Identifying Older People at Risk of Abuse During Routine Screening Practices
Published in: Journal of the American Geriatrics Society, v. 51, no. 1, Jan. 2003, p. 24-31
Posted on RAND.org on December 31, 2002
OBJECTIVES: To examine the association between various characteristics of community-based older people and a constructed measure of potential elder abuse. DESIGN: Cross-sectional design. SETTING: Public community-based long-term care programs in Michigan. PARTICIPANTS: Individuals aged 60 and older seeking home and community-based services in Michigan between November 1996 and October 1997 (N = 701). MEASUREMENTS: Data were collected using the Minimum Data Set for Home Care (MDS-HC) assessment. The dependent variable is a constructed measure of potential elder abuse reflecting physical and emotional abuse and neglect. Independent variables include demographic characteristics; diagnoses; behavioral measures; and cognitive, physical, and social functioning. RESULTS: Several measures of social support and social function were strongly associated with the signs of a potentially abusive environment: brittle support (odds ratio (OR) = 3.5, 90% confidence interval (CI) = 1.5-8.1), older person feels lonely (OR = 2.4, 90% CI = 1.3-4.5), and older person expresses conflict with family/friends (OR = 2.3, 90% CI = 1.2-4.3). Home care participants' alcohol abuse, psychiatric illness, lack of ease interacting with others, and short-term memory problems were also significantly associated with the signs of potential elder abuse. CONCLUSIONS: The results of this study suggest that the signs of potential elder abuse are associated with a diminishing social network and poor social functioning, although some characteristics of the older person's health are contributing factors. Improved understanding of the link between those characteristics and potential abuse will help healthcare providers, case managers, and others identify older people at high risk of abuse.