Use of Actigraphy to Evaluate Sleep in the ICU: A Systematic Review

SESSION TITLE: Critical Care - ICU Management SESSION TYPE: Original Investigation Poster PRESENTED ON: Wednesday, October 26, 2016 at 01:30 PM - 02:30 PM PURPOSE: Given recent guidelines recommending sleep promotion in critically ill patients, actigraphy has gained attention as a feasible and low-cost method to estimate sleep during large-scale ICU quality improvement and research efforts. This systematic review aimed to evaluate existing evidence regarding the use of actigraphy to evaluate sleep in the critically ill and aid in designing future interventions to improve sleep in the ICU. METHODS: We searched PubMed, EMBASE, CINAHL, Cochrane Library, and ProQuest from inception until September 2015. Citations were screened by two raters using pre-specified inclusion criteria. Included studies were assessed using established risk of bias instruments. RESULTS: Of 4,241 citations identified, 15 studies met eligibility criteria. These studies included 11 observational studies, 3 randomized controlled trials, and 1 nonrandomized controlled trial. Six (40%) studies evaluated patients in medical or combined medical-surgical ICUs, 5 (33%) in cardiothoracic surgical ICUs, 2 (13%) in coronary care units, 1 (7%) in a surgical ICU, and 1 (7%) in a burn unit. Six (40%) studies measured actigraphy in mechanically ventilated patients, while 3 (20%) enrolled no mechanically ventilated patients and 6 (40%) did not specify mechanical ventilation status. Nine (60%) studies enrolled ≤20 patients. Ten (67%) involved ICU actigraphy measurement for >24 hours. As measured using actigraphy, mean total night sleep time (measured over an 8 to 12 hour period) ranged from 2.4 to 7.8 hours, 24-hour sleep time from 4.3 to 10.3 hours, nocturnal awakenings from 1.4 to 58.5 per night, and sleep efficiency from 33% to 95%. Seven (47%) studies compared actigraphy with other quantitative measures of sleep (i.e., polysomnography and/or nurse/patient assessments) and demonstrated that total sleep time and sleep efficiency were higher with actigraphy versus these other methods. Additionally, actigraphy demonstrated fewer nighttime awakenings than polysomnography, but more awakenings than nurse or patient assessments. When used during ICU-based interventions involving experimental and control groups, actigraphy detected statistically significant differences in sleep measures in 2 of 4 (50%) studies; those observing differences enrolled more patients (mean = 53 versus 16) and collected actigraphy recordings for more nights (mean = 4 versus 3.5) than those that did not. CONCLUSIONS: Investigations involving actigraphy in the ICU involve heterogeneous study designs, report variable results regarding sleep duration, nighttime awakenings, and sleep efficiency, and suggest that actigraphy may overestimate sleep quantity compared to other measures. Since actigraphy has been used successfully to evaluate the effect of sleep-promoting interventions in the ICU, its greatest utility in this setting may be in large-scale interventional studies. CLINICAL IMPLICATIONS: Although actigraphy has been used to estimate sleep in the ICU, existing studies are heterogeneous and suggest that actigraphy may overestimate sleep duration in this setting. Additional studies are needed to further evaluate the utility of actigraphy to estimate sleep in the critically ill. DISCLOSURE: The following authors have nothing to disclose: Kristin Schwab, Bonnie Ronish, Dale Needham, Michael Ong, Jennifer Martin, Biren Kamdar No Product/Research Disclosure Information