Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Am J Crit Care. Author manuscript; available in PMC 2018 January 1.
Published in final edited form as:
PMCID: PMC5629184

Feasibility of Continuous Actigraphy in Patients in a Medical Intensive Care Unit



Poor sleep and immobility are common in patients in the medical intensive care unit (MICU) and are associated with adverse outcomes. Interventions to promote sleep and mobilization in the MICU are gaining popularity, but feasible instruments to measure their effectiveness are lacking. Actigraphy may be useful for large-scale, continuous measurement of sleep and activity, but its feasibility in MICU patients has not been rigorously evaluated.


To evaluate the feasibility of continuous actigraphy measurement in consecutive MICU patients.


Wrist and ankle actigraphy data were collected for 48 hours in consenting MICU patients. Actigraphy-based measures of estimated sleep and activity were summarized by using descriptive statistics. Agreement between wrist and ankle measurements was evaluated using Cohen κ statistics (for sleep quantity) and intraclass correlation coefficients (for activity).


Overall, 35 of 48 (73%) eligible patients were enrolled, including 10 requiring mechanical ventilation. Of these patients, 34 (97%) completed the 48-hour actigraphy period; 20 (57%) found the devices comfortable. Wrist devices logged a mean (SD) of 33.4 (8.8) hours of estimated sleep (72% [19%] of recording period) and 19.6 (17.2) movements per 30-second epoch. Ankle devices recorded 43.2 (4.1) hours of estimated sleep (93% [7%] of recording period) and 5.1 (6.0) movements per 30 seconds.


Uninterrupted actigraphy is feasible and generally well tolerated by MICU patients and may be considered for future large-scale studies. Wrist and ankle actigraphy measurements of sleep and activity in this setting agree poorly and cannot be used interchangeably.

Poor sleep and immobility are common in the medical intensive care unit (MICU)18 and are risk factors for delirium and long-term physical impairments.911 Hence, as part of efforts to improve patients’ outcomes, sleep and mobility promotion have gained particular attention and are therefore recommended in recent clinical practice guidelines.12

One barrier to ICU-based sleep and mobility promotion efforts is a lack of feasible tools for measuring sleep and activity.13 Polysomnography has demonstrated that sleep in critically ill patients is fragmented, short in duration, and frequently occurs during daytime hours2,3; however, its widespread use in ICU populations is challenging because of the cumbersome equipment, the prohibitive cost, and the need for expert interpretation of atypical tracings of unclear significance.3 For these reasons, polysomnography is not feasible to use throughout a patient’s ICU stay.14 In the area of mobilization, a recently developed ICU mobility scale provides an ordinal measure of activity levels.15 However, the scale provides only the single highest level of mobility over the period of observation (eg, 12 or 24 hours) rather than continuous recordings.

Sleep in ICU patients is fragmented, brief, and often occurs during daytime hours.

As an alternative, actigraphy devices have been demonstrated to be comfortable, affordable, practical, and feasible for continuous, long-term recording of sleep and activity in both research and clinical settings.16 Additionally, these devices have been used to evaluate sleep and activity in ICU-based studies1724 and may be useful in large-scale interventions. However, the feasibility of actigraphy has not been evaluated in a heterogeneous population of critically ill patients. Hence, our objective was to assess the feasibility of actigraphy in a population of MICU patients.


This prospective observational study was done to assess the feasibility of 48-hour continuous wrist and ankle actigraphy in consecutively enrolled MICU patients. All patients or their surrogates provided informed consent, and the institutional review board approved this study.

Setting and Participants

Our MICU has 24 private rooms and a nurse to patient ratio of 1 to 2. Bedside nursing staff work 7 am to 7 pm and 7 pm to 7 am shifts. Routine daily blood sampling, radiology studies, and bathing occur primarily during the 7 pm to 7 am shift.

