Use of Do-Not-Resuscitate Orders for Critically Ill Patients with ESKD

Significance Statement Despite having high rates of comorbid illnesses and significantly shortened life expectancy, patients with ESKD may be unrealistically optimistic about their own prognoses. In their study of do-not-resuscitate (DNR) orders upon admission to intensive care in a large sample of critically ill patients from hospitals across the United States, the authors observed more frequent DNR orders among patients with diseases of other major organs compared with patients without major organ disease. However, despite overall higher rates of comorbidity and hospital mortality in patients with ESKD, they found no such difference in the use of DNR orders between patients with and without ESKD. The lower use of DNR orders in this highly comorbid group might reflect a missed opportunity to avoid overmedicalization and unnecessary patient suffering at the end of life. Background Despite having high comorbidity rates and shortened life expectancy, patients with ESKD may harbor unrealistically optimistic expectations about their prognoses. Whether this affects resuscitation orders is unknown. Methods To determine whether do-not-resuscitate (DNR) orders differ among patients with ESKD compared with other critically ill patients, including those with diseases of other major organs, we investigated DNR orders on admission to intensive care units (ICUs) among 106,873 patients in the United States. Results Major organ disease uniformly associated with increased risk of hospital mortality, particularly for cirrhosis (adjusted odds ratio [aOR], 2.67; 95% confidence interval [95% CI], 2.30 to 3.08), and ESKD (aOR, 1.47; 95% CI, 1.31 to 1.65). Compared with critically ill patients without major organ disease, patients with stroke, cancer, heart failure, dementia, chronic obstructive pulmonary disease, and cirrhosis were statistically more likely to have a DNR order on ICU admission; those with ESKD were not. Findings were similar when comparing patients with a single organ disease with those without organ disease. The disconnect between prognosis and DNR use was most notable among Black patients, for whom ESKD (compared with no major organ disease) was associated with a 62% (aOR, 1.62; 95% CI, 1.27 to 2.04) higher odds of hospital mortality, but no appreciable difference in DNR utilization (aOR, 1.06; 95% CI, 0.66 to 1.62). Conclusions Unlike patients with diseases of other major organs, critically ill patients with ESKD were not more likely to have a DNR order than patients without ESKD. Whether this reflects a greater lack of advance care planning in the nephrology community, as well as a missed opportunity to minimize potentially needless patient suffering, requires further study.