Table 1 Primary outcome measures for observational sub-study

Sleep quality and overnight activity levels

Philips Actiwatch 2—Wrist Actigraph

In order to collect objective data on participant sleep quality, sleep fragmentation, total sleep and overall activity levels, we will use wrist actigraphy. Actigraphs are wearable sensors that allow logging of movement data and have been extensive used in clinical research. Actigraphy data has been validated against gold-standard polysomnography methods and offers a reliable tool for measuring sleep outside of a sleep laboratory environment [71].

Days 0, 3, 7 and 12, consisting of one initial interview and a maximum of three follow-up interviews, unless patient is discharged or moved to another room prior to day 12. Days 3, 7 and 12 data collection will occur ±1 day to accommodate for weekends and public holidays.

We will be using a Philips Actiwatch 2, which is a small, rugged, waterproof wrist worn data logger with long battery life and will provide us with a measure of rest-activity patterns and sleep. The Philips Actiwatch range has been applied in over 30 clinical trials to date including the study of sleep-wake patterns in older acute patients [72, 73].

Upon recruitment, the research officer will apply the Actiwatch on the participant’s non-dominant wrist and re-check application and wearing behaviour at every researcher-participant contact point thereafter.

Daytime sleepiness

Karolinska Sleepiness Scale (KSS) [74]

The KSS is a short 9-item self-report questionnaire that is a measure of a situational sleepiness. The KSS is sensitive to daily changes in levels of sleepiness [75].

As above.

Insomnia

Insomnia Severity Index (ISI) [76]

Insomnia is an important manifestation of sleep disturbance and thus an important construct to measure. The ISI is a brief validated 7-item self-report measure of the individual’s subjective perception of insomnia (sleep onset, maintenance and early and unintended waking) as well as amount of concern generated due to those symptoms. The ISI has been utilised in prior studies on insomnia prevalence in older admitted populations [77].

As above.