Background Premature newborns often experience oxygen saturations outside policy-specified targets, which may be associated with increased morbidity. Nurse workload may affect oxygen management.

Objective To examine the relationship between number of patients assigned to neonatal intensive care unit (NICU) nurses and achievement of oxygen saturation goals in premature newborns.

Design The authors linked nurse–patient assignment data with continuous oxygen saturation data for infants <29 weeks' gestation in a single NICU between January and June 2008. The proportion of time oxygen saturation was in policy-specified target range (85–92%) and proportion of time hyperoxaemic (98–100%) were determined for multiple 6 h monitoring periods. Each period was characterised by a single nurse, respiratory support mode and fraction of inspired oxygen (Fio₂) level (0.22–0.49 or ≥0.5). The nurse:patient ratio for the infant's nurse for each monitoring period was determined. Factors associated with Spo₂ target achievement and hyperoxaemia were identified.

Results The authors analysed 1019 monitoring periods from 14 infants with a mean (SD) birth weight of 860 (270) g and gestational age of 26.6 (1.6) weeks. The mean (range) postmenstrual age for all monitoring periods was 31.6 (24.1–40.7) weeks. Eighty-seven nurses provided care. In a multivariate cross-classified hierarchical regression, the nurse:patient ratio, postmenstrual age, respiratory support mode and Fio₂ were significantly associated with oxygen saturation outcomes. Fewer patients per nurse was significantly associated with a higher saturation target achievement among patients on high-frequency ventilation, and with reduced hyperoxaemia among patients on nasal cannula.

Conclusions Fewer patients per nurse may be associated with improved achievement of oxygen saturation goals and may be an important modifiable factor influencing oxygen-related outcomes in premature newborns. This effect may vary with mode of respiratory support.