Patient and Caregiver Benefit From a Comprehensive Dementia Care Program

1-Year Results From the UCLA Alzheimer's and Dementia Care Program

David B. Reuben, MD, AGSF; Zaldy S. Tan, MD, MPH; Tahmineh Romero, MS; Neil S. Wenger, MD, MPH; Emmett Keeler, PhD; Lee A. Jennings, MD, MSHS

Disclosures

J Am Geriatr Soc. 2019;67(7):2267-2273. 

In This Article

Methods

This study describes 1-year patient and caregiver outcomes for the first 1091 participants in the UCLA ADC Program; reports the percentages who demonstrated patient, caregiver, or either benefit; and identifies predictors of who benefits. The study was approved by the UCLA Institutional Review Board.

Participants

Patients in the UCLA ADC Program were referred to the program by their primary care or specialist physicians, who agreed to comanage care with the DCMs. The only additional eligibility requirements were that they were UCLA patients and they did not live in a nursing home at the time of referral to the program.

Description of the Program

The UCLA ADC Program utilizes an NP DCM supervised by a physician dementia specialist to tailor and facilitate dementia care delivery in collaboration with the primary care physician (comanagement). NPs can write orders, communicate directly through the electronic health record, and facilitate clinical care. Dementia care is based in the healthcare system, which partners with community-based organizations to provide comprehensive, coordinated, patient-centered care. Key components include:

• Structured needs assessments of patients and their caregivers.

• Creation and implementation of individualized dementia care plans.

• Ongoing dementia care management by a DCM supervised by a physician dementia specialist, including, as needed:

   • In-person sessions

   • Telephone follow-up to monitor implementation of dementia care plans

   • Teaching dementia management skills to caregivers

   • Consultation with neurology, geriatric psychiatry, psychology, or geriatrics

   • Caregiver support groups

   • Caregiver education through a community lecture series

   • Referral to community-based organizations for services (eg, adult day care, counseling) as well as caregiver training, including vouchers, if needed, for temporary services.

• Monitoring and revising care plans, as needed, including active monitoring (a minimum of a telephone call every 4 months) and support of caregivers.

• Access 24/7, 365 days a year for assistance and advice. Night, weekend, and holidays are covered by the UCLA geriatrician on call.

Measures

Mini-Mental State Examination (MMSE) measures cognition with ranges from 0 to 30, with lower scores indicating greater cognitive impairment.[11]

Functional Activities Questionnaire (FAQ) measures functional status and ranges from 0 to 30, with higher scores indicating more functional dependence.[12]

Cornell Scale for Depression in Dementia is a 19-item validated tool used to assess depressive symptoms in patients with dementia. Scores range from 0 to 38, and a score of 11 or greater indicates probable depression.[13]

Functional status was measured using Basic Activities of Daily Living (ADLs)[14] and Instrumental ADLs (IADLs) scales,[15] which have been validated and are well established in research and clinical use. We also administered the FAQ, which measures functional status and ranges from 0 to 30, with higher scores indicating more functional dependence.[12]

Neuropsychiatric Inventory Questionnaire (NPI-Q)[16] is a validated survey that assesses the caregiver's perception of the severity of 12 dementia-related psychiatric and behavioral symptoms and the level of distress experienced by the caregiver as a result of these symptoms. NPI-Q Severity score ranges from 0 to 36, and NPI-Q Distress score ranges from 0 to 60, with higher scores indicating more severe symptoms and distress, respectively. In a nursing home population, the minimal clinically important difference was determined to be 2.8 to 3.2 points for severity and 3.1 to 4.0 points for distress.[17]

Modified Caregiver Strain Index (MCSI)[18] is a 13-item validated tool used to assess severity of caregiver strain. The index targets financial, physical, psychological, and social aspects of strain and is scored from 0 to 26, with higher scores indicating greater levels of strain.

Patient Health Questionnaire-9 (PHQ-9)[19] is a nine-item validated tool used to assess depressive symptoms in the caregiver using the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria for major depression and is scored from 0 to 27, with scores greater than 10 indicating moderate symptoms and scores greater than 20 indicating severe depressive symptoms.

