Epidemiology of Advance Directives in Extended Care Facility Patients Presenting to the Emergency Department

INTRODUCTION: We conducted an epidemiologic evaluation of advance directives and do-not-resuscitate (DNR) prevalence among residents of extended care facilities (ECF) presenting to the emergency department (ED). METHODS: We performed a retrospective medical record review on ED patients originating from an ECF. Data were collected on age, sex, race, triage acuity, ED disposition, DNR status, power-of attorney (POA) status, and living will (LW) status. We generated descriptive statistics, and used logistic regression to evaluate predictors of DNR status. RESULTS: A total of 754 patients over 20 months met inclusion criteria; 533 (70.7%) were white, 351 (46.6%) were male, and the median age was 66 years (IQR 54–78). DNR orders were found in 124 (16.4%, 95% CI [13.9–19.1%]) patients. In univariate analysis, there was a significant difference in DNR by gender (10.5% female vs. 6.0% male with DNR, p=0.013), race (13.4% white vs. 3.1% non-white with DNR, p=0.005), and age (4.0% <65 years; 2.9% 65–74 years, p=0.101; 3.3% 75–84 years, p=0.001; 6.2% >84 years, p<0.001). Using multivariate logistic regression, we found that factors associated with DNR status were gender (OR 1.477, p=0.358, note interaction term), POA status (OR 6.612, p<0.001), LW (18.032, p<0.001), age (65–74 years OR 1.261, p=0.478; 75–84 years OR 1.737, p=0.091, >84 years OR 5.258, P<0.001), with interactions between POA and gender (OR 0.294, P=0.016) and between POA and LW (OR 0.227, p<0.005). Secondary analysis demonstrated that DNR orders were not significantly associated with death during admission (p=0.084). CONCLUSION: Age, gender, POA, and LW use are predictors of ECF patient DNR use. Further, DNR presence is not a predictor of death in the hospital.