LAUSR

AIMS AND OBJECTIVES This study aimed to examine the association between physical restraint duration and undesirable outcomes amongst inpatients comorbid with dementia and pneumonia in acute care hospitals. BACKGROUND Physical restraints are frequently used in the management of patients, especially amongst patients with dementia. No previous study investigated the potential undesirable effects of physical restraints in patients with dementia. METHODS This was a cohort study using a nationwide discharge abstract database in Japan. Patients aged ≥65 years with dementia hospitalised for pneumonia or aspiration pneumonia between April 1, 2016 and March 31, 2019 were identified. The exposure was physical restraint. The primary outcome was hospital discharge to the community. Secondary outcomes included hospitalisation costs, functional decline, in-hospital mortality, and institutionalisation for long-term care. RESULTS A total of 18,255 inpatients with pneumonia and dementia in 307 hospitals were included in this study. Of them, 21.5% and 23.7% had physical restraint during full and partial days of hospital stays, respectively. Discharge to the community incidence rates was lower in the full-restraint vs. the no-restraint group (27 vs. 29 per 1000 person-days; HR, 1.05 [95% CI, 1.01-1.10]) and the partial-restraint vs. the no-restraint group (17 vs. 29 per 1000 person-days; HR, 1.79 [95% CI, 1.71-1.87]). The risks of functional decline were higher in the full-restraint vs. the no-restraint group (27.8% vs. 20.8%; RR, 1.33 [95% CI, 1.22, 1.46]) and the partial-restraint vs. the no-restraint group (29.2% vs. 20.8%; RR, 1.40 [95% CI, 1.29, 1.53]). CONCLUSIONS The use of physical restraints was associated with a lower incidence rate of discharge to the community and an increased risk of functional decline at discharge. Further research is needed to judge the benefit-risk balance of physical restraints in acute care settings. RELEVANCE TO CLINICAL PRACTICE Understanding the risk of physical restraints allows medical staff to improve the process of decision making in everyday practice. No Patient or Public Contribution. REPORTING METHODS The reporting of this article conforms to the STROBE statement.