Dementia Burden Scale-Caregiver (DBS-CG)[20] is a composite of the NPI-Q Distress, MCSI, and PHQ-9 scales, with items transformed linearly to be on a 0 to 100 possible range and then averaged, with higher scores indicating higher caregiver burden. The minimal clinically important difference for the DBS-CG is five points.

Definition of Clinical Benefit

We captured benefit in two ways: (1) for those in the two worst tertiles at the baseline, improvement by the minimal clinically important difference over 1 year (yes/no) or (2) for those who were in the lowest symptom tertile at the baseline, by maintaining low symptoms at baseline and 1 year (yes/no) (Table 1). We measured patient symptoms using the NPI-Q Severity scale (the only patient outcome measure anticipated to benefit from the program) and caregiver symptoms using the DBS-CG scale. Benefit on the NPI-Q severity scale was defined as improving by at least three points, the minimal clinically important difference,[17] or having baseline and 1-year scores of 6 or lower (the upper limit of the baseline tertile of fewest patient symptoms) (Supplementary Figure S1). DBS-CG benefit was defined as improving by at least five points, the minimal clinically important difference,[20] or having baseline and 1-year scores of 18.8 or lower (the upper limit of the baseline tertile of fewest caregiver symptoms) (Supplementary Figure S2). Defining benefit in this manner allowed us to capture both improvement (by the minimal clinically important difference or more) and maintenance of low symptoms (baseline and 1-year symptoms are low).

Supplementary Figure S1.

Patient clinical benefit. Patients who benefitted are in blue shaded area, either because of improvement of at least the minimal clinically important difference (right of the diagonal) or maintenance of low symptoms at baseline and 1 year (below horizontal line).

Supplementary Figure S2.

Caregiver clinical benefit. Caregivers who benefitted are in blue shaded area, either because of improvement of at least the minimal clinically important difference (right of the diagonal) or maintenance of low symptoms at baseline and 1 year (below horizontal line).

Analyses

Baseline characteristics of the cohort who survived to 1 year, stratified by missing at year 1, were described using proportions for categorical variables and means with SDs or medians with interquartile ranges for continuous variables. Differences between missing/not missing were tested using the χ 2 test for categorical variables and the Student t-test or Wilcoxon test for continuous variables, depending on whether mean or median was reported. We used item-level data from FAQ, MMSE, Cornell, and IADLs to impute full predictor scales for patients who had at least one item on the scale. For missing outcomes, we required at least one item in the NPI-Q Severity scale and at least one item in each of the three component scales (NPI-Q Distress, PHQ-9, and MCSI) to impute DBS-CG scores.

One-year changes in NPI-Q Severity and DBS-CG among patients and caregivers were used to measure the benefit from the program. To adjust for the bias that could arise from systematic differences between complete cases and patients with missing data, we used inverse probability weighting (IPW).[21] The estimated probability of being a complete case was calculated, performing a logistic regression modeling missing patient or caregiver outcomes at year 1 (yes/no), adjusting for patient's and caregiver's baseline demographic and clinical characteristics. The predictors of being a complete case include patient's age, sex, race, education, dementia type, and MMSE category and the tertiles of FAQ, Cornell, and NPI-Q Severity; and the caregiver's sex, relation to patient, and the tertiles of MCSI, PHQ-9, and NPI-Q Distress. The tertiles were modeled as continuous variables (test of trend). We examined the distribution of the estimated inverse probability of being a complete case, and there were no influential weights.

Continuous outcomes were modeled performing IPW linear regression. For binary outcomes, we used IPW univariable and multivariable logistic regressions to calculate unadjusted and adjusted odds ratios and their 95% confidence intervals, respectively. To model clinical benefit (improving [yes/no] by the minimal clinically important difference or maintaining low symptoms [yes/no]), for patients and caregivers, we included predictors that were significant in univariable logistic regressions plus the clinically relevant covariates. As a sensitivity analysis, all the models were redone without applying IPW.

All tests were two sided, and P < .05 was considered statistically significant. Analyses were performed using SAS, version 9.4 (SAS Institute Inc).

processing....