We identified eligible patients from systematic daily screening of the MICU census and electronic medical records. We enrolled patients aged 18 years and older who were being cared for by the MICU team. Exclusion criteria included (1) previous study enrollment, (2) expected ICU stay of less than 24 hours from the time of enrollment, (3) neither a wrist nor an ankle available for actigraphy placement (eg, because of amputations, deformities, or placement of medical devices), (4) anticipated sterile procedure requiring device removal during the 48-hour actigraphy monitoring period, (5) pending transfer to a hospital general inpatient area or outside facility, (6) inability to provide informed consent or no surrogate present to provide informed consent on the patient’s behalf, (7) non–English-speaking patient, and (8) moribund or palliative status.

Actigraph Evaluation

We used the Actiwatch Spectrum (Philips-Respironics) because of its widespread use and previous validation for monitoring sleep and activity.25,26 Its compact size and light weight (16 g, compared with 21 g or more for most other actigraph devices) were additional considerations for its use in critically ill patients in the ICU.

Actigraph Setup and Removal

Actigraphy was started at 12 noon (or soon after) on the first day of recording, with most patients receiving 2 devices: 1 on the dominant wrist and 1 on the dominant ankle. Nondominant locations were used when dominant sides were unavailable (eg, because of medical devices). Patients in whom both wrists or both ankles were unavailable received only 1 device on an available extremity. Consistent with prior ICU-based studies,27 the devices were programmed to log activity levels across discrete 30-second epochs. Devices were removed after approximately 48 hours, after which data were downloaded using Actiware software. At device removal, patients rated the device as “comfortable, barely noticed,” “moderately comfortable,” or “very uncomfortable.”

Data Collection

In addition to actigraphy data, research staff recorded each patient’s age, gender, race, ICU admission diagnosis, organ failure status (evaluated using daily Sequential Organ Failure Assessment score28), and daily mechanical ventilation status from the medical record. Patients or proxies also reported baseline sleep quality and sleep problems (from questions adapted from the Pittsburgh Sleep Quality Index29) and activity levels (adapted from prior publications15,30,31).

Statistical Analysis

Demographic and clinical data were summarized using median and interquartile range for continuous variables and proportions for categorical variables. Actigraphy data were summarized using mean and standard deviation. Raw actigraphy data analyzed included activity levels (a continuous variable of the number of movements per epoch) and sleep versus wake, which was assigned as a binary variable for each epoch using an established scoring algorithm within the Actiware software. The agreement between wrist and ankle readings for sleep versus wake was calculated using the Cohen κ statistic. Additionally, the agreement of wrist and ankle activity levels was evaluated using intraclass correlation coefficients (ICCs), estimated by using linear mixed-effects models that clustered activity levels by patient and by epoch nested within each patient. Subgroup analyses were performed in particular patients’ epochs to characterize agreement at different points of the study. A modified Bland-Altman plot was produced to visualize patterns of agreement between wrist and ankle activity levels. Descriptive analyses were performed using Stata version 14.0 (StataCorp). The ICCs and Cohen κ statistics were calculated using SAS version 9.4 (SAS Institute Inc).

Finally, a sample size of 35 patients was calculated to achieve a feasibility proportion of 90%, with a 95% CI of plus or minus 10%.



Of 135 consecutive MICU patients screened from November 2014 to January 2015, 48 (36%) met eligibility criteria, of whom 35 (73%) provided informed consent to participate (Figure 1). Enrolled patients had a median age of 60 (interquartile range, 45–70) years, 17 (49%) were female, and 5 (14%) reported a history of sleep disorders (Table 1). These patients were admitted primarily for respiratory failure (40%), with 10 (29%) patients receiving mechanical ventilation and 10 (29%) receiving sedative infusions during their enrollment period.

Figure 1
Patient flow diagram.
Table 1
Characteristics of the 35 patients in the study

Actigraph Recording

Overall, 34 wrist and 34 ankle actigraph recordings were initiated: 33 in patients who received actigraphs on both a wrist and an ankle and 1 of each in patients who received only a single actigraph on either a wrist or ankle (because of wound dressings that precluded placement of a second actigraph). Of 35 enrolled patients, 34 completed the 48-hour actigraph recording period with at least 1 device in place, yielding 189 595 wrist and 189 607 ankle epochs for analysis (Table 2). One patient (3%) completed 34.2 of 48 hours (71% target recording time) because of serial magnetic resonance imaging scans requiring actigraph removal. Regarding device comfort, 20 patients (57%) rated the devices as “comfortable, barely noticed” or “moderately comfortable”; 6 (17%) rated them as “very uncomfortable”; and 8 (23%) were unable to respond (ie, because of delirium or coma).

