Improving sleep on the inpatient general, non-stroke neurology service: A quasi-experimental interventional trial.

Abstract

Hospitalized patients often experience poor sleep, which is associated with worse health outcomes, increased rates of delirium, and readmissions. To improve overnight sleep for clinically stable general neurology patients at a single academic center. We conducted a quasi-experimental pre-post clinical trial of two sleep-protective interventions: (1) a sleep-friendly order set that reduced overnight interruptions by discontinuing vital sign checks/neurological examinations and retiming medications/blood draws; and (2) a "sleep menu" of comfort items. The primary outcome measure was sleep duration. Secondary outcomes were awakenings and patient responses to the Richards-Campbell Sleep Questionnaire. Safety outcomes were rates of delirium, ICU transfers, and 30-day readmissions. We compared the groups using linear mixed-effects models to account for repeated measures (multiple nights per participant). We analyzed data from 66 pre-intervention participants (238 nights) and 61 post-intervention participants (244 nights). The interventions did not increase sleep duration (pre-intervention median 5.5 h [interquartile range (IQR) 4.0,6.6], post-intervention median 5.4 h [4.1, 6.7], p = .84). There was marginal improvement in nightly awakenings (pre-intervention 2 [1, 4], post-intervention 2 [1, 3], p = .04). There were no significant differences in subjective sleep measures or safety outcomes. Targeted care team interruptions all significantly decreased post-intervention, though staff intrusions remained a common patient-reported barrier to sleep. Fragmented and reduced sleep was common in our general neurology population. Although our interventions did not improve sleep outcomes-apart from a modest drop in nocturnal awakenings-there was a significant decrease in care team interruptions. Poor inpatient sleep is multifactorial, and meaningful change likely requires comprehensive interventions.