(105) Effectiveness of Mono and Dual Antidementia Pharmacotherapy Among US Nursing Home Residents

Background: Half of nursing home (NH) residents have a diagnosis of Alzheimer’s disease (AD) or related dementia (RD). Cholinesterase inhibitors (ChEIs) and memantine have been widely used for the symptomatic treatment of ADRD for decades. Recent evidence indicates their use in NHs has been declining, and there is widespread variation. A 2018 review of guidelines highlighted discrepant recommendations and called for well-designed longitudinal studies.

Objectives: To compare the effectiveness of combination therapy, monotherapy, and no treatment with ChEIs and memantine.

Methods: We conducted a new-user cohort study using 2011-2018 Minimum Data Set (MDS) and Medicare Parts A and D claims among US NH residents with an AD or selected RD diagnosis and Medicare fee-for-service enrollment. We followed new-users for a 90-day titration period after first ChEI or memantine dispensing and those titrated to a therapeutic dosage were assigned to ChEI, memantine, or combination treatment groups as of day 90 (index date). A random MDS assessment was selected for non-users and day 90 afterwards was their index date. Effectiveness outcomes (higher scores are worse) included changes in functioning in the activities of daily living (ADL) score (range 0-28), the cognitive function scale (CFS; range 0-3), and the Distressed Behavior in Dementia Indicator (DBDI) scale (range 0-18) from baseline (before initiation) to the assessment nearest 90-days post-titration. Propensity score matching (1:1) was performed separately for the three groups. Mixed effects linear models with weights for missingness compared effectiveness measures over time. Subgroup analyses were performed by disease severity and AD diagnosis.

Results: Groups were well-balanced after matching (n=40,150 ChEI, n=13,371 memantine, and 7,098 combination users and non-users each). Almost half of residents were 85 years or older. Less than half had an AD diagnosis.

Among those with ADRD, ChEIs alone (CFS: -0.06; 95% CI: -0.07 to -0.05; ADLs: -0.41; 95% CI: -0.48 to -0.34) and in combination (CFS: -0.03; 95% CI: -0.05 to 0.00; ADLs: -0.28; -0.46 to -0.11) had small favorable differences in cognition and ADL functioning, but not the DBDI; a small difference was observed in ADLs for memantine (-0.30; 95% CI: -0.42 to -0.19). Favorable findings for CFS and ADLs were concentrated among those with moderate to severe impairment. Among those with AD, findings were generally consistent, but magnitudes varied.

Conclusions: Modest benefits for cognition and ADLs were observed with treatment of NH residents for ADRD, with larger effects among those with more advanced disease. Next, we will compare safety and survival outcomes and examine heterogeneity of effects to determine for which NH residents benefits of treatment outweigh risks.