A published clinical trial protocol details the design of a randomized controlled trial evaluating the Disaster PrepWise-Caregiving program, a behavioral intervention designed to strengthen disaster preparedness and psychological resilience among family caregivers of community-dwelling individuals with Alzheimer’s disease and related dementias. Authored by lead investigator Ai Iizuka, corresponding author Sato Ashida, and colleagues from the University of Iowa, the University of California San Diego, and the University of California Irvine, the protocol appeared in Frontiers in Public Health. The investigation evaluates whether providing personalized emergency planning assistance alleviates caregiving stress while bolstering coping capacity. Although disaster materials exist, few address the demands of dementia caregiving, which often disrupt care continuity during emergencies. By testing the intervention among two hundred and fifty caregivers, the trial aims to evaluate efficacy on psychological and social adaptation outcomes over six months.
The public health rationale reflects the severe vulnerabilities confronting dementia care networks during natural hazards. In the United States, approximately twelve point seven million family caregivers provided nineteen point six billion hours of unpaid care for people with dementia in 2025, and an estimated seven point four million older adults live with Alzheimer’s dementia in 2026. Concurrently, environmental emergencies have imposed growing burdens; in 2024, the nation documented twenty-seven weather and climate disasters exceeding one billion dollars each, with total losses reaching one hundred eighty-two point seven billion dollars. Studies examining caregivers affected by hurricanes and flooding show that care responsibilities complicate evacuation decisions, emergency supplies, and recovery resource utilization, with care recipients sometimes resisting evacuation. Despite these risks, fewer than twenty-five percent of older adults possess an emergency plan, and only ten percent are registered on disaster registries.
To meet these needs, researchers adapted the Disaster PrepWise intervention, which was originally developed through community-engaged processes to help older households establish comprehensive disaster management plans. Endorsed by the Federal Emergency Management Agency, the program adopts an all-hazards approach addressing diverse emergencies, including extreme weather, fires, tornadoes, floods, and pandemics. While pilot testing in older adults demonstrated measurable gains in preparedness behaviors and emergency support networks, its efficacy among dementia caregivers remained unexamined. The adapted program, termed Disaster PrepWise-Caregiving, embeds caregiving requirements directly into five core modules covering household assessment, personal emergency networks, emergency documents, medication supplies, and emergency kits. Unlike the original tool where caregiving considerations were optional, the adapted program requires every participant to incorporate care continuity, evacuation accommodations, and care recipient needs into their primary emergency planning.
The intervention and its proposed mediation pathways rely on two theoretical frameworks: the Extended Parallel Process Model and the Stress Process Model. In parallel, the Stress Process Model provides a conceptual foundation for understanding caregiver well-being under stress. Within this framework, social and psychological coping assets—specifically self-efficacy, specialized preparedness, and social support networks—operate as vital buffers.
The investigation is organized as a prospective five-year randomized controlled trial spanning participant recruitment from March 2023 through May 2026, with each enrolled caregiver followed for approximately six months. Recruitment deploys two primary strategies across the state. First, researchers query the Enterprise Data Warehouse for Research, which aggregates electronic medical records from Epic systems at University of Iowa Hospitals and Clinics to identify patients diagnosed with dementia. Informational letters, study flyers, and interest forms with prepaid envelopes are mailed to patients’ families, who can also enroll digitally via a quick-response code. Trained research staff make up to three follow-up contacts by phone or email. Second, collaborating outpatient clinics and caregiver organizations distribute recruitment flyers publicly. Because the study represents minimal physical and psychological risk, the institutional review board waived written consent, permitting research personnel to obtain verbal consent prior to baseline assessment.
Study inclusion and exclusion criteria define a broad, community-centered population of caregivers. Eligible participants must be adults aged eighteen or older who provide or coordinate care for an individual diagnosed with Alzheimer’s disease or a related dementia residing within the state. Family caregivers include biological and non-biological relatives as well as close friends, irrespective of whether they reside in the same household as the care recipient. Furthermore, the trial includes caregivers supporting individuals who do not currently require daily living assistance. Caregivers are also excluded if they present cognitive or physical impairments that prevent them from providing informed consent or completing research surveys, as verified using standard institutional evaluation screening tools.
Participants assigned to the intervention group receive a multi-stage curriculum delivered by trained volunteer interventionists using the electronic Disaster PrepWise tool. The initial session lasts approximately ninety minutes and takes place at a location chosen by the participant, such as their private home, a local library, or remotely via videoconference or telephone. The interventionist reviews the five core modules, enters participant details into electronic forms, and generates personalized emergency plans and action steps. Individual plans are stored securely on a HIPAA-compliant server accessible only by the participant with private credentials; researchers receive data only regarding module completion extent. Four weeks later, interventionists conduct a fifteen-to-thirty-minute telephone session to refine plans, assist with unfinished action items, and address questions. Eight weeks following the initial meeting, a final telephone encounter reinforces annual plan updates and provides closing assistance.