Table 2
Actigraph data summary (N = 35)

Sleep Recordings

At a medium threshold setting for classifying epochs as sleep or wake, the wrist and ankle actigraphs logged a mean (SD) of 33.4 (8.8) and 43.2 (4.1) hours of estimated sleep, respectively, accounting for 72% (19%) and 93% (7%), respectively, of each patient’s total recording time. During the 10 pm to 6 am nighttime period, sleep accounted for 80% (14%) and 95% (6%) of the recording period for wrist and ankle, respectively. At the low, medium, and high thresholds for estimated sleep, the κ statistic for agreement of wrist and ankle was from 0.12 to 0.34 (Table 3).

Table 3
Agreement of wrist versus ankle actigraphy during 48-hour sleep measurement, using κ statistica

Activity Recordings

During the recording period, mean (SD) wrist and ankle activity counts totaled 19.6 (17.2) and 5.1 (6.0) units per 30-second epoch, respectively, with maximum levels of 1418 and 1922 units, respectively (Table 2, Figure 2). Activity counts equaled 0 during 122 259 (64%) and 157 795 (83%) wrist and ankle epochs, respectively.

Figure 2
Actigrams depicting 24-hour activity recordings as measured using wrist actigraphy. Figure 2A (top panel) depicts a healthy adult. Figures 2B (middle panel) and 2C (bottom panel) depict activity levels averaged by epoch for the 34 wrist actigraphy devices ...

Among 183 878 paired wrist and ankle epochs registered by 33 patients, wrist activity levels exceeded ankle levels 58 911 (32%) times and ankle activity levels exceeded wrist levels 14 745 (8%) times. Wrist and ankle activity levels were nonzero but equal during 518 epochs (0.3%) and equaled 0 during 109 704 epochs (60%; Figure 3). Epoch-by-epoch wrist-versus-ankle ICCs were 0.241, 0.246, 0.231, and 0.234 for 48-hour, 7 pm to 7 am, 7 am to 7 pm, and 10 pm to 6 am recording periods, respectively.

Figure 3
Modified Bland-Altman plot of 183 878 paired wrist-ankle activity levels from actigraph recordings in 33 critically ill patients shows the relationship between wrist activity (x axis) and the difference in wrist and ankle activity levels (y axis). Each ...


This study demonstrated that continuous actigraphy in consecutively enrolled MICU patients was feasible, as 34 of 35 patients (97%) completed the 48-hour actigraphy recording period, including 33 who wore both wrist and ankle actigraphs. Consistent with prior research,3234 patients’ activity levels were low, with sleep estimates totaling greater than two-thirds of the recording time and with 64% and 83% of 30-second wrist and ankle epochs, respectively, logging zero activity. Wrist-versus-ankle correlation and agreement of activity and sleep levels were poor. Compared with wrist actigraphs, ankle actigraphs logged more zeroes, generally lower activity levels, and higher estimated sleep totals.

This study was motivated in part by recent ICU-based sleep and early rehabilitation studies that demonstrated the benefits of these interventions but were limited by a lack of practical large-scale continuous measures of sleep and mobility.31,35,36 Given that actigraphy has been used in prior ICU interventional studies to demonstrate improvements in sleep and activity17,18 and in observational studies to estimate sleep,1923 activity,22,24 sedation,20,37 and delirium,38 the use of actigraphy during future sleep and rehabilitation interventional studies seems logical. However, prior ICU-based evaluations of actigraphy were limited in scope and generalizability because of enrollment of convenience samples,19,37,39 small sample sizes,17,20,32,40 recording times of 24 hours or less,19,32,37 and inclusion of only low-or high-acuity patients or exclusively surgical ICU patients.2022 Although some studies documented no complications involving actigraphy in critically ill patients,20,22,23 this study is unique in its evaluation of day-to-day feasibility of actigraphy in a busy ICU setting across a heterogeneous spectrum of patients receiving care.

