Agreement Between Actigraphy-Based Sleep Measures and Routine Nursing Sleep Assessments in Acute Psychiatric Inpatient Care

Introduction Sleep is routinely assessed by psychiatric mental health nurses in acute inpatient wards, where overnight observation forms part of everyday clinical care. These records are repeatedly collected, but observational assessment and activity-based monitoring capture different aspects of sleep and may not be interchangeable. Aim To examine the agreement between actigraphy-based total sleep time and routine nursing sleep assessments in an acute psychiatric inpatient ward. Method This cross-sectional observational study included 94 patients admitted to an acute psychiatric ward. Participants wore wrist actigraphy devices for up to seven available nights while sleep duration was concurrently documented by nursing staff as part of routine care. Pearson correlation coefficients assessed linear association and intraclass correlation coefficients (ICC) assessed agreement. Results Across all participants, nursing-assessed and actigraphy-based total sleep time showed a moderate linear association (Pearson r = 0.629) and good ICC (0.767). Bland-Altman analysis showed a mean bias of 0.20 hours, with 95% limits of agreement from -1.79 to 2.19 hours. Agreement varied across diagnostic groups and was strongest among patients with bipolar mania (r = 0.826; ICC = 0.886). Discussion Routine nursing sleep assessments showed little systematic bias relative to actigraphy at the group level, but wide limits of agreement indicate that the methods are not interchangeable for individual patients. Patient-reported sleep was not measured and should be included in future comparisons.