In contrast, participants in the control condition complete study assessments at baseline, three months, and six months without receiving intervention materials during the trial. Contact with the study team is restricted to assessment administration, but controls are invited to receive the program upon completing their six-month survey. Randomization occurs after the baseline assessment is completed, utilizing sequences generated in R via the blockrand package. Allocation is stratified by rurality, maintaining a targeted rural-to-urban ratio of two to three, and concealed through a central web platform to prevent bias. Because the educational intervention cannot be concealed from participants or interventionists, blinding is maintained among outcome assessors, data entry assistants, and the principal investigator. Intervention staff have no access to research outcome databases, and research surveyors cannot access participant intervention plans, ensuring rigorous separation between intervention delivery and outcome assessment.
The trial relies on established psychometric measures to capture primary outcomes, hypothesized mediators, and exploratory variables. The primary outcomes are caregiver resilience and caregiver stress assessed at baseline, three-month, and six-month intervals. Resilience is quantified using the four-item Connor-Davidson Resilience Scale, which substituted the originally planned twenty-five-item version before enrollment to reduce survey length while preserving validity and divergent validity against stress. Caregiver stress is assessed using the ten-item Cohen Perceived Stress Scale. Secondary mediating outcomes comprise caregiver self-efficacy in managing care and using community services, caregiver preparedness evaluated with an eight-item scale, and social support networks measured by adapting the CaREnet instrument and prior Disaster PrepWise metrics to assess network composition, relationship quality, and crisis assistance. Caregiver burden serves as an exploratory outcome using the four-item Zarit Caregiver Burden Interview Screening Version to facilitate integration with existing literature.
Statistical analysis plans specify analysis of covariance models adjusting for baseline values, prespecified covariates, and sex as a biological variable to evaluate between-group differences in resilience and stress at six months. Exploratory analyses under Aim 1b evaluate urban-rural differences by incorporating treatment-by-rurality interaction terms. For Aim 2, mediation analyses use the mediation package in R to examine whether three-month changes in self-efficacy, preparedness, and support networks mediate intervention impacts on six-month primary outcomes, estimating indirect effects via non-parametric bootstrapping with ninety-five percent confidence intervals. The original power calculation assumed a two-sided alpha of zero point zero five, requiring one hundred forty-six participants for resilience and one hundred twenty for stress to achieve eighty-five percent power. Initial targets planned to enroll two hundred participants to yield one hundred fifty completers under an assumed twenty-five percent attrition rate; however, after observing higher early attrition, researchers expanded recruitment to two hundred fifty caregivers.
Data management, participant tracking, and audit trails are maintained within the REDCap platform, while surveys are collected via Qualtrics and exported into R and SAS programs. Missing data will be handled under missing-at-random assumptions using inverse probability weighting with stabilized propensity score weights. The principal investigator oversees participant safety, and an independent Data and Safety Monitoring Committee convenes twice yearly to review trial progress and operating procedures. Per institutional standards, quantitative survey data will be archived in the University of Iowa open-access repository with persistent Digital Object Identifiers for at least ten years, becoming accessible within two years of study completion. Alongside efficacy testing, Aim 3 involves a formative mixed-methods evaluation guided by the Practical, Robust Implementation Science Model, gathering implementation feedback from caregivers, aging service organizations, and advisory boards to inform future scalable translation across community health settings.
Subject of Research: Disaster preparedness intervention for family caregivers of persons with dementia
Article Title: Emergency preparedness support for caregivers of persons with dementia: a randomized controlled trial of the Disaster PrepWise-Caregiving Program
Article References: Iizuka, A., Killian, E., Donohoe, M., Cho, H., Seaman, A., Lynn, F. B., Reisinger, H. S., Ramirez, M. R., & Ashida, S. (2026). Emergency preparedness support for caregivers of persons with dementia: a randomized controlled trial of the Disaster PrepWise-Caregiving Program. Frontiers in Public Health, 14, Article 2010784. https://doi.org/10.3389/fpubh.2026.2010784
Image Credits: AI Generated
DOI: 10.3389/fpubh.2026.2010784
Keywords: Disaster PrepWise, dementia caregiving, Alzheimer's disease, randomized controlled trial, emergency preparedness, caregiver resilience, caregiver stress, Connor-Davidson Resilience Scale, Perceived Stress Scale, caregiver self-efficacy, social support networks, Extended Parallel Process Model, Stress Process Model, rural caregiving, disaster planning, Zarit Caregiver Burden Interview
News Source: Cassandra Pierce. (November 21, 2026). Randomized Trial Evaluates Disaster Preparedness Program for In-Home Dementia Caregivers. Scienmag.



