JMIR Publications

ABSTRACT

Background:

Informal caregivers make up a large share of the care provided to people living with dementia (PLwD), and experience an elevated risk for anxiety, depression and caregiving burden. Technology-based interventions, such as mobile applications, aim to support informal caregivers by providing information, training, and mental or social support. Considering the diverse nature of the caregiver experience, Artificial Intelligence (AI) is being incorporated into such interventions to create more tailored and specific support; however, AI hesitancy may serve as a significant barrier to utilization.

Objective:

The study aimed to preliminarily assess the level of AI acceptance in adult informal caregivers of PLwD and examine psychological and contextual factors as potential determinants of AI acceptance.

Methods:

Adult, unpaid caregivers were recruited through social media and community partners. A cross-sectional, web-based survey was administered via Qualtrics. AI acceptance was measured using the Attitude Towards Artificial Intelligence (ATTARI-12) scale, while psychological flexibility (PF) was measured using the Personalized Psychological Flexibility Index (PPFI). Social needs were assessed with the Accountable Health Communities Health Related Social Needs (AHC-HRSN) tool. Descriptive, correlational and regression analyses were performed to examine the associations between these factors and AI acceptance.

Results:

Overall, 31 informal caregivers of PLwD completed the survey. With an average age of 60 years old (SD 10.6), the majority of caregivers were female (29/31, 94%), Caucasian (25/31, 81%), highly educated (24/31, 77% completed some form of higher education) and currently serving in a caregiver role (21/31, 68%). Mean PPFI and ATTARI-12 scores were moderate (66.4 [SD 10.3] and 3.01 [SD 0.55], respectively), demonstrating a neutral attitude regarding AI use. In bivariate analyses, ATTARI-12 differed by the caregiver-perceived illness severity (p < .001) and perceived financial status (p = .055). Linear regression suggested that the PPFI acceptance subscale (β=0.05, 95% CI 0.002-0.097; p = .04) and harnessing subscale (β=0.034, 95% CI -0.004-0.073; p=.08) were associated with ATTARI-12 scores.

Conclusions:

This study suggests that caregivers’ acceptance of such technology may be non-linearly influenced by both their financial status and their care recipient’s disease severity. Additionally, the study provides justification for using the ATTARI-12 and PPFI scales to assess AI acceptance and psychological flexibility in informal caregivers of PLwD, respectively. Future research should explore these factors across a larger and more diversified cohort to further generalize our findings.