Predictors of do not resuscitate orders in the nursing home.

BACKGROUND The outcome of cardiopulmonary resuscitation of residents of long-term care facilities is poor. However, only about one half of residents of long term care facilities have a do not resuscitate (DNR) order. The remainder usually have resuscitation by order or by default policy. Understanding predictors of DNR may help clinicians address end-of-life issues with the older long-term care population. OBJECTIVES To determine (1) the prevalence of DNR orders, and (2) predictors of DNR orders in older institutionalized individuals in a large community teaching nursing home. METHODS A cross-sectional chart review study of 177 consecutively located older patients from an 899-bed academic long-term care facility. RESULTS The prevalence of a DNR order was 40%. The frequency of ordering DNR was greater in subjects who were 85 years or older compared with subjects who were younger than 85 years (57% vs. 30%, P < .05). Ordering DNR was associated significantly with race (49% of whites compared with 13% African Americans, P < .05) but not with sex. Subjects with a DNR order were more likely to have been diagnosed with depression (52% vs. 35%, P < .05) but not dementia, and overall had greater number of medical conditions (5.9 +/- 2 vs. 5.1 +/- 2, P < .05) compared with subjects without DNR orders. The frequency of DNR orders did not significantly differ between subjects who were able to ambulate (with or without assistance) compared with subjects who were wheelchair or bed bound. Using logistic regression analysis, only age (with a B of -1.04 and P of .017) and race (with a B of 1.4 and a P of .01) were independent predictors of DNR status. CONCLUSION Fewer than half of this sample of long-term care residents had a DNR order. Among seven factors studied, only age and race were independent predictors of DNR status in the nursing home.