We found that actigraphy was feasible in a heterogeneous population of MICU patients whose organ failure scores paralleled those of other critically ill populations.41 Additionally, patients tolerated actigraphy well, as only 1 patient’s device was removed by staff (because of a magnetic resonance imaging study). Notably, however, only 36% of patients met the basic eligibility criteria for this study, and 27% declined enrollment despite the minimal risk and short duration of study participation, suggesting the need for careful evaluation of eligibility criteria, consent rates, and procedures for future research.

Wrist actigraphy may be a feasible method for quantifying changes in sleep and activity in ICU patients.

Finally, we performed ankle actigraphy with the understanding that wrist placement may not be feasible in some patients with intravenous and intra-arterial catheters, restraints, wounds, or anticipated procedures involving the upper extremities. Although both wrist and ankle actigraphy were well tolerated by patients, ankle activity levels equaled 0 (ie, no movement was detected) more often than did wrist activity levels (83% versus 64% of epochs), thus yielding longer periods of inactivity interpreted as estimated sleep. Hence, correlation and agreement of wrist-versus-ankle activity and sleep measures were both poor. We identified only 1 prior investigation of 20 medical and cardiovascular ICU patients that demonstrated a higher correlation (ρ = 0.69) between wrist and ankle actigraphy; however, that study used a short 2-hour measurement period, which may have resulted in overestimation of the correspondence between the 2 measures.39 Although we did not assess the validity or superiority of wrist or ankle actigraph recordings, given American Academy of Sleep Medicine guidelines recommending wrist over ankle actigraphy recordings42 and the fact that patients in our study generally tolerated wrist actigraphy, wrist placement should be the preferred mode of actigraphy measurement in ICU patients.

Strengths of our study include enrollment of consecutive patients, epoch-by-epoch analysis, and comparison of wrist and ankle placements. A limitation was the relatively small sample of patients who were all studied in a single ICU. This may reduce the generalizability of our findings to other institutions. Additionally, we estimated sleep duration with a software-based algorithm that has not been validated in critically ill populations. By using an algorithm for scoring sleep and wake in ambulatory adults, it is possible that critically ill patients were incorrectly scored as sleeping when they were awake but immobile because of weakness, restraints, sedation, or severe illness. In future research in critically ill populations, ICU-specific actigraphy interpretation algorithms should be developed to address this potential limitation.

In conclusion, continuous actigraphy monitoring for 48 hours was feasible and well tolerated by MICU patients. Ankle actigraphy was well tolerated but logged substantially more inactivity and yielded higher estimates of sleep duration than wrist actigraphy, which is the standard measurement method. Given its ease and low cost of use, wrist actigraphy may be a feasible method for quantifying changes in sleep and mobility in future larger-scale efforts to evaluate interventions designed to improve sleep and/or mobilization in the ICU setting.


We thank Prerna Gupta, Nikita Mathew, and Chloe Krasnoff for their assistance with data collection, entry, and quality assurance.


B.B.K. is supported by a grant through the UCLA Clinical Translational Research Institute (CTSI) and the National Institutes of Health/National Center for Advancing Translational Sciences (UL1TR000124, UL1TR001881).


To purchase electronic or print reprints, contact American Association of Critical-Care Nurses, 101 Columbia, Aliso Viejo, CA 92656. Phone, (800) 899-1712 or (949) 362-2050 (ext 532); fax, (949) 362-2049; gro.ncaa@stnirper


