Optimizing Antipsychotic Use across Nursing Homes: Implications for Dementia Care

Objectives: The study objectives were to (1) examine individual and contextual factors associated with initiation of antipsychotics among short-stay nursing home residents with dementia; (2) evaluate the relationship of nursing home level initiation of antipsychotics with all-cause hospitalization and mortality; and (3) examine individual and contextual factors associated with discontinuation of antipsychotics among short-stay nursing home residents with dementia. Methods: This study involved retrospective cohort study design using Medicare claims, Minimum Data Set (MDS), and the Certification and Survey Provider Enhanced Reporting (CASPER) data from 2007-2009. The participants included individuals aged 65 years or older with Medicare continuous eligibility six months prior to the nursing home admission. The study sample consisted of short-stay (a nursing home stay of ≤ 100 days) elderly nursing home residents with dementia. The study excluded patients with any antipsychotic use in six months prior to the nursing home admission (objectives 1 and 2) or patients with no antipsychotic use in six months prior to the nursing home admission (objective 3) based on Part D. The outcome measures of interest were measured after nursing home admission and included patient-level antipsychotic initiation (objective 1) and discontinuation (objective 3) and 30-days all-cause hospitalization and mortality (objective 2). The primary independent variable for objective was contextual level predisposing factor, nursing home level antipsychotic initiation rate (objectives 1 and 2) and nursing home level antipsychotic discontinuation (objective 3). The nursing home level antipsychotic imitation rate was categorized into 0% (no use) and tertiles (low, medium, or high) based on antipsychotic use rate groups. The nursing home level antipsychotic discontinuation rate was categorized into less than equal to or greater than the median value. Multilevel Andersen Behavioral model was used to select individual and contextual level predisposing, enabling and need factors. The differences in covariate distributions were evaluated using chi-squared (χ2) tests for categorical variables and t-tests for continuous variables. Hierarchical logistic regression model was used to achieve all the study objectives. Results: For the objectives 1 and 2, a total of 9,611 dementia patients was identified residing in 2,548 nursing homes. Overall antipsychotic initiation rate was 11.2%; nursing home level initiation rate was 0% in 890, <12.6% in 659 (low), 12.6%-16.7% in 465 (medium) and 16.8%-70% in 534 (high) nursing homes. For the first objective, hierarchical logistic regression model revealed that nursing home level antipsychotic initiation rate was associated with patient’s increased likelihood of initiating with an antipsychotic (Medium: OR 1.36, 95% CI 1.09-1.71, High: OR 1.48, 95% CI 1.20-1.82). Among patient-level factors, predisposing (male), enabling (dementia unit), and need (aggressive behavior, moderate and severe cognitive performance, psychosis, delirium, and antianxiety medication use) characteristics were associated with higher likelihood of antipsychotic initiation. For the second objective, nursing home level antipsychotic rate was not associated with 30-day hospitalization (low, OR 1.01, 95% CI 0.89-1.15; medium, OR 0.97, 95% CI, 0.84-1.13; high, OR 1.11, 95% CI 0.97-1.28) or 30-day mortality (low, OR 0.98, 95% CI 0.79-1.23; medium, OR 0.87, 95% CI, 0.66-1.14; high, OR 0.90, 95% CI 0.70-1.16) compared to 0% initiation rate. However, patient level antipsychotic use was associated with increased risk of 30-day hospitalization but not with 30-day mortality. Patient level antipsychotic use was associated with increased risk of 30-day hospitalization but not with 30-day mortality. For the third objective, the cohort consisted of 389 dementia patients who used antipsychotics prior to admission to 146 nursing homes. Overall antipsychotic discontinuation rate was 34.7%; nursing home level discontinuation rate was 0 to 66.6% in 60 (≤ median) and >66.6% in 86 (> median) nursing homes. Hierarchical logistic regression model revealed that nursing home level antipsychotic discontinuation rate was not associated (OR: 1.50, 95% CI: 0.88-2.55) with patient’s likelihood of discontinuing antipsychotics. But nursing homes with larger bed size were more likely to discontinue antipsychotics (OR: 1.98, 95% CI: 1.16-3.39) compared to smaller bed size nursing homes. Among patient-level, need factors such as depressed mood disorder and anxiety were associated with lower likelihood, and cognitive performance scale and Elixhauser comorbidities were associated with higher likelihood of antipsychotic discontinuation. Conclusions: Both patient and contextual level factors influenced initiation and discontinuation of antipsychotics among nursing home residents with dementia. The study revealed that antipsychotic prescribing practices play an important role in the use of antipsychotics in nursing home residents with dementia. However, nursing home level antipsychotic initiation was not associated with short-term all-cause hospitalization or mortality. But patient-level antipsychotic use was associated with higher risk of 30-day hospitalization. Systematic efforts are needed by nursing homes to optimize antipsychotic use in dementia patients by initiating or discontinuing these potent medications based the risk-benefit profile of these agents in individual patients.

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Pharmaceutical Health Outcomes and Policy

Grantor

University of Houston

Year dc:date.issued

2017

Author dc:creator

• Sura, Sneha

Advisor dc:contributor.advisor

• Aparasu, Rajender R.

Committee members dc:contributor.committeemember

• Johnson, Michael L.
• Chen, Hua
• Holmes, Holly M.
• Chan, Wenyaw

• Dementia
• Nursing Homes
• Antipsychotics
• Multilevel Andersen Behavioral Model

Statement dc:rights

• The author of this work is the copyright owner. UH Libraries and the Texas Digital Library have their permission to store and provide access to this work. Further transmission, reproduction, or presentation of this work is prohibited except with permission of the author(s).

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/10657/2960

OAI identifier oai:identifier

oai:uh-ir.tdl.org:10657/2960