Healthcare Decision-Making and Outcomes for People Living with Alzheimer's Disease

There are unique aspects and potential barriers to providing healthcare to people living with Alzheimer’s Disease and related dementias (ADRD), since they may be less capable than others of communicating their symptoms and health histories or following through independently on prescribed treatments. People living with ADRD often rely on family members and other caregivers for support. The co-occurring presence of ADRD, particularly in advancing and later stages, may also alter the decision-making calculus surrounding other diagnoses, such as cancer or cardiac treatment. In each of five analytic components, the Program Project considers how healthcare and health outcomes vary between patients with and without ADRD, and across the continuum of ADRD progression. 

Project 1 (Baicker and Wright) tests a machine learning tool called HEARTSPOT that draws from the electronic medical records of a massive database of past patients to help clinicians decide whether to recommend invasive treatment for patients who arrive in an emergency department with acute coronary syndrome. Project 2 (Simon) looks at the differences in healthcare use and health outcomes that result when the provider is of the same race as the patient, as compared with when they are of different races. Project 3 (Baicker, Handel, and Kolstad) uses high frequency audit log data to drill down on the complex process of diagnosis and treatment decision-making in emergency departments, and how decisions are influenced by not just biomedical factors, but also the clinical environment, patient conditions and characteristics, team interactions, clinician fatigue and stress, and information flow. Project 4 (Gottlieb and Layton) analyzes the causal effect of administrative hassles on healthcare access and outcomes, including billing hassles, prior authorization restrictions, complex applications for public insurance programs, and eligibility redeterminations for these programs. Project 5 (Sacarny) examines how Medicare payment rates affect the quality and safety of care for skilled nursing patients, the mechanisms like staffing that drive these effects, "spillover" effects onto long-stay patients, and the implications of payment changes for disparities in care access and quality. In each context, the co-occurring presence of ADRD may impact the decision-making process, the healthcare decisions made, and the health outcomes associated with those decisions.

In addition to its analytic components, the Program Project includes an administrative core, a data core, and a measurement core that seeks to improve how we identify individuals with ADRD, and how we track the progression of ADRD from early symptoms of mild dementia, through the stages of preclinical and clinical cognitive decline in existing and new databases.

This set of projects is the latest phase of a Program Project on health and well-being at older ages that the NBER has hosted for several decades. Information on some of the sub-projects that were supported by earlier Program Projects may be found on the Project History page.