Using the electronic health record to identify suicide risk factors in an Alaska Native Health System.

Abstract

Suicide rates are higher in some Alaska Native and American Indian communities than the general U.S. population. Screening for suicide risk typically requires self-report, but many people may not engage with conventional screening because of distrust or are reluctant to disclose thoughts of suicide. Resource-sensitive methods of detecting suicide risk are needed. This study identifies routinely collected electronic health record data to identify demographic, clinical, and utilization factors associated with suicide-related visits in a tribal health care system. In this retrospective, case-control study, cases were defined as any person with a suicide-related visit from 2012 to 2015. Cases and controls were matched by age, sex, and urban/rural residence. We used conditional logistic regression to estimate odds ratios, which were interpreted as prevalence ratios (PR) based on the rare outcome assumption. The dataset included 314 cases and 1,169 controls. In the year before the index visit, cases had higher prevalence of poisoning or overdose (PR = 13.4, 95% confidence interval [CI] [3.5, 51.7]), emergency department and urgent care visits (PR = 15.8, 95% CI [6.6, 38.1]), and hospitalizations (PR = 4.5, 95% CI [3.0, 6.8]). Electronic health records can be used to identify factors that are significantly associated with suicide risk among those who may not be flagged by screening. Risk detection through electronic health record assessment might increase clinical workload in the short term, but this change would be offset by downstream prevention of suicide-related events. Such efforts could improve suicide risk detection and help to improve suicide-related health disparities in Alaska Native and American Indian populations.