Increasing geriatric population and patients with multiple chronic conditions have exacerbated the crisis of a rapid increase in number of Alzheimer?s disease (AD) patients. AD is the primary cause of dementia, and is responsible for over eleven million disability-adjusted life years (DALY) globally. Overwhelming evidence indicates that progression of AD is associated with increased dependency, resource utilization, and costs of care. The diagnosis of AD and onset of dementia not only places an emotional burden on patients, but the costs associated with Alzheimer?s care also places substantial financial burden on families, who often must make significant changes in their economic lifestyles to extend care. To mitigate the health, emotional, and financial burden associated with AD, we need to identify those at risk long before they develop symptoms of dementia. Early symptoms of dementia, such as memory impairment, may not be apparent during the current practice of primary care visits, unless directly assessed or adequately informed. Hence, it is not surprising that many studies reveal delayed and/or undocumented diagnosis of dementia among primary care providers. To address this critical unmet need, we propose to use machine learning and advanced predictive analytics and perform retrospective and prospective analysis of data, including clinical data from existing patient medical records, financial data from reimbursement claims, and demographic data from population health records to identify clinical, co-morbid, pharmaceutical, environmental, and other variables related to onset and progression of dementia. The overarching goal of the present study is to develop a comprehensive decision support tool that would predict cost drivers, outcome measures and trajectory of dementia onset in AD patients. To achieve our goals, we are working in collaboration with Chris Craver, Senior Director, Health Data Analytics at Vizient Inc., and Dr. Ronald Petersen, Director of Alzheimer?s Disease Research Center at Mayo Clinic Rochester. We have now curated data from over 72,000 AD patients and 100,000 age-matched controls to date, and are now proposing to pursue studies with the following specific aims.
Aim #1 : Using descriptive analytics, create an interactive, secure, web-based interface to enable benchmarking of key metrics for AD and AD dementia, including population health metrics, co-morbidities and risk factors, therapeutic drug regimens, and the true costs of episodes of care.
Aim #2 : Develop an electronic health record (EHR) agnostic, HIPAA- compliant tool for predicting the development of AD-dementia, its disease trajectory and outcomes, medication utilization, total annual cost of care, and any diagnostics or therapeutics to augment adverse events. Successful completion of studies proposed in this application would enable physicians and care providers to better understand their AD patients with dementia. Additionally, product developed through this grant study would benefit key stakeholders in healthcare market place, including Medicare, Medicaid, commercial insurers, hospitals and health systems, in addition to AD patients and caregivers.

Public Health Relevance

While Alzheimer?s disease (AD) and aging have several common pathogenic events and underlying risk factors, diagnosis of AD not only affects the patient but also places enormous emotional and financial burden on families and care providers. There remains an urgent need to reduce this burden by providing information, in advance, on the course of AD and costs associated with care. The work proposed in this application seek to address this problem by using advanced analytics and machine learning techniques, and identifying patients at risk of cognitive decline and dementia. By empowering physicians, care providers, insurers and health systems with relevant patient information, we will delineate trajectory of the disease, and help refine the current guidelines for dementia. Our rationale for such a comprehensive approach is that by augmenting clinical interventions and reducing inefficient resource consumption, we will advance value-based care.

Agency
National Institute of Health (NIH)
Institute
National Institute on Aging (NIA)
Type
Small Business Innovation Research Grants (SBIR) - Phase I (R43)
Project #
1R43AG058480-01
Application #
9465969
Study Section
Special Emphasis Panel (ZRG1)
Program Officer
Hsiao, John
Project Start
2017-09-15
Project End
2018-08-31
Budget Start
2017-09-15
Budget End
2018-08-31
Support Year
1
Fiscal Year
2017
Total Cost
Indirect Cost
Name
Aprihealth, Inc.
Department
Type
DUNS #
079317296
City
Dallas
State
TX
Country
United States
Zip Code
75248