Changes over time in POLST use and content by race and ethnicity among California nursing home residents

Abstract

Background

Physician Orders for Life‐Sustaining Treatment (POLST) are commonly used for nursing home (NH) residents. Treatment orders differ across race and ethnicity, presumably related to cultural and socioeconomic variation and levels of access to care and trust. Because national efforts focus on addressing the underpinnings of racial and ethnic differences in treatment (i.e., access to care and trust), we describe POLST use and content by race and ethnicity.

Methods

California requires NHs to document POLST completion and content in the Minimum Data Set. We describe POLST completion and content for all California NH residents from 2011 to 2016 ( N = 1,120,376). Adjusting for resident characteristics, we compared changes in completion rate and differences by race and ethnicity in POLST content—orders for cardiopulmonary resuscitation (CPR), do not resuscitate (DNR), CPR with full treatment, DNR with selective treatment or comfort orders, and if unsigned.

Results

POLST completion increased across all racial and ethnic groups from 2011 to 2016; by 2016, NH residents had a POLST two‐thirds or more of the time. In 2011, Black residents had a POLST with a CPR order 30.4% of the time, Hispanic residents 25.6%, and White residents 19.7%. By 2016, this grew to 42.5%, 38.2%, and 28.1%, respectively, with Black and Hispanic residents demonstrating larger increases than White residents ( p < 0.001). Increases over time in POLST with CPR and full treatment were greater for Black and Hispanic residents compared to White residents. The increase in POLST with DNR and DNR with Selective treatment and Comfort orders was greater for White compared to Black patients ( p < 0.001). Unsigned POLST with CPR and DNR orders decreased across all racial and ethnic groups.

Conclusions

Racial and ethnic differences in POLST intensity of care orders increased between 2011 and 2016 suggesting that efforts to mitigate factors underlying differences were ineffective. Studies of newer POLST data are imperative.