EVALUATING THE IMPACT OF STAR-VA ON VETERANS’ PSYCHOTROPIC MEDICATION USE IN CLCS

Abstract The impact of STAR-VA on psychotropic drug use among residents with behavioral symptoms of dementia was evaluated through a difference-in-differences framework. STAR-VA residents enrolled 2013-2017 were evaluated longitudinally pre-post intervention. The primary outcome was the number of as needed administrations with an indication of ‘anxiety’ or ‘agitation’. The analytical cohort included 214 training cases and 1,870 controls from untrained sites meeting eligibility criteria. STAR-VA cases were less white (48% vs. 54%), less black (11% vs. 14%), and had significantly longer median length of stay (830 vs. 261 days), respectively. STAR-VA cases had on average 3.5 as needed doses/month of psychotropic medication before the intervention and 1.7 after, controls averaged 1.8 doses/month. After adjustment for person-time-fixed effects, enrollment was associated with 55% (95% CI:30, 68) reduction or an average 0.8 as needed psychotropic doses/month. Findings demonstrate effectiveness in decreasing as-needed psychotropic drug use among CLC residents, supporting continued implementation of STAR-VA.

average 0.8 as needed psychotropic doses/month. Findings demonstrate effectiveness in decreasing as-needed psychotropic drug use among CLC residents, supporting continued implementation of STAR-VA.

VARIATIONS IN SUSTAINING IMPLEMENTATION OF STAR-VA: THE ROLE OF KNOWLEDGE RESERVOIRS AND OTHER FACTORS
Jennifer L. Sullivan, 1 Kim Curyto, 2 Omonyêlé l. adjognon, 3 Jacqueline Pendergast, 1 Laura O. Wray, 4 and Michele Karel 5 , 1. Center for Healthcare Organization and Implementation Research (CHOIR), VA Boston Healthcare System,Boston,Massachusetts,United States,2. vA western new york healthcare system,batavia,New York,United States,3. center for healthcare organization and implementation research, boston VA healthcare system,boston,Massachusetts,United States,4. VA Center for Integrated Healthcare,VA Western NY Healthcare System (116N),Buffalo,New York,United States,5. Office of Mental Health and Suicide Prevention,Veterans Health Administration,Underhill,Vermont,United States Variation in STAR-VA sustainability across 20 trained VA Community Living Centers was explored using prospective qualitative methods utilizing the knowledge reservoirs framework including seven domains: People, Routines, Artifacts, Relationships, Information space, Culture, and Structure. We conducted directed content analysis of transcripts to identify facilitators and barriers of successful program sustainment. We found that people, usual routines, information sharing, and team relationships were the most often mentioned facilitators by CLC staff. Common reported barriers were people, team relationships, and work culture. Overlap was found in knowledge reservoirs acting as both facilitators and barriers at the same site, most often for people/teams, team relationships, and work culture. Results will be used to develop a sustainability intervention focused on addressing reported barriers. Most notably, a focus on having the appropriate team members, positive team relationships, usual routines, and a supportive work culture are critical for STAR-VA sustainability efforts. associated with a trend toward increased all cause mortality, driven by cancer deaths (results published NEJM September 2018). After the LDA intervention was halted in June 2017, ASPREE was extended as an observational cohort follow-on study, ASPREE-XT, to measure potential delayed LDA effects on ASPREE outcomes. The ASPREE study primary results will be summarized, and the rationale for and performance of the novel DFS geriatric outcome discussed. New results of the analysis of dementia as a secondary outcome will also be presented (both for overall dementia and Alzheimer's disease). We will also examine the unexpected increased allcause mortality attributed to cancer deaths, despite no significant difference between groups for all incident cancer, and effects of LDA on incident metastatic disease. The important implications of the ASPREE results for prescribing LDA for primary prevention in health elderly will be discussed, and the ASPREE-XT study design and progress described. Lastly, the breadth of the ASPREE sub-studies including the Biobank, Brain Imaging studies and Genomics, and opportunities to access the rich ASPREE data and collaborate with ASPREE investigators will be reviewed.

