Impact of dementia special care units for short-stay nursing home patients

Abstract

Background: Improving quality of care provided to short-stay patients with dementia in nursing homes is a policy priority. However, it is unknown whether dementia-focused care strategies are associated with improved clinical outcomes or lower utilization and costs for short-stay dementia patients.

Methods: We performed a national survey of nursing home administrators in 2020-2021, asking about the presence of three dementia-focused care services used for their short-stay patients: (1) a dementia care unit, (2) cognitive deficiency training for staff, and (3) dementia-specific occupational therapy. Using Medicare claims, we identified short-stay episodes for beneficiaries residing in surveyed skilled nursing facilities (SNFs) with and without dementia. We compared clinical, cost, and utilization outcomes for dementia patients in SNFs, which did and did not offer dementia-focused care services. As a counterfactual control, we compared these differences to those for non-dementia patients in the same facilities. Our primary quantity of interest was an interaction term between a patients' dementia status and the presence of a dementia-focused care tool.

Results: The study population included 102,860 Medicare episodes of care from 377 SNF survey respondents in 2018-2019. In adjusted comparisons of the interaction between dementia status and the presence of each dementia-focused care tool, dementia care units were associated with a 1.5-day increase in healthy days at home in the 90 days following discharge (p = 0.01) and a 3.1% decrease in the likelihood of a subsequent SNF admission (p = 0.001). Cognitive deficiency training was also associated with a 2.0% increase in antipsychotics (p = 0.03), whereas dementia-specific occupational therapy was associated with a 1.2% increase in falls (p = 0.01) per patient episode.

Conclusions: Self-reported use of dementia care units for short-stay patients was associated with modestly better performance in some, but not all, outcome measures. This provides hypothesis-generating evidence that dementia care units could be a promising mechanism to improve care delivery in nursing homes.

Keywords: dementia; nursing home; short-stay patients.

Figure 1:

Adjusted difference in claims-based outcome measures for patients with and without dementia with short-stays in SNFs with and without dementia-focused care tools The point and bars represent the coefficient estimate and 95% confidence interval from the adjusted regression results comparing outcomes for the interaction of dementia status (with vs. without dementia) and the presence of each of the dementia-focused care tools used for short-stay patients (dementia care unit, dementia occupational therapy, cognitive deficiency training). Negative coefficient estimates are generally associated with better outcomes (e.g., lower costs and lower readmission rates are good outcomes) while positive values are associated with worse outcomes. The exception is healthy days at home, for which higher numbers are better.

Figure 2:

Adjusted difference in MDS outcome measures for patients with and without dementia with short-stays in SNFs with and without dementia-focused care tools. The point and bars represent the coefficient estimate and 95% confidence interval from the adjusted regression results comparing outcomes for the interaction of dementia status (with vs. without dementia) and the presence of each of the dementia-focused care tools used for short-stay patients (dementia care unit, dementia occupational therapy, cognitive deficiency training). Negative coefficient estimates are generally associated with better outcomes while positive values are associated with worse outcomes.