JMIR Publications

ABSTRACT

Background:

Web-based interventions aimed at supporting informal caregivers (ICGs) of persons with dementia (PLWD) have the potential to improve caregivers’ well-being and psychological health. However, few interventions are widely implemented for this population, and none of the prior reviews have systematically examined the use of the behavior change techniques (BCTs), theories, and agents in web-based interventions for ICGs of PLWD. To better understand this implementation gap, we reviewed the literature to map BCTs, theories, and agents deployed in the studies. Furthermore, because there is an emerging consensus that retention could be shaped by participants’ characteristics and behavioral factors, we explored relationships between these features and retention rates across the studies.

Objective:

We pursued three objectives: 1) to map behavioral theories, change agents, and BCT involved in the web-based interventions for ICG of PLWD; 2) to examine the relationship between behavioral change elements and retention in the studies; 3) to examine the relationship between participants’ characteristics and study retention.

Methods:

We conducted a literature review in PubMed, Embase, and PsycINFO, utilizing the following keywords and their corresponding MeSH terms: dementia, caregivers, online or web-based intervention. The time limits were January 1998 to February 2021. Utilizing the BCTv1 taxonomy, a taxonomy that specifies the active ingredients in interventions, two coders collected, summarized, and analyzed the frequency distributions of the BCTs. Similarly, they abstracted and analyzed participant characteristics, behavior change theories, behavior change agents (BCA), and retention rates of the studies.

Results:

The average age, the proportion of spousal ICGs, and retention rates were 62.3% (SD=7.4), 50.6% (SD=24.0%), and 74.6% (SD=11%) respectively. Only 52% of the studies utilized behavior change theories, but 88% included BCA. The most common BCTv1 clusters were "shaping knowledge" and "social support." The median number of "BCTv1 clusters" was 5 (IQR=3). We observed a negative correlation between the proportion of spousal ICGs and the retention rate (r=-0.5; P=.03) and between the number of BCTv1 clusters and retention rates (r=-0.53; P=.01). None of the other participant characteristics or behavioral factors were associated with retention rates.

Conclusions:

We found that almost half of the studies were not informed by behavior change theories. In addition, spousal involvement and a higher number of BCTs were each associated with lower retention rates. In planning future studies, researchers should consider matching participants' characteristics with their intended intervention as the alignment might improve their retention rates. Clinical Trial: NA