John McNeil 1 , 1. Department of Epidemiology & Preventive Medicine, Monash University, Melbourne VIC, Australia, Melbourne VIC, Australia, Australia
Disability-free survival (DFS), defined as survival free of disability and dementia was the primary outcome measure of the ASPREE clinical trial. As previously reported, there was no benefit of low dose aspirin on the primary end point of dementia, physical disability or death, but bleeding risks were increased. In total, 1,835 participants reached the primary endpoint, confirmed amongst approximately 3,000 who had triggered for one of the end-points. Dementia was the most labor intensive component of DFS. Several previous primary prevention aspirin studies had identified a reduction of vascular events counterbalanced by an increase in serious bleeding, leaving the question of net outcome to an intuitive decision. DFS was chosen because it balances the positive and negative effects of a preventive drug such as aspirin. It also encapsulates the primary purpose of a preventive drug in older people i.e., to prolong a healthy lifespan rather than prevent a defined disease.  Clayton VIC,Victoria,Australia,4. Division of Cancer Prevention,National Cancer Institute,Bethesda,Maryland,United States In ASPREE, we previously reported a surprising increase in cancer-related deaths associated with initiating aspirin.

EFFECT OF INITIATING ASPIRIN ON CANCER EVENTS IN THE HEALTHY ELDERLY
We now report primary incident cancer events. Aspirin was not associated with risk of incident cancer (HR=1.04, 95% CI 0.95-1.14), including non-metastatic cancer (HR=0.99, 95% CI, 0.89-1.11) and colorectal cancer (HR=1.02; 95% CI, 0.81-1.30). However, risk of incident metastatic cancer was elevated with aspirin (HR=1.18; 95% CI,0.96-1.46), although this could be attributable to chance. In ASPREE, the increase in cancer deaths associated with initiation of aspirin was not accompanied by a significant increase in overall incident cancer after 4.7 years. However, there did appear to be an increase in the incidence of advanced cancer in the aspirin arm. These data support the possibility that aspirin may adversely affect short-term outcomes among elderly participants with undiagnosed cancers (e.g. prevalent tumors at enrollment or early incident tumors) and/or may have differential effects according to age.

THE ASPREE STUDY: DEMENTIA DUE TO ALZHEIMER'S DISEASE OUTCOMES Raj C. Shah 1 , 1. Rush University, Chicago, Illinois, United States
In the ASPREE clinical trial, aspirin 100mg daily in health older adults did not delay onset of dementia, a pre-specified secondary outcome over a period of 5 years. We examine whether low-dose aspirin versus placebo is related to incident dementia due to Alzheimer's disease. Older communitydwelling participants free of dementia, physical disability, and conditions requiring aspirin treatment were recruited (n=19,114). Participants were administered a cognitive test battery during follow-up and participants with suspected dementia underwent a more extensive dementia assessment. An expert international panel adjudicated dementia according to DSM-IV criteria, with sub-classification according to NIA-AA criteria. Over a median 4.7 years, 575 participants had a confirmed dementia diagnosis. In analysis of the 41% of cases classified as dementia probably due to Alzheimer's disease, no difference in the incidence between the treatment arms was found (HR=0.96, 95% CI =0.74, 1.24). Plans for continued assessment of cognition in ASPREE-XT will be presented.

POTENTIAL IMPLICATIONS OF ASPREE ON ASPIRIN PRIMARY PREVENTION GUIDELINES
Anne M. Murray 1 , 1. Berman Center for Outcomes and Clinical Research, Minneapolis, Minnesota, United States The 2016 USPSTF guidelines for aspirin to prevent CVD and colorectal cancer noted insufficient evidence to assess the balance of benefit and harm in those 70 and older. The long-awaited ASPREE trial, conducted in healthy elderly aged 70 and older (65 for US minorities), evaluated aspirin's effect on disability-free survival, a composite of death, dementia, or persistent physical disability. CVD and cancer were prespecified secondary endpoints, positioning ASPREE's results to substantially inform the evidence gap noted in the USPSTF guidelines. Low-dose aspirin over 5 years did not lower CVD events or colorectal risk, but significantly increased bleeding. The ASPREE-XT observational follow-up study over the next 5-7 years will observe for potential legacy effects of aspirin on the primary and secondary outcomes of ASPREE, thus adding further evidence to define the risk-benefit profile of aspirin for primary prevention in healthy elderly.