Emergency Department Visits Prior to Suicide and Homicide: Linking Statewide Surveillance Systems

• Published in: Crisis : the journal of crisis intervention and suicide prevention 2016, Vol.37 (1), p.5-12
• Main Authors: Cerel, Julie, Singleton, Michael D, Brown, Margaret M, Brown, Sabrina V, Bush, Heather M, Brancado, Candice J
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Cause of Death; Databases, Factual; Death and Dying; Emergency Service, Hospital - utilization; Emergency Services; Epidemiological Monitoring; Female; Homicide; Homicide - prevention & control; Homicide - statistics & numerical data; Human; Humans; Information Storage and Retrieval; Insurance, Health - statistics & numerical data; Male; Medicaid - statistics & numerical data; Medicare - statistics & numerical data; Mental Disorders - epidemiology; Odds Ratio; Poisoning - epidemiology; Risk Assessment; Risk Factors; Suicide; Suicide - prevention & control; Suicide - statistics & numerical data; United States - epidemiology; Veterans - statistics & numerical data; Violence; Wounds and Injuries - epidemiology; Young Adult
• ISSN: 0227-5910; 2151-2396
• DOI: 10.1027/0227-5910/a000354

Background: Emergency departments (EDs) serve a wide range of patients who present at risk of impending suicide and homicide. Aims: Two statewide surveillance systems were probabilistically linked to understand who utilizes EDs and then dies violently within 6 weeks. Method: Each identified case was matched with four randomly selected controls on sex, race, date of birth, resident zip code, and date of ED visit vs. date of death. Matched-pair odds ratios were estimated by conditional logistic regression to assess differences between cases and controls on reported diagnoses and expected payment sources. Results: Of 1,599 suicides and 569 homicides in the 3-year study period, 10.7% of decedents who died by suicide (x = 13.6 days) and 8.3% who died by homicide (x = 16.3 days) were seen in a state ED within 6 weeks prior to death. ED attendees who died by suicide were more likely to have a diagnosis of injury/ poisoning diagnosis or mental disorder and more likely to have Medicare. Those who died by homicide were more likely to have a diagnosis of injury/poisoning and less likely to have commercial insurance. Conclusion: It is essential for research to further explore risk factors for imminent suicide and homicide in ED patients who present for psychiatric conditions and general injuries.