The feasibility and reliability of actigraphy to monitor sleep in intensive care patients: an observational study

Crit Care. 2021 Jan 29;25(1):42. doi: 10.1186/s13054-020-03447-8.


Background: Sleep amongst intensive care patients is reduced and highly fragmented which may adversely impact on recovery. The current challenge for Intensive Care clinicians is identifying feasible and accurate assessments of sleep that can be widely implemented. The objective of this study was to investigate the feasibility and reliability of a minimally invasive sleep monitoring technique compared to the gold standard, polysomnography, for sleep monitoring.

Methods: Prospective observational study employing a within subject design in adult patients admitted to an Intensive Care Unit. Sleep monitoring was undertaken amongst minimally sedated patients via concurrent polysomnography and actigraphy monitoring over a 24-h duration to assess agreement between the two methods; total sleep time and wake time.

Results: We recruited 80 patients who were mechanically ventilated (24%) and non-ventilated (76%) within the intensive care unit. Sleep was found to be highly fragmented, composed of numerous sleep bouts and characterized by abnormal sleep architecture. Actigraphy was found to have a moderate level of overall agreement in identifying sleep and wake states with polysomnography (69.4%; K = 0.386, p < 0.05) in an epoch by epoch analysis, with a moderate level of sensitivity (65.5%) and specificity (76.1%). Monitoring accuracy via actigraphy was improved amongst non-ventilated patients (specificity 83.7%; sensitivity 56.7%). Actigraphy was found to have a moderate correlation with polysomnography reported total sleep time (r = 0.359, p < 0.05) and wakefulness (r = 0.371, p < 0.05). Bland-Altman plots indicated that sleep was underestimated by actigraphy, with wakeful states overestimated.

Conclusions: Actigraphy was easy and safe to use, provided moderate level of agreement with polysomnography in distinguishing between sleep and wakeful states, and may be a reasonable alternative to measure sleep in intensive care patients. Clinical Trial Registration number ACTRN12615000945527 (Registered 9/9/2015).

Keywords: Actigraphy; Intensive care; Polysomnography; Sleep; Sleep disturbance; Sleep quality.

Publication types

  • Observational Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Actigraphy / methods*
  • Actigraphy / standards*
  • Actigraphy / statistics & numerical data
  • Adult
  • Aged
  • Feasibility Studies
  • Female
  • Humans
  • Intensive Care Units / organization & administration
  • Intensive Care Units / statistics & numerical data
  • Male
  • Middle Aged
  • Polysomnography / methods
  • Polysomnography / standards*
  • Polysomnography / statistics & numerical data
  • Prospective Studies
  • Reproducibility of Results

Associated data

  • ANZCTR/ACTRN12615000945527