Pulse oximetry in ventilated preterm newborns: reliability of detection of hyperoxaemia and hypoxaemia, and feasibility of alarm settings

Acta Paediatr. 1995 Jun;84(6):613-6. doi: 10.1111/j.1651-2227.1995.tb13709.x.


The aim of our study was to evaluate the reliability of pulse oximetry in detecting both hyper- and hypoxaemic states and to create clinically feasible alarm limits. A total of 792 readings of a pulse oximeter and corresponding values of arterial oxygen tension from 146 (79M, 67F) artificially ventilated preterm newborns with indwelling umbilical artery catheters were compared. Predictive value analysis of pulse oximeter readings related to arterial oxygen tension confirmed the ability of the pulse oximeter to identify both hypoxaemia and hyperoxaemia. However, a clinically feasible and safe range of alarm limits for maintenance of arterial oxygen tension of 40-90 mmHg (5.3-12 kPa) could only be established at a sensitivity level less than 0.9. At a level of 0.85, the alarm range on the pulse oximeter was 92.5-95%. Based on these findings, we are concerned about using pulse oximetry as the sole means of oxygen monitoring for preterm infants receiving supplementary oxygen. A combination of the pulse oximetry with other methods of blood-gas monitoring seems mandatory.

MeSH terms

  • Blood Gas Monitoring, Transcutaneous*
  • Humans
  • Hyperoxia / diagnosis*
  • Hypoxia / diagnosis*
  • Infant, Newborn
  • Infant, Premature
  • Infant, Premature, Diseases / diagnosis*
  • Monitoring, Physiologic
  • Prospective Studies
  • Reproducibility of Results
  • Respiration, Artificial