A Cluster Randomized Trial: Mixed Methods Comparison of 2 Approaches to Promote Nonpharmacologic, Resident-Centered Dementia Care in Nursing Homes.

Objectives: Nursing home dementia care initiatives have prioritized nonpharmacologic approaches to manage behavioral symptoms. This study compares the effectiveness of 2 nonpharmacologic approaches on resident outcomes.

Design And Intervention: Cluster randomized controlled trial using a convergent mixed methods design to compare the team-based and problem-based approaches to dementia care. The team-based arm received facility-wide training, providing a shared language and knowledge across staff. The problem-based arm received discipline-specific web-based training for certain staff.

Setting And Participants: Residents living with dementia (n = 2728) in 53 nursing homes; 327 staff.

Methods: Outcomes were antipsychotic medication use, behavioral symptoms, rejection of care, and wandering captured with the Minimum Data Set. Exploratory quantitative analysis compared change in outcomes between baseline and 6-month follow-up, across arms using a difference-in-difference model. Qualitative data were collected via staff interviews. Integration of quantitative and qualitative data determined whether these sources provided confirmatory, emergent, or contradictory information on outcomes.

Results: Resident wandering increased 3.51 percentage points (95% CI 0.6, 6.4) in the team-based arm compared to the problem-based arm, without increase in adverse events. Data integration confirmed this finding. Staff in the team-based arm described wandering as a positive coping strategy for the resident. We found mixed results for the other 3 outcomes. There was no statistically significant difference in antipsychotic medication use across arms, but staff interviews did reveal differences. There were trends of increased behavioral symptoms in the team-based arm and increased rejection of care in the problem-based arm (neither statistically significant), which were confirmed by qualitative findings.

Conclusions And Implications: Integration results suggest a benefit for resident wandering in the team-based approach, compared with the problem-based approach. Although findings are exploratory, training for all nursing home staff that accounts for diverse education and training needs may influence care delivery and have benefits for residents living with dementia.