Measurement and Comparison of Suicide Prevention Efforts in Emergency Departments and Crisis Centers Using the Zero Suicide Framework (Board#20)

Abstract : Crisis response centers (CRCs) were designed to meet the needs of individuals exhibiting suicidal thoughts, behaviors, or other emotional distress by providing immediate care while helping to find subsequent mental health services. However, many suicidal youth will be first encountered in an emergency department (ED). Following an increase in suicidal ideation and history of attempts (Hill et al., 2021) as well as suicide-related ED admissions (Yard et al., 2021) during the COVID-19 pandemic, EDs must be prepared to encounter suicidal youth. The need for a multidisciplinary approach to suicide-related care in EDs has been identified and widely communicated (Petrik et al., 2015), and frameworks have been proposed and implemented in some hospitals (Betz & Boudreaux, 2016), yet tools to assess an ED’s readiness to encounter and respond to suicidal individuals remain void in the literature.

The Zero Suicide Institute (ZSI) has developed a framework for the implementation of suicide-related care in healthcare systems (ZSI, 2022a) and an assessment (Zero Suicide Organizational Self-Study; ZSI, 2022b) to determine adherence to it. By following best practices across the core domains of the framework, organizations can be prepared to meet the needs of suicidal individuals and ensure healthy transitions to other care-providing systems. Ideally, the components of the Zero Suicide Framework work collaboratively to achieve the most effective outcome for youth (Stapelberg et al., 2020).

The current study adapted the Zero Suicide (ZS) Organizational Self-Study to be relevant to the expected roles of Pennsylvania (PA) CRCs and EDs. We will report on (1) the preparedness of CRCs and EDs, (2) whether EDs are more, less, or as prepared to provide suicide-related care as CRCs, and (3) whether differences in readiness exist in urban vs. rural counties in PA.

Method: 12 CRCs and 18 EDs were surveyed across 10 PA counties. The survey contained items relating to one of five components of the ZS Framework (Lead, Train, Identify, Engage, Improve) and relevant to the roles of both systems when encountering a potentially suicidal individual. Data collection occurred from February 2020 until December 2022.

Results: EDs (M = 3.25, SD = 1.08) and CRCs (M = 3.26, SD = 1.05) scored above a moderate score of ‘3’ on the ZS implementation scale for Lead. Both systems reported high average scores in their overall policies and procedures (ED; M = 4.06, SD = .99, CRC; M = 4.00, SD = .85) and suicide prevention teaming (ED; M = 3.41, SD = 1.66, CRC; M = 3.09, SD = 1.45), though scores for including loss/attempt survivors were lower (ED; M = 2.13, SD = 1.63, CRC; M = 2.30, SD = 1.64). Train scores were higher (ED; M = 3.81, SD = 1.56, CRC; M = 3.82, SD = 1.25) with best-practice advanced suicide prevention training being required for clinical staff in 61% of EDs and 58% of CRCs. Scores were similarly high for Identify (ED; M = 4.22, SD = y, CRC; M = 4, SD = 1.12), with 72% of EDs and 67% of CRCs using a validated suicide screening tool and 72% of EDs and 67% of CRCs performing risk assessments using established measures/protocol when an individual screens positive. Both systems scored moderately well for Engage (ED; M = 3.23, SD = .80, CRC; M = 3.16, SD = 1.13), with 67% of EDs and 75% of CRCs including discussions of lethal means reduction for all at-risk youth. With respect to Improve (ED; M = 3.50, SD = 1.71, CRC; M = 3.18, SD = 1.47), 50% of organizations in each system reported tracking, aggregating, and examining suicide-related data for quality improvement purposes. Upon comparing scores across systems, a chi-square test did not yield any significant results. Finally, a Mann-Whitney U test by population density did not reveal any differences in scores between CRCs located in rural vs urban counties. However, urban EDs did score significantly higher (p < .05) than rural EDs in suicide prevention teaming, advanced training, suicide risk assessment, and lethal means reduction.

Discussion: The results of our analyses suggest that while EDs and CRCs have demonstrated adherence to training recommendations and screening/assessment responsibilities, there is room for improvement in suicide-related care after identifying risk and tracking outcomes for quality improvement. In accordance with Joint Commission (2015) requirements, all reported using some kind of suicide screening, with 72% of EDs using an evidence-based tool. Scores in the Engage domain, specifically regarding lethal means reduction strategies, were variable for both systems. Urban EDs reported utilizing significantly better lethal means reduction strategies than rural units. Given that similar differences were seen for suicide prevention teaming, advanced training, and assessment, resource availability may impede some rural EDs from providing the level of care seen in more funded, urban EDs. Since no significant differences were found for any ZS items in CRCs, it appears that a similar level of preparedness can be expected whether presenting to an urban or rurally located center. This may reflect that funding and regulation standards both are directed from the state, thus limiting differences based on location.

A few limitations are present with the current study that, if addressed in future research, could provide stronger insight into these findings. First, due to a small sample size, groups of organizations were too small to adequately compare systems by population density. With a larger number of participating organizations, future research could examine whether a similarly observed difference between urban and rural EDs translates to EDs being more prepared than CRCs in urban environments, or vice versa in rural environments. Second, the questions included in our adapted form of the ZS Organizational Self-Study did not use a traditional Likert-type response structure. Answer choices were string responses that, to the research team and ED/CRC experts, ranged from a baseline/non-adherent (“1”) level to the best-practice (“5”) strategy in each ZS component. Nonetheless, the survey applies a well-supported framework to a system of care that lacks an assessment of suicide readiness. Future research should seek to improve the adapted survey, as well as utilize the framework in additional settings.