Timing of do-not-resuscitate orders and health care utilization near the end of life in cancer patients: a retrospective cohort study

Purpose: The objectives are to explore the prevalence of DNR orders, the factors influencing them, and the association between DNR signing and health care utilization among advanced cancer patients. Methods: This was a retrospective cohort study. Data from cancer decedents in three hospitals in China from January 2016 to December 2017 during their last hospitalization before death were obtained from the electronic medical records system. Results: In total, 427 cancer patients were included; 59.0% had a DNR order. Patients who had solid tumors, lived in urban areas, had more than one comorbidity, and had more than five symptoms were more likely to have DNR orders. The cut-off of the timing of obtaining a DNR order was 3 days, as determined by the median number of days from the signing of a DNR order to patient death. Patients with early DNR orders (more than 3 days before death) were less likely to be transferred to the intensive care unit and undergo cardiopulmonary resuscitation, tracheal intubation, and ventilation, while they were more likely to be given morphine and psychological support compared with those with late (within 3 days before death) and no orders. Conclusions: Advanced cancer patients with solid tumors living in urban areas with more symptoms and comorbidities are relatively more likely to have DNR orders. Early DNR orders are associated with less aggressive procedures and more comfort measures. However, these orders are always signed late. Future studies are needed to better understand the timing of DNR orders.