Infants will be observed in the neonatal intensive care unit for 4 hours. The observer will note the timing of oxygen saturation alarms, staff response, and interrupted staff activities.

False alarms may be particularly prevalent in the neonatal intensive care unit (NICU) where uncontrolled motion of infants can intensify the problem. The performance of pulse oximeters is of particular importance in the NICU because of the danger that both hyperoxemia and hypoxemia pose to newborns; hyperoxemia can lead to chronic lung disease or retinopathy of prematurity, and hypoxemia depresses spontaneous ventilation. High false alarm rates contribute to the problem of noise in the NICU. They also have the potential to endanger patients if clinicians become inured to the continual alarms and ignore some that may be clinically relevant. There have been few studies on the utility of pulse oximeter alarms in the NICU, particularly with the new-generation technology. This study will build on the small body of existing literature on alarm rates in new-generation pulse oximeters in neonates and provide details about the relationship of the alarms to clinical interventions.

This is an observational study of 50 infants at three hospitals in the United States. The observer will be an experienced nurse with comprehensive training to ensure consistency. The infants and clinical staff will be observed for four hours continuously. Observers will note the timing of alarms, response, interrupted clinical staff activities, and any clinical interventions, and timing of interventions. Clinical staff will also be questioned on whether the alarm was consistent with a desaturation event. Infant characteristics may affect the frequency of alarms. Data collected will include age, gender, weight, ethnicity, diagnosis, and medications.

This study will evaluate the proportion of nuisance alarms relative to the proportion of clinically relevant alarms. It will also evaluate the differences in alarm frequencies across infant characteristics and characterize nurse activities interrupted by the alarms.

Inclusion Criteria:

• Infants in the NICU continuously monitored by pulse oximetry

Exclusion Criteria:

• Infants whose legal guardians do not consent

neonates in the intensive care unit

The study oximeter's detection could have been 60 seconds before or after the recorded clinical intervention because of sampling bias. There were many desaturations that led to alarms that had no interventions.