Our Research Projects

Co-investigators: Roy Adams, PhD

Care partners, unpaid family or friends who partner with patients to manage care, are at the forefront of Alzheimer’s disease and related dementias care. Yet, care partners often go unidentified and unsupported by healthcare systems, which compounds the demands of Alzheimer’s disease and related dementias on patients, clinicians, and care partners. In response to this need, we will develop and test “Smart Rep”, Smart Automation to REspond to Patient portal messages, which sends messages in response to patient portal messages to both increase documentation of care partners in the electronic health record, and support care partners with content and resources relevant to their messaged concerns.

This project which will involve diverse methods to analyze claims data from FAIR Health (which holds the nation’s largest collection of private healthcare claims) and Medicare claims (subject to approval by the Centers for Medicare & Medicaid Services [CMS]), will illuminate the economic costs and differences in utilization of the range of healthcare services provided to patients with Alzheimer’s disease and related dementias (ADRD) by geographic area and a host of demographic factors, including race and ethnicity.

By illuminating salient treatment patterns and costs among different populations, the study will result in innovative, actionable insights that can be used by stakeholders to address gaps in understanding regarding the extent to which and how ADRD care is delivered to and paid for among diverse populations. Through a rigorous, innovative claims-based data analysis utilizing Medicare Advantage, private claims and Medicare data to examine services, costs and comorbid conditions associated with ADRD, the study will inform a wide range of new reimbursement, care delivery and payment approaches for a variety of healthcare stakeholders (e.g., clinicians, payors, policy makers, consumers and others) at the micro and macro levels.

Accurate claims-based identification of ADRD diagnoses is important for clinical care, population health management, and risk adjustment. In Medicare Advantage, capitated payments may lead to greater care coordination, chronic care management, and lower utilization of care, including for patients with ADRD. However, capitated payments also strongly incentivize MA insurers to increase coding intensity for prospective risk-adjusted payment levels. In this project, we aim to compare the accuracy of diagnostic codes in Medicare Advantage and Traditional Medicare to identify ADRD in the period between the removal and reintroduction of Medicare risk adjustment for ADRD. Our study will help to develop more accurate diagnostic coding and risk adjustment, which are essential to improving clinical care for patients with ADRD.

Co-investigators: Dan Polsky, PhD, Julia G. Burgdorf, PhD

Home health care services has provided crucial health services for persons living with dementia (PLWD) living in the community. Prior work has demonstrated that PLWD use more intensive home health and at greater costs than persons without dementia. However much of our understanding of home health use by PLWD is in the context of traditional Medicare. Given the recent growth in Medicare Advantage and the fundamental differences in the reimbursement models, it is critical to understand how PLWD use home health care with Medicare Advantage. The purpose of this project is to explore differences in home health care use by PLWD between traditional Medicare and Medicare Advantage. This pilot will leverage linked traditional Medicare claims, Medicare Advantage encounter data, and clinician assessments in a national sample of PLWD. We will address essential literature gaps by examining details of the home health episodes (e.g. services used, length of stay) and exploring variations across different Medicare Advantage plan types. Findings will provide an important foundation for future research on the accessibility, quality, and equity of home health care for PLWD.

Co-investigators: Ge Bai, PhD, Jeromie Ballreich, PhD, Jenny Markell, PhD Student

The Medicare program is the primary source of health care spending for the 6.5 million Americans with Alzheimer’s disease and related dementias (ADRD), accounting for 45% of all ADRD health care-related cost, or $146 billion, in 2022. Patients with ADRD require several pharmacologic treatments to maintain their health and quality of life, as well as to treat underlying comorbidities. This study will examine how the use of pharmaceutical treatments through mail order and retail pharmacies affects drug accessibility and affordability for Medicare beneficiaries with ADRD.

Co-investigators: Cynthia Boyd, MD, Theodore Jack Iwashyna, MD, Jessica Ma, MD, Megan Shepherd-Banigan, PhD, John Pura, PhD, Jennifer Wolff, PhD

Among non-veteran persons living with dementia, the use of mechanical ventilation has doubled in recent years. This growth occurred without an increase in survival and despite evidence that surrogate decision makers of persons with advanced dementia almost universally desire comfort, rather than life prolongation, to be the patient’s goal of care. Trends in mechanical ventilation have not been studied among veterans living with dementia, despite the high prevalence of dementia among veterans over the age of 65. In addition, no study has investigated the association of palliative care use and mechanical ventilation for veterans living with dementia, though prior work has shown that use of palliative care is associated with decreased critical care resource use. In this study, we will construct a national cohort of veterans living with dementia to examine trends in mechanical ventilation and explore the association between receipt of palliative care and mechanical ventilation in this population.

