DESCRIPTION (provided by applicant): Dr. Dunlay is a cardiologist specializing in the care of patients with heart failure (HF). Her long term goal is to become an independent health services and outcomes researcher in the field of HF. Her short term goal is to strengthen her research expertise and add new skills becoming proficient in the tools necessary to perform this type of research. Key components of her training plan during the K23 period include: * Adding to her existing skill set by doing the following: o Enhancing health services research knowledge through coursework o Mastering use of advanced data analysis techniques through coursework and practice o Expanding her knowledge of resources available to outcomes researchers o Developing end of life and survey research content expertise * Fostering existing and developing new collaborative relationships by presenting her work at national conferences, applying for participation in Early Career Committees, and working with internal and external experts in health services and outcomes research. * Publishing high impact results: Dr. Dunlay will publish 5 senior author papers resulting from the proposed aims and be a co-author on at least 8 additional papers demonstrating evidence of successful collaboration * Obtain RO1 level funding: Dr. Dunlay will submit RO1 proposals during the 3rd and 5th years of the award Dr. Dunlay is highly committed to making the successful transition to an independent clinical investigator during the K23 award period. The receipt of this award is critical to her attaining he goals. ENVIRONMENT The Mayo Clinic has a longstanding dedication to patient-centered medicine and patient-oriented research. Through its comprehensive infrastructure, the Mayo Clinic is both committed and highly capable of training physician investigators who will become future leaders in cardiovascular research. The Mayo Clinic has identified health services and outcomes research as an area of critical need and an institutional priority, as evidenced by the recent development of the Center for the Science of Healthcare Delivery, directed by Dr. V ronique Roger, mentor of this proposal. Dr. Dunlay has assembled a Career Advisory Committee of established experts and experienced mentors who are fully committed to her development into an independent clinical investigator during the K23 award period. Finally, these studies will be conducted under the auspices of the Rochester Epidemiology Project (REP), which provide a unique well-established framework for the performance of community-based health services research such as is being proposed herein. In summary, the institutional environment and mentoring team will ensure Dr. Dunlay's successful transition toward an independent career in health services and outcomes research. RESEARCH Background: Patients with advanced HF have high healthcare resource utilization including repeated hospitalizations that adversely affect their quality of life and are a major public health problem. However, it is unknown how end of life patient preferences and advance care planning contribute to differences in healthcare resource use in advanced HF. We will address these gaps in knowledge through the proposed studies. Specific Aims: First, we will examine healthcare utilization at the end of life and how it may be impacted by the type of HF (HFpEF, HFrEF), frailty, and use of palliative care and hospice services (Aim 1). Next, we will investigate how often advance care planning including patient-provider conversations regarding end of life occur (Aim 2). Finally, we will determine whether advance care planning and patient preferences contribute to differences in healthcare resource use, including hospitalizations (Aim 3). Study Design and Outcome Measures: The studies will be conducted in Olmsted County, MN under the auspices of the REP. First, among a population-based cohort of patients with HF who were prospectively enrolled into a study of the mentor of this proposal from 2003-2012, we will examine healthcare resource utilization (hospitalizations, outpatient and emergency department visits) in the last 6 months of life among those who died. Using Andersen-Gill models which account for repeated events, the risk of hospitalization at the end of life will be examined by type of HF (HFpEF vs. HFrEF) and frailty, adjusting for potential confounders. Differences in healthcare use among those who received palliative medicine consultation and hospice services will also be examined. Second, we will prospectively recruit consecutive Olmsted County residents hospitalized with HF and assess their end of life preferences and involvement in advance care planning using a face-to-face questionnaire and compare differences by type of HF and frailty. Finally, the enrolled patients will be followed and the association between end of life preferences and healthcare resource use will be determined using Andersen-Gill models. Potential Public Health Contribution: Data on end of life preferences in advanced HF and their association with resource utilization are lacking. The proposed aims will address these critical gaps in knowledge, thereby informing us on potential disparities in the care of HF patients that may be interfering with the shared decision- making process and, if improved, may lead to a reduction in hospitalizations and improved patient satisfaction Relationship to the NHLBI Mission: These studies aim to determine ways to properly assess and treat patients with HF so that they can live more fulfilling lives, in alignment with th NHLBI mission statement. Relationship to the Career Goals of the Candidate: The proposed studies will provide necessary advanced exposure to the performance of population-based health services research in HF and will be the foundation for the candidate to establish a career in HF health services and outcomes research.

PROJECT NARRATIVE Patients with advanced heart failure often experience repeated hospitalizations that adversely affect their quality of life and are a major public health problem, but whether patient preferences and advance care planning lead to differences in the risk of hospitalization is unclear. We will investigate how often advance care planning including patient-provider conversations regarding end of life occur in community patients with heart failure, and whether these factors contribute to differences in healthcare resource use, including hospitalization. The proposed aims will provide critical information to guide our efforts to improve the care of patients with heart failure to reduce hospitalizations and improve patient satisfaction.