Abstract
Background: The bioethics literature contains speculation but little data about sociodemographic differences between patients for whom ethics committees (EC) are consulted for conflict about life-sustaining treatment (LST) and the broader hospital population that these committees serve. To provide an empirical context for this discussion, we examined differences in five sociodemographic factors between patients for whom an EC was consulted for conflict over LST and the general inpatient population, hypothesizing that nonwhite patients were most likely to be disproportionately represented. Methods: This was a retrospective cohort study comparing self-reported race/ethnicity, household income, insurance status, primary language, and religious affiliation between patients involved in EC consultations for conflict over LST and the general adult inpatient population at a large academic hospital from January 1, 2007, through December 31, 2013. Results: There were 169 EC consultations involving conflict over LST and 243,197 adult inpatient admissions over the study period. On bivariate analysis, patients for whom the EC was consulted for conflict over LST were more likely to be nonwhite, low-income, and to speak a primary language other than English. They were not more likely to be underinsured or to report a religious affiliation. On multivariate analysis, patients for whom the EC was consulted for conflict over LST were more likely to be nonwhite (odds ratio [OR], 2.5; 95% confidence interval [CI], 1.7–3.7; p < .001). They were also more likely to be low income (OR, 1.8; 95% CI, 1.1–3.0; p = .02). They were not more likely to speak a primary language other than English (OR, 1.4; 95% CI, 0.9–2.1; p = .17). None of the assessed variables were associated with conflict over LST requiring additional meetings beyond the first EC consultation. Conclusions: Nonwhite and low-income status patients are disproportionately represented in EC committee consultation for conflict over LST compared to the general inpatient population.