Impact of Care Delivery and Therapeutic Changes on Patients with Alzheimer's Disease and Alzheimer's Disease Related Dementias

Recently the Medicare program introduced policies that permitted greater flexibility in the provision of outpatient and inpatient care, including relaxing rules governing the remote delivery of care. Unfortunately, there is limited information on how the changes in care delivery made possible by these policies have impacted patient outcomes especially in the setting of ongoing changes in the diagnosis of Alzheimer’s Disease and Alzheimer’s Disease Related Dementias (AD/ADRD) and with new therapeutic options. In this project, we will examine the impact of changes in outpatient and inpatient care on clinical event rates and deaths among older patients with AD/ADRD living in the community. These patients are particularly vulnerable to disruptions in their care, and might not be able to articulate their needs. Those with yet additional disadvantages such as frailty or lower incomes may be more susceptible to adverse effects because of greater need or lower capacity to respond. Some care delivery changes, however, might have increased the number or quality of clinician interactions with patients, e.g., more frequent tele-health visits because of less travel. Thus, to examine the impact of the changes in care delivery under the Medicare policies, we will address three aims: 1) To assess and refine study variable definitions in the setting of evolving diagnostic and therapeutic options; 2) To examine the impact of changes in outpatient care on clinical event rates, e.g., emergency department and hospitalization rates; and 3) To examine the impact of changes on mortality. Notably, changes in visit and referral patterns in recent years could impact information capture; thus, diagnosis-based definitions of AD/ADRD status, frailty, or delirium developed in the earlier time periods could become progressively less accurate. We will examine these care patterns, assess the accuracy of claims-based measures, and develop prediction models using machine learning methods and linked, overlapping datasets; we also will account for temporal variation in care delivery at the county level. Given these complex data and analytic issues, we will work with NIH to improve the data resources and methods for this type of work. The information from this large natural experiment is critical for preparing for future shocks to the health system, to determine which of the Medicare policies, if any, should be extended, and to inform debates concerning the balance between local flexibility and national standards within Medicare.