Use of restraints in long-term residential care facilities in Hong Kong SAR, China: predisposing factors and comparison with other countries.

Abstract

To the Editor: Use of physical and chemical restraint may be considered to be an indicator of the quality of care in an institutional setting. In Hong Kong, 9% of the population aged 65 years and older reside in long-term care facilities. The quality of care in these facilities is only recently being reviewed, with a view to setting up regulatory guidelines. We examined the prevalence of and factors associated with the use of chemical and physical restraints among residential care homes for elderly people in China as a quality of care indicator in comparison with other countries. A total of 1820 of 1914 residents in 14 residential case institutions were successfully assessed (95% response rate) using the Minimum Data Set–Residential Assessment Instrument translated into Chinese (1). Participants were classified into four categories to enable comparison with other countries, ranging from those with minimal impairment (Low ADL [activities of daily living] 1 Low CP [cognitive performance]) to those highly dependent (High ADL 1 High CP), where high ADL represents an ADL score . 9 and high CP represents a cognitive performance scale score . 1 (2). The overall prevalence of chemical restraint was 11.4%. Bedside rails were used in 62.5% of residents, while other physical restraints were used in approximately 25% of residents, with trunk restraint being most commonly used, followed by limb restraint and chair restraint. In multivariate logistic regression, factors positively associated with physical restraint use were dependent ADL, impaired balance, dementia, need for artificial nutrition and hydration, and socially inappropriate behavior. Low staff number was not a significant factor. Dependency and impaired balance were not associated with increased risk of falls, suggesting that the perceived risk may be much higher than the actual risk, prompting use of physical restraints that may not be indicated. Adjusting for case mix, the use of physical restraint was