Background: Training family caregivers (hereafter "caregiver') can improve their own and their care recipients' health care experiences, yet only half of caregivers of U.S. Veterans report receiving sufficient training.
Objectives: To assess the implementation of a caregiver skills training program for caregivers of veterans with functional limitations, delivered in the U.S. Veterans Affairs healthcare system (VAHCS).
Methods: We conducted a Type III hybrid implementation-effectiveness, stepped wedge cluster randomized trial to evaluate implementation and effectiveness of an evidence-based caregiver skills training at eight VAHCSs (ClinicalTrials.gov ID: NCT03474380). Participants included veteran patients referred to five VA home- and community-based services who were receiving informal care (n=898) and their caregivers. Caregiver characteristics are presented for the subset of the eligible caregivers who completed a phone survey (n=435). Enrolled sites could invite other caregivers to attend the training based on their site's needs so we present total caregivers trained and eligible caregivers trained.
All eight enrolled sites received implementation support to implement the caregiver training program using the Replicating Effective Programs strategy over four randomly assigned time intervals. In this design, the usual care group is Veterans from the time intervals pre-training implementation and the intervention group is Veterans from the time intervals post-training implementation. Half the enrolled sites were randomized to also receive a complexity-science intervention focused on multi-disciplinary team collaboration called CONNECT (REP+CONNECT).
Implementation penetration was measured as proportion of eligible caregivers enrolled in training. Fidelity was measured as proportion of sites delivering two training rounds of the full curriculum in each six-month time interval. CONNECT was hypothesized to increase implementation penetration and fidelity.
Results: The control group had 327 Veterans and intervention group had 571 Veterans. The number of caregivers enrolled in training was 291 regardless of eligibility status, with 165 enrolled caregivers. The mean Veteran age was 76.5; 95% were male; 18.9% were Black, 4.9 % were multiracial or Asian American, American Indian, or Pacific Islander (AAPI), 5% were of unknown race, and 76.2% were white. Caregivers were 12 years younger on average than Veterans with a mean age of 64; 18% were Black, 11% were multiracial or AAPI, 1.6% were of unknown race, and 70% were white.
Fidelity to the intervention was high across all sites: 35 training rounds were delivered over 40 rounds expected (88%). Fidelity was markedly higher among REP sites compared to REP + CONNECT sites (95% vs 80%). Penetration ranged from 21.4% at REP+CONNECT sites to 35.4% at REP sites. The overall penetration showed rate was 29%.
Conclusions: Fidelity was high across all sites while penetration was relatively lower. The complexity-science based team-building intervention CONNECT did not enhance implementation outcomes and in fact implementation outcomes were better at REP sites. The last 6-month time interval of the intervention occurred at the start of the Covid19 pandemic. Where sites effectively shifted to virtual training delivery, reduced fidelity occurred largely in this final time interval. It is possible to implement a caregiver skills training program with fidelity in a large integrated health system using structured implementation support (REP).