Accounts of killings of civilians by police in the United States (U.S.) have attracted considerable public attention. In this study, using all civilian deaths (N = 1099) in the U.S. in 2015, compiled independently by The Washington Post and The Guardian newspapers, we identified characteristics of each interaction between the police and the deceased, such as whether the decedent was armed. We expanded the database to include systemic factors possibly related to these deaths, and examined death rates by demographics, presence of mental illness, and state-level predictors. Twenty-three percent (251 of 1099) of individuals killed during interactions with police in 2015 displayed signs of a mental illness. Race (African-American [RR = 2.57] compared to non-Hispanic Whites [95% CI 2.08-3.18]) and presence of mental illness (RR = 7.16 compared to no mental illness, 95% CI 6.21-8.25) were strongly associated with such fatalities. Individuals with mental illness were more likely to be armed with a knife (OR = 3.1, 95% CI 2.1-4.6), and were more likely to have been killed at home (OR = 2.8, 95% CI 1.9-4.0). The death rates for persons with evidence of mental illness during interactions with police are high. Our finding that many persons with mental illness were killed at home and were not brandishing a firearm suggests that more effective de-escalation methods might reduce the incidence of fatal outcomes.
Copyright © 2018 Elsevier Ltd. All rights reserved.