Abstract 115: Resuscitation Preferences of Older Patients Hospitalized with Heart Failure: Temporal Trends and Associations with Patient and Hospital Factors

Background: Preferences regarding end-of-life (EOL) care in patients (pts) with HF may be influenced by personal, cultural, and health system factors. We examined characteristics associated with cardiopulmonary resuscitation preferences among pts hospitalized for HF and explored whether rates of Do Not Resuscitate (DNR) status have changed over time. Methods: Using the California State Inpatient Databases from the Healthcare Cost and Utilization Project and the American Hospital Association Annual Survey Databases, we identified 297,156 pts aged ≥ 60 years hospitalized for HF between 2004 and 2011. The first eligible hospitalization for each patient was selected. DNR status documented within the first 24 hours of admission was assessed. Multivariable logistic regression was used to evaluate associations of DNR status with demographic factors, year of admission, median household income quartile, and hospital teaching status. Results: 39,658 (13.4%) pts had a DNR order. Higher proportions with DNR were found among women than men (15.4% vs. 11.0%, p<.0001) and white pts than non-white pts (16.7% vs. 7.3%). Proportions of pts with DNR increased over the study period for both men and women (from 10.0% to 12.3% and 13.7% to 17.2%, respectively, p<.0001 for trend with time) and in both white pts (15.0% to 19.2%) and non-white pts (6.5% to 7.9%), but more rapidly for white pts (p=.02). In multivariable analysis adjusting for age, gender, race, and year of admission, odds ratios (ORs) for DNR were 1.5 per 5-year increase in age, 1.2 for women vs. men, and 2.0 for white vs. non-white pts. After adding income and hospital teaching status to the model, the respective ORs associated with each incremental increase in income quartile and receiving treatment in a teaching hospital were 1.1 and 1.2 (p<.0001 for both). ORs for age, gender, race, income, and hospital teaching status were essentially unchanged after adjustment for comorbidities. Conclusion: DNR status among pts hospitalized with HF is influenced by age, gender, race, and income in addition to hospital teaching status. The proportions of pts with DNR status have increased, but unequally among demographic subgroups. Continued efforts to understand the factors that influence preferences around resuscitation and EOL care are needed to provide truly patient-centered care. ![][1] [1]: /embed/graphic-1.gif