.9 events per hour. The incidence of these events decreased with advancing post-menstrual age. Rates of intermittent hypoxia one month post discharge were greatly decreased from discharge with improvements of 42% - 57% seen, with DSI 4% reducing to 25.5 events per hour. This study did not show a significant association between intermittent hypoxia and post-discharge growth, possibly because of the small sample size in this study subgroup.

There were few clinically relevant differences on reports edited manually compared with automatically edited reports, with some difference when compared to unedited reports. We recommend automatically editing oximetry reports as this gives similar results to manual editing for the majority of measures, however the nadir of the fall in oxygen saturation is often artefact, and even after automatic editing is the one measure that may remain false. The 24-hour oximetry reports were clinically similar to 12-hour recordings and therefore we suggest 12-hour oximetry studies are sufficient.