A dementia care management intervention: which components improve quality?

OBJECTIVES: To analyze whether types of providers and frequency of encounters are associated with higher quality of care within a coordinated dementia care management (CM) program for patients and caregivers.
DESIGN: Secondary analysis of intervention-arm data from a dementia CM cluster-randomized trial, where intervention participants interacted with healthcare organization care managers (HOCMs), community agency care managers (CACMs), and/ or healthcare organization primary care providers (HOPCPs) over 18 months.
METHODS: Encounters of 238 patient/caregivers (dyads) with HOCMs, CACMs, and HOPCPs were abstracted from care management electronic records. The quality domains of assessment, treatment, education/support, and safety were measured from medical record abstractions and caregiver surveys. Mean percentages of met quality indicators associated with exposures to each provider type and frequency were analyzed using multivariable regression, adjusting for participant characteristics and baseline quality.
RESULTS: As anticipated, for all 4 domains, the mean percentage of met dementia quality indicators was 15.5 to 47.2 percentage points higher for dyads with HOCM--only exposure than for dyads with none (all P < .008); not anticipated were higher mean percentages with increasing combinations of provider-type exposure-up to 73.7 percentage points higher for safety (95% confidence interval 65.2%-82.1%) with exposure to all 3 provider types compared with no exposure. While greater frequency of HOCM-dyad encounters was associated with higher quality (P < .04), this was not so for other provider types.
CONCLUSIONS: HOCMs' interactions with dyads was essential for dementia care quality improvement. Additional coordinated interactions with primary care and community agency staff yielded even higher quality.