A significant share of long-term care needs related to ADRD are met informally by family members, many of whom also work, and the increasing demand for care has potential to impede economic growth by reducing formal employment among the working-age population. A full accounting of the economic burden of ADRD must include the substantial spillovers in the form of reduced labor force participation and lost earnings of family caregivers. In this project, we will study the labor supply response to ADRD-related caregiving among adult men and women in Denmark and document the distribution of economic costs within families based on demographic characteristics. Our study will contribute to a greater understanding of the substantial, but often difficult to measure economic consequences that ADRD imposes on families and on national budgets pointing to targeted policies to mitigate these costs.

My research seeks to improve our understanding of the determinants of cognitive decline. I propose to use economic modeling to address two fundamental problems in this literature: dynamic selection and measurement error. Dynamic selection occurs when the composition of data changes over the age profile due to mortality. For example, dynamic selection biases the relationship between cigarette smoking and cognitive health because smokers experience excess mortality. Measurement error occurs when we use proxy measures (e.g., word recall scores) in the place of latent variables (e.g., cognitive health). Economists have developed statistical models for both dynamic selection and measurement error, but my research is the first to apply these methods to cognitive health and its determinants.

The purpose of this project is to understand the role of care partners in health care in the context of caring for an older adult with hearing loss and dementia. Caring for the growing number of older adults with dementia presents complex medical and communication challenges. These challenges may be compounded in the context of hearing loss where the additional burden of hearing loss among older adults with dementia may place added stressors on cognitive reserve, hinder communication further, and create additional emotional/social burdens on care partners. With this work, we aim to understand how care partners influence health management of older adults with hearing loss and/or dementia to inform future work to improve communication between older adults with dementia, care partners, and clinicians — critical for the health and social/emotional support of older adults and their care partners.

Hip fractures are a significant cause of morbidity and mortality in older adults and have a substantial impact on health care service utilization and cost. Each year over 300,000 older adults are hospitalized for hip fracture in the United States with an estimated annual direct cost of $17 billion dollars. Given the increase in older adults nationally, these numbers are expected to grow. The risk of hip fracture is nearly three times higher in older adults with dementia owing to higher rates of osteoporosis and a greater propensity for falls. Nearly 40% of older adults with hip fracture are cognitively impaired and are more likely to experience post-operative complications, longer lengths of hospital stay, hospital readmission and greater loss of independent function. This is a vulnerable group that require a higher treatment focus and more resources to meet their complex needs, though little is known about the incremental service utilization and cost associated with dementia for hip fracture. Therefore, we will compare health care service utilization and cost for hip fracture among older adults with versus without dementia using 20% Medicare FFS claims and will determine if dementia is predictive of higher cost independent of standard risk adjustment variables. This research has important implications for future bundled payment policy and the provision of value-based care for these individuals.

Older adults with dementia go to the Emergency Department (ED) more frequently than older adults without dementia. The ED is a potentially overstimulating environment for older adults with dementia given the background noise, large number of unfamiliar people and interrupted opportunities for sleep. Further, the ED is associated with higher costs of care for older adults with dementia compared to those without dementia, placing financial strain on patients, families and society. Given these risks, reducing the time spent in the ED, particularly from long extremes is important for older adults with dementia. In this study we will determine if length of stay in the ED is longer for older adults with dementia and if there are any facility characteristics associated with longer lengths of stay for older adults with dementia. These results will serve as important foundational information that could be used to target ED dementia care interventions. 

The purpose of this project is to use data on genetic propensities to develop ADRD to better understand the life course trajectories of people who may eventually develop ADRD, including their educational attainment, work life, income, financial decisions, and wealth accumulation. Of particular interest are individuals who are not diagnosed with ADRD or who do not exhibit significant cognitive decline according to standard measures (either because they lack access to medical care or, alternatively, because their condition is mild enough or their environment protective enough that cognitive decline falls under the radar). Using genetic data to better understand people who may not exhibit cognitive decline using standard measures or who are not diagnosed can help us to better understand the full extent of ADRD. Preliminary results suggest that many people with a genetic propensity for ADRD and who are not diagnosed and who do not exhibit strong cognitive decline according to standard measures still exhibit other negative outcomes (e.g., earlier retirement or less wealth) suggesting that the ways we currently diagnose or recognize illness are not sufficient to capture the full population in need of treatment. Findings will not only help us to recognize more potential patients, but may also help us to understand treatment. If some people with a genetic propensity for ADRD do not exhibit symptoms, it may be by chance, but may also be because policy-modifiable environmental factors are protective. Understanding what these factors are and why they are protecting could eventually be used to inform best treatment practices. 

