Association between physician orders for life-sustaining treatment for scope of treatment and in-hospital death in Oregon

Objectives To examine the relationship between Physician Orders for Life-Sustaining Treatment (POLST) for Scope of Treatment and setting of care at time of death. Design Cross-sectional. Setting Oregon in 2010 and 2011. Participants People who died of natural causes. Measurements Oregon death records containing cause and location of death were matched with POLST orders for people with a POLST form in the Oregon POLST registry. Logistic regression was used to measure the association between POLST orders and location of death. Results Of 58,000 decedents, 17,902 (30.9%) had a POLST form in the registry. Their orders for Scope of Treatment were comfort measure only, 11,836 (66.1%); limited interventions, 4,787 (26.7%); and full treatment, 1,153 (6.4%). Comfort measures only (CMO) orders advise avoiding hospitalization unless comfort cannot be achieved in the current setting; 6.4% of participants with POLST CMO orders died in the hospital, compared with 44.2% of those with orders for full treatment and 34.2% for those with no POLST form in the registry. In the logistic regression, the odds of dying in the hospital of those with an order for limited interventions was 3.97 times as great (95% CI = 3.59-4.39) as of those with a CMO order, and the odds of those with an order for full treatment was 9.66 times as great (95% CI = 8.39-11.13). Conclusions The association with numbers of deaths in the hospital suggests that end-of-life preferences of people who wish to avoid hospitalization as documented in POLST orders are honored.