Advance Care Planning for African American Caregivers of Relatives With Dementias: Cluster Randomized Controlled Trial.

Only four churches were used, with two sites per condition, so site-level variation was limited and standard errors may be underestimated. Participants were recruited from African American megachurches in metropolitan Chicago, so the sample may not represent caregivers in other communities or cultures. Within randomized sites, caregivers were assigned by convenience rather than individually randomized. Intention and written care plan outcomes were not measured before the intervention; the study assumed all participants started with no. The study tested eight outcomes at two time points, and although a conservative Bonferroni correction was applied, multiple comparisons remain a concern. Many caregivers did not know the care recipient's dementia diagnosis, which may affect interpretation of dementia knowledge outcomes. Of 358 enrolled, 264 completed the week 20 posttest, so long-term findings rest on fewer participants.

Do not conclude that this education increases written end-of-life care plans. The trial found better knowledge and self-efficacy, but intentions and written care plans did not differ from the control group.