Caring Transitions – A Care Coordination Intervention to Reduce Suicide Risk Among Youth Discharged From Inpatient Psychiatric Hospitalization
Abstract
Abstract.Background: Suicide risk following youth psychiatric hospitalization is of significant concern. This study evaluated Linking Individuals Needing Care (LINC), a theory-driven, comprehensive care coordination approach for youth discharged from crisis services. Aims: To pilot LINC's potential effectiveness in increasing service utilization and decreasing suicide risk. Method: Participants were 460 youth patients who received LINC for approximately 90 days following discharge from crisis services. Service utilization, depressive symptoms, and suicide-related variables were measured at baseline and 30, 60, and 90 days after baseline. Results: Patients significantly increased the use of various beneficial, least restrictive services (individual therapy, medication management, and non-mental health supports) over the 90-day intervention. Significant decreases were observed in depressive symptoms, suicide ideation, and engagement in suicide-related behaviors. Limitations: Absence of a comparison group and nonparticipating families limit causal conclusions and generalizability. Conclusions: LINC may be a promising new approach following inpatient hospitalization that can engage and retain youth in services, likely resulting in improved treatment outcomes. This approach was designed emphasizing patient engagement, suicide risk assessment and management, safety planning, community networking, referral/linkage monitoring, coping and motivational strategies, and linguistic/culturally responsive practices to meet service and support needs of high-risk suicidal youth.
In the United States, rates of suicide and self-injurious thoughts and behaviors (SITBs) have been rising, especially among adolescents, whose mortality rate has increased by 76% in the last decade (Curtin & Heron, 2019; Plemmons et al., 2018). Consequently, US hospital encounters for SITBs have nearly doubled (Plemmons et al., 2018), while the utilization of mental health services remains low (Hom et al., 2015), particularly following discharge from psychiatric hospitalization (Fontanella et al., 2020). Given that lack of engagement/retention in mental health services is a significant predictor of subsequent youth suicide risk (Czyz & King, 2015; Mirkovic et al., 2020; Wolff et al., 2018), there is a need to improve access and follow-up care.
Care coordination is a public health approach involving timely and coordinated strategies to address service utilization barriers and health outcomes (McDonald et al., 2007). These interventions have been found to decrease hospital readmissions, increase service engagement, and improve health and service satisfaction (Gelkopf et al., 2016; Gorin et al., 2017; Grupp-Phelan et al., 2019). Unfortunately, most studies investigating care coordination strategies have primarily focused on adults (e.g., Motto & Bostrom, 2001; Wang et al., 2016), not on high-risk youth, and have not integrated principles/components across theoretical frameworks to teach providers (e.g., care coordinators) how to engage patients and address service utilization barriers.
To address these gaps, an innovative, theory-driven care coordination intervention (Linking Individuals Needing Care [LINC]) was developed. This intervention incorporates engagement and service use principles from multiple health care theories (Andersen & Davidson, 2007; Karver et al., 2006; Miller & Rollnick, 2002; Salzer et al., 1997), with lessons learned from prior research (e.g., Tricco et al., 2014) and consumer input (Gryglewicz et al., 2015). LINC, a 90-day intervention, infuses suicide risk management and care coordination strategies via caring contacts provided in any mode of communication. Considering naturalistic studies of youth discharged from inpatient care show that mental health functioning and risk of SITBs and psychiatric rehospitalization remains elevated after care (Czyz & King, 2015; Mirkovic et al., 2020; Wolff et al., 2018), this study addresses an important gap in the literature as it seeks to evaluate a theoretical and consumer-driven, clinical intervention designed to reduce subsequent suicide risk among a vulnerable population of youth.
Thus, the aim of the study was to evaluate the utility and potential effectiveness of implementing the LINC intervention in inpatient settings providing emergency care to suicidal youth. It was hypothesized that patients exposed to the intervention would report increased engagement (retention) in formal (mental health) and informal (non-mental health) services and experience decreased depressive symptoms and SITBs.
Method
Study Design
A longitudinal pilot study was conducted with at-risk youth during and following inpatient psychiatric hospitalization for suicide risk.
Sample and Setting
Participants were recruited from inpatient facilities at three behavioral health organizations in one southeastern state in the United States. Upon inpatient admission, intake coordinators screened patients using the Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001). Patients with a score of 10+ and/or who indicated suicide risk were further assessed by therapists using the Columbia Suicide Severity Rating Scale (C-SSRS; Posner et al., 2011) and a clinical assessment. Patients were eligible to participate in the study if they met the following criteria: (a) positive endorsement of SITBs, (b) self-report of multiple risk factors (e.g., history of nonsuicidal self-injury, victimization, and substance use), and/or (c) need for intensive support due to recurrent psychiatric hospitalizations and/or mental health or SITB history. A total of 1,125 patients between the ages of 10 and 17 (M = 14.6, SD = 1.8) met the study criteria, of whom 75.2% were female, 82.5% heterosexual, and 80.4% non-Hispanic (18.4% African American/Black and 15.9% multiracial). More than 86% of participants had a mood disorder, and 64% had a family history of mental illness. For 827 (73.5%) patients, contact with parents was made, which resulted in 460 youth/parents assenting/consenting for receipt of the LINC intervention and participation in the research study conducted between July 2016 and May 2020 (among nonparticipants, 280 parents perceived care was not needed and 87 were not interested). Study protocols were approved by a university institutional review board.
