Abstract
ObjectiveSleep disruption occurs frequently in hospitalised patients. Given the potential of music intervention as a non-pharmacological measure to improve sleep quality, we aimed to assess and quantify current literature on the effect of recorded music interventions on sleep quality and quantity in the adult critical care and surgical populations.DesignSystematic review and meta-analysis.Data sourcesEmbase, MEDLINE Ovid, Cochrane Central, Web of Science and Google Scholar.Eligibility criteria for studiesRandomised controlled trials assessing the effect of music on sleep quality in critically ill and surgical patients.MethodsThe electronic databases were systematically searched from 1 January 1981 to 27 January 2020. Data were screened, extracted and appraised by two independent reviewers. Primary outcomes were sleep quality and quantity, assessed with validated tools. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed. Random effects meta-analysis was performed, and pooled standardised mean differences (SMDs) with 95% CIs were reported.ResultsFive studies (259 patients) were included in qualitative (risk of bias) and quantitative analysis (meta-analysis). Pooled data showed a significant effect of recorded music on subjective sleep quality in the critical care and surgical population (SMD=1.21 (95% CI 0.50 to 1.91), p<0.01, excluding one non-English study; SMD=0.87 (95% CI 0.45 to 1.29), p<0.01). The SMD of 1.21 corresponded to a 27.1% (95% CI 11.2 to 42.8) increase in subjective sleep quality using validated questionnaires. A significant increase in subjective sleep quantity of 36 min was found in one study. Objective measurements of sleep assessed in one study using polysomnography showed significant increase in deeper sleep stage in the music group.ConclusionsRecorded music showed a significant improvement in subjective sleep quality in some critical care and surgical populations. Therefore, its use may be relevant to improve sleep, but given the moderate potential for bias, further research is needed.PROSPERO registration numberCRD42020167783.