1. Kamdar BB, Needham DM, Collop NA. Sleep deprivation in critical illness: its role in physical and psychological recovery. J Intensive Care Med. 2012;27(2):97–111. [PMC free article] [PubMed]
2. Freedman NS, Gazendam J, Levan L, Pack AI, Schwab RJ. Abnormal sleep/wake cycles and the effect of environmental noise on sleep disruption in the intensive care unit. Am J Respir Crit Care Med. 2001;163(2):451–457. [PubMed]
3. Watson PL, Pandharipande P, Gehlbach BK, et al. Atypical sleep in ventilated patients: empirical electroencephalography findings and the path toward revised ICU sleep scoring criteria. Crit Care Med. 2013;41(8):1958–1967. [PMC free article] [PubMed]
4. Richards KC, Bairnsfather L. A description of night sleep patterns in the critical care unit. Heart Lung. 1988;17(1):35–42. [PubMed]
5. Aaron JN, Carlisle CC, Carskadon MA, Meyer TJ, Hill NS, Millman RP. Environmental noise as a cause of sleep disruption in an intermediate respiratory care unit. Sleep. 1996;19(9):707–710. [PubMed]
6. Helton MC, Gordon SH, Nunnery SL. The correlation between sleep deprivation and the intensive care unit syndrome. Heart Lung. 1980;9(3):464–468. [PubMed]
7. Hilton BA. Quantity and quality of patients’ sleep and sleep-disturbing factors in a respiratory intensive care unit. J Adv Nurs. 1976;1(6):453–468. [PubMed]
8. Brower RG. Consequences of bed rest. Crit Care Med. 2009;37(10 Suppl):S422–428. [PubMed]
9. Figueroa-Ramos MI, Arroyo-Novoa CM, Lee KA, Padilla G, Puntillo KA. Sleep and delirium in ICU patients: a review of mechanisms and manifestations. Intensive Care Med. 2009;35(5):781–795. [PubMed]
10. Trompeo AC, Vidi Y, Locane MD, et al. Sleep disturbances in the critically ill patients: role of delirium and sedative agents. Minerva Anestesiol. 2011;77(6):604–612. [PubMed]
11. Fan E, Dowdy DW, Colantuoni E, et al. Physical complications in acute lung injury survivors: a two-year longitudinal prospective study. Crit Care Med. 2014;42(4):849–859. [PMC free article] [PubMed]
12. Barr J, Fraser GL, Puntillo K, et al. American College of Critical Care Medicine. Clinical practice guidelines for the management of pain, agitation, and delirium in adult patients in the intensive care unit. Crit Care Med. 2013;41(1):263–306. [PubMed]
13. Watson PL. Measuring sleep in critically ill patients: beware the pitfalls. Crit Care. 2007;11(4):159. [PMC free article] [PubMed]
14. Knauert MP, Yaggi HK, Redeker NS, Murphy TE, Araujo KL, Pisani MA. Feasibility study of unattended polysomnography in medical intensive care unit patients. Heart Lung. 2014;43(5):445–452. [PMC free article] [PubMed]
15. Hodgson C, Needham D, Haines K, et al. Feasibility and inter-rater reliability of the ICU Mobility Scale. Heart Lung. 2014;43(1):19–24. [PubMed]
16. Martin JL, Hakim AD. Wrist actigraphy. Chest. 2011;139(6):1514–1527. [PubMed]
17. Shilo L, Dagan Y, Smorjik Y, et al. Effect of melatonin on sleep quality of COPD intensive care patients: a pilot study. Chronobiol Int. 2000;17(1):71–76. [PubMed]
18. Taguchi T, Yano M, Kido Y. Influence of bright light therapy on postoperative patients: a pilot study. Intensive Crit Care Nurs. 2007;23(5):289–297. [PubMed]
19. Kroon K, West S. ‘Appears to have slept well’: assessing sleep in an acute care setting. Contemp Nurse. 2000;9(3–4):284–294. [PubMed]
20. Mistraletti G, Taverna M, Sabbatini G, et al. Actigraphic monitoring in critically ill patients: preliminary results toward an "observation-guided sedation". J Crit Care. 2009;24(4):563–567. [PubMed]
21. Chen JH, Chao YH, Lu SF, Shiung TF, Chao YF. The effectiveness of valerian acupressure on the sleep of ICU patients: a randomized clinical trial. Int J Nurs Stud. 2012;49(8):913–920. [PubMed]
