Do-Not-Resuscitation Orders Among Elderly Patients with Heart Failure
Open Access
- Author:
- Callahan, Katherine
- Graduate Program:
- Health Policy and Administration
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- March 13, 2022
- Committee Members:
- Chris Hollenbeak, Program Head/Chair
Chris Hollenbeak, Co-Chair & Dissertation Advisor
Lisa Kitko, Outside Unit & Field Member
Yubraj Acharya, Co-Chair of Committee
Melissa Hardy, Major Field Member - Keywords:
- advance care planning
end-of-life care
do-not-resuscitate orders
heart failure
elderly - Abstract:
- The purpose of this dissertation is to contribute to the discussion about the potential impact of the advance care planning (ACP) process on end-of-life care. This dissertation aims to achieve this through three research studies that explore documentation of a do-not-resuscitate (DNR) order and patient outcomes. In the first study, the association between end-of-life care and costs is explored in the context of increasing healthcare expenditures. Although the primary benefit of ACP is that it aligns patients’ wishes with the treatment they receive, a secondary benefit may be cost-savings. Data from the Healthcare Cost and Utilization Project’s National Inpatient Sample is used to examine whether a documented DNR order among 700,922 hospital inpatients with heart failure (HF) was associated with costs and other outcomes between the years of 2011 and 2016. The results indicate that the presence of a DNR among this population is associated with lower total costs, a shorter hospital stay, and a higher risk of mortality, particularly when a patient with a DNR in place dies. These findings add to the evidence that ACP may aid in containing health care costs, and policies that promote ACP may not only help patients get the care they prefer but may also lower healthcare costs The second study in this dissertation examines the validity of International Classification of Diseases (ICD) codes (V49.86 for ICD-9 and Z66 for ICD-10) to identify the presence of a DNR order. Elderly HF patients (N=1,719) that were admitted to the Penn State Milton S. Hershey Medical Center between 2013 and 2016 were included in this study. Merging data from the electronic medical record (EMR) and the cost-accounting database allowed a comparison between a DNR in ICD codes to a DNR present in the EMR. Measures of agreement and disagreement suggested that the ICD code for DNR presence in the ICD codes is valid and has excellent diagnostic characteristics, including sensitivity and specificity. In addition, when the ICD measures of DNR were used in analyses focused on estimating the association between DNR and outcomes, they provided similar estimates of association compared to DNR in the EMR. This supports the use of ICD codes to identify DNR in future research studies. In addition, the use of ICD codes to identify DNR will facilitate the study of ACP as ICD codes are more readily available in large, observational datasets. The third study of this dissertation uses data from the California Office of Statewide Health Planning and Development (OSHPD) to investigate the association between a recent payment rule that provides reimbursement for physicians to have ACP conversations with their patients and DNR use among racial and ethnic minorities—groups with well-documented disparities in end-of-life care. In addition, this study examines the timing of DNR orders among racial and ethnic minority groups, as early DNRs are thought to distinguish cases in which the DNR reflects patient preference regarding resuscitation on admission from those in which the DNR order results from failed salvage treatment. The results of this study suggest that although the reimbursement policy was associated with an increase in ACP participation overall, the rate of increase of DNR participation was lower for racial and ethnic minorities. This paper adds to the discussion about how to best decrease disparities in end-of-life care and suggests that reimbursement incentives alone do not appear to be enough to reduce racial and ethnic disparities in DNR participation. Future interventions may need to be multilevel and specific to the targeted group they are intended to benefit from.