We will examine 12-month health care use patterns among Veterans 65+ with pre-existing mental illness and a new Alzheimer’s Disease and Related Dementias (ADRD) diagnosis. Specifically, we will examine the extent to which pre-existing mental illness is a risk factor for poor quality care for individuals with ADRD. The incidence of ADRD is projected to increase exponentially in the coming decades and mental illness is associated with a 2-fold higher risk for developing ADRD. Co-occurrence of ADRD and mental illness likely complicates health care delivery for this population, yet there is little understanding of the implications of this co-occurrence on service utilization patterns which drive Medicare costs. Findings from this study will provide information about the extent to which pre-existing mental illness impedes high quality care for older adults with ADRD, as evidenced by higher ED use, inpatient visits, and 30-day readmissions. This information will provide a foundation upon which to develop health system strategies that improve care coordination and quality for this understudied population. 

Dementia and sensory impairment are salient features of aging that share common risk factors, and sensory impairment is itself a risk factor for cognitive decline. These commonly co-occurring conditions may be associated with impairment in financial capacity, putting older adults at risk for financial loss and fund mismanagement. In addition, caring for older adults with chronic conditions may be associated with financial difficulties among caregivers. Financial capacity is central to older adults’ independence and well-being, and their caregivers’ welfare and capacity for caregiving. Therefore, a deeper understanding of the interplay between dementia and sensory impairment and the impact of the resultant cumulative risk on financial wellbeing of older adults and their caregivers is required. This project aims to examine the economic impact (income, economic wellbeing, financial literacy) of having co-occurring dementia and sensory impairment (vision and hearing impairments) on older adults, and their family caregivers, using nationally representative data. The resulting findings will highlight gaps and opportunities for the development of supportive services and policies to help older adults with financial management. 

Alzheimer’s Disease Related Dementia (ADRD) is poorly coded in electronic health records (EHRs). To address this gap, this study aims to assess the reliability of ADRD coding in EHRs, and then compare the value of unstructured EHR data in identifying patients with ADRD. Results of this study can enhance our understanding of EHR’s data value in accurately identifying ADRD patients. Study conclusions may improve the use of EHRs to assess patient eligibility in clinical trials and denominator selection for health services research projects targeting the older adults. 

The purpose of this project is to determine whether the use of supportive services varies by race for black and white caregivers of persons with Alzheimer’s Disease and Related Dementias (ADRD). Persistent health disparities experienced by Black persons living with dementia (PLWD) likely place an added demand on family and unpaid caregivers. There is limited understanding of supportive services accessed by caregivers of PLWD and how utilization may vary by caregiver and care recipient characteristics, particularly across race groups. More information is needed to address the needs of this growing population, inform the design of relevant and effective programs and practices, and increase the capacity of providers and policymakers to serve diverse groups of people impacted by ADRD. Findings from this pilot will provide important information about the current landscape of support for racially diverse ADRD caregivers and better inform strategies to help persons with ADRD and their family and unpaid caregivers. 

Dementia represents a threat to retirement security for patients and families because characteristic brain changes make it difficult to remember routine financial characteristics and alter risk perception, increasing susceptibility to fraud and exploitation. Yet little is known about the prevalence and magnitude of financial losses due to dementia and the potential unmet need for assistance with financial management among persons with Alzheimer’s Disease and Related Dementias (ADRD). With no public and/or private sector policies to protect the financial interests of persons with ADRD, there is an urgent need to understand the magnitude of the issue and develop methods to monitor incidence over time so that effective policies can be developed. This pilot project will collect new survey data to assess the prevalence and magnitude of financial losses associated with cognitive impairment and to assess the demographic characteristics of patients and families affected by these losses. 

The purpose of this pilot study is to contribute new and more robust estimates of state-level measures of Alzheimer’s Disease and Related Dementias (ADRD) prevalence and potential variability in prevalence from multiple national datasets (American Community Survey, Behavioral Risk Factor Surveillance System, National Health Interview Survey, and Medicare Claims) with both self-reported measures and diagnosed ADRD. The evidence generated by this research will directly support Drabo’s planned career development application to NIA around ADRD diagnosis and care burden. The anticipated outcome from this proposed training plan and research agenda is to contribute stronger and more robust information regarding state-level estimates of ADRD prevalence to set the stage for further work to quantify ADRD burden, and to permit dementia-related research in various rich, less restrictive, and relatively user-friendly data sources. 

Older adults with Alzheimer’s Disease and Related Dementias (ADRD) have uniquely challenging care needs, large comorbidity burdens, and are heavy users of health care; they comprise an estimated one-third of the patient population in Medicare-funded home health care. Current risk-adjustment models for Medicare home health reimbursement fail to account for whether a patient has ADRD. This omission has the potential to threaten access to care for this vulnerable subpopulation. Using linked survey, patient assessment, and claims data, this pilot will provide the first evidence regarding the relationship between patient cognitive impairment and costs of care during Medicare-funded home health care. Findings will offer a better understanding of the resources needed to provide skilled care at home for older adults with ADRD and may inform future revisions to the Medicare home health reimbursement model.