Agitation Prevalence in People With Dementia in Australian Residential Aged Care Facilities: Findings From Machine Learning of Electronic Health Records

Developing an ontology of non-pharmacological treatment for emotional and mood disturbances in dementia

Emotional and mood disturbances are common in people with dementia. Non-pharmacological interventions are beneficial for managing these disturbances. However, effectively applying these interventions, particularly in the person-centred approach, is a complex and knowledge-intensive task. Healthcare professionals need the assistance of tools to obtain all relevant information that is often buried in a vast amount of clinical data to form a holistic understanding of the person for successfully applying non-pharmacological interventions. A machine-readable knowledge model, e.g., ontology, can codify the research evidence to underpin these tools. For the first time, this study aims to develop an ontology entitled Dementia-Related Emotional And Mood Disturbance Non-Pharmacological Treatment Ontology (DREAMDNPTO). DREAMDNPTO consists of 1258 unique classes (concepts) and 70 object properties that represent relationships between these classes. It meets the requirements and quality standards for biomedical ontology. As DREAMDNPTO provides a computerisable semantic representation of knowledge specific to non-pharmacological treatment for emotional and mood disturbances in dementia, it will facilitate the application of machine learning to this particular and important health domain of emotional and mood disturbance management for people with dementia.

Reliability and Validity of a Chinese Version of the Cohen–Mansfield Agitation Inventory-Short Form in Assessing Agitated Behavior

Background: Patients with dementia often present agitated behaviors. The Cohen–Mansfield Agitation Inventory-short form (CMAI-SF) is one of the most widely used instruments to evaluate agitated behaviors that affect patients’ quality of life and impose burden on caregivers. However, there is no simplified Chinese version of the CMAI-SF (C-CMAI-SF) in clinical settings. Purpose: This study aimed to develop a Chinese version of the C-CMAI-SF and examine its validity and reliability. Methods: This cross-sectional study included three phases. In Phase I, the original CMAI-SF was translated to Chinese. In Phase II, experts were invited to examine the content validity index (CVI). Phase III was conducted to test the validity and reliability of the C-CMAI-SF. Results: The scale showed good validity and reliability with a scale-level CVI of 0.89, Cronbach’s alpha (measure of internal consistency) of 0.874, and test–retest correlation coefficient of 0.902 (for 257 individuals). Using factor analysis, three factors were identified. Regarding concurrent validity, the C-CMAI-SF score was correlated with the Neuropsychiatric Inventory (agitation aggression subscale) and the Cornell Scale for Depression in Dementia (agitation subscale). Conclusions: The study demonstrated that the C-CMAI-SF is a valid and reliable instrument for evaluating agitated behaviors in people with dementia. Relevance to clinical practice: The C-CMAI-SF is an easy and quick tool used to identify and evaluate agitated behaviors in busy clinical settings.

Algorithmic harms and digital ageism in the use of surveillance technologies in nursing homes

Ageism has not been centered in scholarship on AI or algorithmic harms despite the ways in which older adults are both digitally marginalized and positioned as targets for surveillance technology and risk mitigation. In this translation paper, we put gerontology into conversation with scholarship on information and data technologies within critical disability, race, and feminist studies and explore algorithmic harms of surveillance technologies on older adults and care workers within nursing homes in the United States and Canada. We start by identifying the limitations of emerging scholarship and public discourse on "digital ageism" that is occupied with the inclusion and representation of older adults in AI or machine learning at the expense of more pressing questions. Focusing on the investment in these technologies in the context of COVID-19 in nursing homes, we draw from critical scholarship on information and data technologies to deeply understand how ageism is implicated in the systemic harms experienced by residents and workers when surveillance technologies are positioned as solutions. We then suggest generative pathways and point to various possible research agendas that could illuminate emergent algorithmic harms and their animating force within nursing homes. In the tradition of critical gerontology, ours is a project of bringing insights from gerontology and age studies to bear on broader work on automation and algorithmic decision-making systems for marginalized groups, and to bring that work to bear on gerontology. This paper illustrates specific ways in which important insights from critical race, disability and feminist studies helps us draw out the power of ageism as a rhetorical and analytical tool. We demonstrate why such engagement is necessary to realize gerontology's capacity to contribute to timely discourse on algorithmic harms and to elevate the issue of ageism for serious engagement across fields concerned with social and economic justice. We begin with nursing homes because they are an understudied, yet socially significant and timely setting in which to understand algorithmic harms. We hope this will contribute to broader efforts to understand and redress harms across sectors and marginalized collectives.