Extension Connection Evaluation

This study will perform a retrospective analysis to evaluate the effectiveness of a
multicomponent educational intervention to improve medication use and management of
behavioral and psychological symptoms of dementia, relative to a statewide intervention that
has been ongoing. Medication use and symptom severity outcomes will be compared among 29
intervention counties and 10 demographically similar non-intervention counties. The long-term
effectiveness of the statewide intervention will also be evaluated. Medicare and Minimum Data
set data from 2011 to 2014 will be used to evaluate prescribing and other outcomes in 1)
outpatients with dementia, and 2) nursing home residents, based on exposure to interventions
among healthcare and nursing home providers. Measurable goals included increasing the
appropriateness of antipsychotic prescribing and reducing anticholinergic use. Effect
modification will be evaluated by level of participation in the intervention and other
prescriber and nursing home facility characteristics.

Inclusion Criteria:

1. Current Medicare fee-for-service plan and Part D enrollment for at least 3 months
prior to the month of interest and during the month of interest (allowing a gap of one
month to be considered as continuous enrollment if the person is enrolled during the
month of interest)

2. Age greater than or equal to 66 years at the end of a year of interest

3. Any dementia diagnosis prior to the month of interest, or current resident of a
nursing home

4. For the nursing home analysis, at least 14 days in the facility in the month being
classified

Exclusion Criteria:

1. Greater than 15 days of the month of interest when prescription fills are unobservable
from Medicare Part D claims due to hospital inpatient status, skilled nursing facility
resident status, or hospice enrollment

2. A diagnosis of schizophrenia, schizoaffective disorder, bipolar disorder, Huntington's
disease, Down's syndrome, or mental retardation or developmental disability at any
time during the study period. Diagnoses after entry into the study sample will be
included because of the limited look-back period to define some conditions for some
individuals, and because these are generally not conditions with late-life onset so
they are assumed to have been present prior to the first occurrence of a diagnosis in
the available data

3. Currently comatose based on the most recent MDS assessment—if currently in a nursing
home