22. Yilmaz H, Iskesen I. Objective and subjective characteristics of sleep after coronary artery bypass graft surgery in the early period: a prospective study with healthy subjects. Heart Surg Forum. 2007;10(1):E16–20. [PubMed]
23. Bourne RS, Mills GH, Minelli C. Melatonin therapy to improve nocturnal sleep in critically ill patients: encouraging results from a small randomised controlled trial. Crit Care. 2008;12(2):R52. [PMC free article] [PubMed]
24. Verceles AC, Hager ER. Use of accelerometry to monitor physical activity in critically ill subjects: a systematic review. Respir Care. 2015;60(9):1330–1336. [PMC free article] [PubMed]
25. Weiss AR, Johnson NL, Berger NA, Redline S. Validity of activity-based devices to estimate sleep. J Clin Sleep Med. 2010;6(4):336–342. [PubMed]
26. Hyde M, O’Driscoll DM, Binette S, et al. Validation of actigraphy for determining sleep and wake in children with sleep disordered breathing. J Sleep Res. 2007;16(2):213–216. [PubMed]
27. Bourne RS, Minelli C, Mills GH, Kandler R. Clinical review: sleep measurement in critical care patients: research and clinical implications. Crit Care. 2007;11(4):226. [PMC free article] [PubMed]
28. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):707–710. [PubMed]
29. Buysse DJ, Reynolds CF, 3rd, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193–213. [PubMed]
30. Bailey P, Thomsen GE, Spuhler VJ, et al. Early activity is feasible and safe in respiratory failure patients. Crit Care Med. 2007;35(1):139–145. [PubMed]
31. Needham DM, Korupolu R, Zanni JM, et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542. [PubMed]
32. van der Kooi AW, Tulen JH, van Eijk MM, et al. Sleep monitoring by actigraphy in short-stay ICU patients. Crit Care Nurs Q. 2013;36(2):169–173. [PubMed]
33. Winkelman C. Investigating activity in hospitalized patients with chronic obstructive pulmonary disease: a pilot study. Heart Lung. 2010;39(4):319–330. [PMC free article] [PubMed]
34. Winkelman C, Higgins PA, Chen YJ. Activity in the chronically critically ill. Dimens Crit Care Nurs. 2005;24(6):281–290. [PMC free article] [PubMed]
35. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882. [PubMed]
36. Kamdar BB, King LM, Collop NA, et al. The effect of a quality improvement intervention on perceived sleep quality and cognition in a medical ICU. Crit Care Med. 2013;41(3):800–809. [PMC free article] [PubMed]
37. Raj R, Ussavarungsi K, Nugent K. Accelerometer-based devices can be used to monitor sedation/agitation in the intensive care unit. J Crit Care. 2014;29(5):748–752. [PubMed]
38. Osse RJ, Tulen JH, Hengeveld MW, Bogers AJ. Screening methods for delirium: early diagnosis by means of objective quantification of motor activity patterns using wrist-actigraphy. Interact Cardiovasc Thorac Surg. 2009;8(3):344–348. discussion 348. [PubMed]
39. Grap MJ, Borchers CT, Munro CL, Elswick RK, Jr, Sessler CN. Actigraphy in the critically ill: correlation with activity, agitation, and sedation. Am J Crit Care. 2005;14(1):52–60. [PubMed]
40. Shilo L, Dagan Y, Smorjik Y, et al. Patients in the intensive care unit suffer from severe lack of sleep associated with loss of normal melatonin secretion pattern. Am J Med Sci. 1999;317(5):278–281. [PubMed]
41. Ruhl AP, Lord RK, Panek JA, et al. Health care resource use and costs of two-year survivors of acute lung injury. An observational cohort study. Ann Am Thorac Soc. 2015;12(3):392–401. [PMC free article] [PubMed]
42. Ancoli-Israel S, Cole R, Alessi C, Chambers M, Moorcroft W, Pollak CP. The role of actigraphy in the study of sleep and circadian rhythms. Sleep. 2003;26(3):342–392. [PubMed]