Validation of actigraphy for determining sleep and wake in preterm infants

Abstract

Actigraphy has been widely used in adults and children for the determination of sleep and wake. However, there have been limited studies in infants and to date there have been no studies to validate the reliability of actigraphy in preterm infants. To evaluate the usefulness of actigraphy in preterm infants in a neonatal unit setting for determining sleep-wake by comparing results with those recorded from behavioural observations. Thirty-eight studies were carried out in 10 preterm infants (8M/2F) born at 29-34 weeks gestational age. Sleep-wake patterns were assessed over 24 h with behavioural observations and compared to actigraphy (Actiwatch AW64, Mini Mitter Company Inc., Sunriver, OR, USA). The studies were grouped into gestational ages 30-33 weeks (n = 8), 34-36 weeks (n = 20) and 37-40 weeks (n = 10). Overall, on the low-activity threshold we found agreement rates of 84.5-88.9% between actigraphy and behavioural scoring with the predictive value for determining sleep (PVS) being between 91.3% and 95.6% and sensitivity between 88.2% and 96.8%. However, the actiwatch was not reliable for determining wakefulness with low values for predictive value of wake (PVW,31.1-53.7%) and specificity (31.5-33.6%). Actigraphy can be used as a reliable indicator of sleep patterns in preterm infants in the neonatal unit setting.