Assessment of actigraphy as a method to assess sleep quality following inpatient tonsillectomy in children

Abstract

Introduction: The aetiology of sleep disturbance in children following surgery is unclear, despite a reported incidence of 47% [1]. The Actiwatch, a small ‘wristwatch’ device, uses actigraphy, a technology based on the principle of movement detection, to assess sleep quality. It can be attached to a child and set to record details of movement at regular intervals. We sought to determine whether the Actiwatch is able to produce useful data for the assessment of sleep in inpatient children following tonsillectomy. Methods: Prior local ethics committee approval and informed parental consent were obtained. Fifteen ASA 1–2 children, aged 2–12 undergoing inpatient tonsillectomy were recruited. The Actiwatch monitor, programmed to record movement at 25 s intervals, was attached to each child after surgery and then retrieved the following morning. Additional data concerning operative details, drugs administered and adverse events were recorded and a sleep diary was provided for nursing staff and parents to record periods of sleep and wake overnight. Actigraph summary data was retrieved from each wristwatch the following day and details of asleep and awake periods recorded in the sleep diary were compared with actigraph data. Results: All parents approached agreed to take part, however one child refused to wear the ‘watch’ and was not included in the study. One further dataset was not collected due to a watch malfunction, leaving data from 13 children (five male, eight female) remaining. A total of 86 sleep assessments were made in the sleep diaries which, when compared with the actigraph data, gave a sensitivity and specificity of 82% and 90% respectively. The mean, range and standard deviation for overnight sleep parameters for all children were as follows: Total sleep time: 8 h 24 min (range 7:04–10:30 h), sd 70.7 min. Total awake time: 1 h 20 min (range 22 min–2:48 h), sd 41 min. Number of sleep bouts: 59.7 (range 33–109) sd 19.5. Sleep efficiency (% of total time spent asleep): 78.8% (range 69–95.8%), sd 8.8%. Discussion: Risk factors previously identified for poor postoperative sleep in the day case setting include pain, preoperative sleep patterns, parental anxiety and aggressive child behaviour [3]. However, there has been little work published using actigraphy in the inpatient setting to determine which factors are associated with poor sleep, such as the ward environment and perioperative medication. The actigraph is a small robust device that appears to be well tolerated by inpatient children, with only one refusal and one malfunction. We have illustrated a good correlation between observed and recorded sleep, indicating that the actigraph is indeed a useful tool for the assessment of postoperative sleep dysfunction in children. This correlates well with other authors [2] who have found a concordance rate of 85.3% compared with polysomnography in children. We have illustrated a large variation in the recorded sleep patterns between children in the inpatient setting. Both the range and standard deviations of the recorded parameters are large, implying that a direct comparative study of perioperative techniques, for example, would require large sample sizes to ensure adequate power. It is possible that more robust data may be obtained by recording control actigraphy data on a child preoperatively, such that a paired comparison can then be made. This approach will, however, be logistically more challenging as it will require multiple preoperative contacts with the child. Conclusions: We conclude that actigraphy is a practical and reasonably reliable tool for the assessment of sleep in children following tonsillectomy although we have also illustrated a large inter-patient variation of all measured sleep parameters.