Intervention
Caring contacts were made during inpatient hospitalization, within 24–72 h of discharge, followed by weekly sessions for 30 days, and then monthly (up to approximately 90 days). At each contact, providers assessed and managed suicide risk (care/safety plans), identified service use needs/barriers, and utilized linguistic and culturally responsive strategies (e.g., individualized care monitoring plans, age and language-appropriate screening and assessment measures, collaborative decision-making, and inclusion of culture/preferences in care plans) to (a) build rapport, (b) increase symptom distress awareness and knowledge of help-seeking strategies and service navigation, (c) connect and coordinate referrals to community services, (d) motivate and encourage continued use of therapeutic supports, and (e) strengthen use of coping skills and other resources aimed to foster resiliency (an important component of evidence-based treatments for youth with depression/suicide risk; Barrett et al., 2017). Contacts were made via phone or in-person. Educational resources and a listing of community resources were provided to patients at discharge and check-in points. Providers delivering the intervention received an 8-h, face-to-face, skills-based training (using experiential techniques), three 6-h “booster” trainings, and bimonthly supervision (1-h phone sessions) from the lead author and two suicide prevention training experts during the study period. A random review of patient records (30%) was conducted to assess treatment fidelity by the lead author and a trained research assistant. Fidelity was calculated by taking the average percent of endorsed care coordination strategies observed across cases (i.e., 90.53%). Interrater reliability of the rating of fidelity between coders was good (κ = .782, p < .001).
Data Collection
Therapists and/or unit supervisors met with eligible youth/parents within 24–48 h of admission to explain the study and refer to LINC care coordinators. If this was not possible (e.g., discharge occurred over weekend), care coordinators contacted youth/parents by phone within 72 h of discharge. Youth assent/parental consent was obtained prior to study enrollment. Mental health, suicide risk, and service utilization measures were collected from participants at baseline (prior to intervention) and at 30, 60, and 90 days from baseline via phone and/or in-person interviews. Mental health was measured by the PHQ-9 (Kroenke et al., 2001), a 9-item scale used to assess depression severity (0- to 3-point scale; total scores 0–27). The measure has been shown to be reliable and valid with adolescents (Richardson et al., 2010; ∞ = .84, current sample) and was included in the study due to the strong correlation between depression and utilization of crisis services (Plemmons et al., 2018), subsequent suicide risk, and mortality (King et al., 1995; Schlagbaum et al., 2020). The C-SSRS Screen Version, a six question (yes/no) semistructured clinical interview (Posner et al., 2011), was used to assess suicidal ideation (SI; items 1–5 were added to create a SI score [ranged from 0 to 5], ∞ = .76, current sample) and suicide-related behavior (suicide attempts, aborted/interrupted attempts, plans; SRB [item 6]). The C-SSRS demonstrates strong convergent, divergent, and predictive validity for SRB during treatment for adolescents (Posner et al., 2011). Service utilization was assessed using the LINC Care Coordination Monitoring Form (Gryglewicz et al, 2018), a dichotomous (yes/no) measure used to monitor engagement in six different services: individual therapy, family therapy, medication management, non-mental health supports (e.g., after school activities), school services (e.g., counseling), and other services (e.g., faith-based). The number of readmissions to inpatient facilities was obtained via agency records. Patient records, including demographics and related measures, were deidentified to protect patient anonymity.
Data Analysis
Descriptive statistics, χ2 tests, and t-tests were used to examine service use patterns and outcomes at baseline and at 30, 60, and 90 days. Mixed-effects linear regression analyses were performed to examine changes in depressive symptoms and SI, and mixed-effects logistic regression analyses were performed to examine changes in SRB and service utilization (i.e., six different services). In the regression models, demographic and clinical characteristics and baseline outcomes were adjusted for. As measurement occurred every 30 days, “time” in the current study means 30 days. A random effect of BHOs was included in models to account for nested structure of the data. t tests and χ2 comparisons between participants who remained in the intervention and those who did not indicated no differences between groups on demographic and baseline mental health/suicide risk measures. Stata SE 15.1 was used for data analysis (StataCorp, 2017).
Results
Service Utilization Patterns
Most participants remained in the intervention and study: 90.4% (n = 416) at 30 days, 80.2% (n = 369) at 60 days, and 67.6% (n = 311) at 90 days. Their use of nonresidential services increased from 79% at baseline to 86% at 90 days. By the type of services, at 90 days, 71% were linked to individual therapy, 8% family therapy, 66% medication management, 7% non-mental health services, 13% school services, and 32% other informal supports. These increases were statistically significant at p < .05 level with the exception of family therapy and non-mental health services (Table 1). In general, youth were linked to one or two nonresidential services. Moreover, most youth (84.1%) were not readmitted for inpatient psychiatric care over the 90 days. Approximately 12% of youth were readmitted once, and 3.9% experienced two or more readmissions.
Table 2 presents mixed-effects logistic regression results predicting change over time in service use patterns while holding demographic and clinical characteristics, baseline depressive symptoms, and SI constant. Time was a significant predictor of all services, suggesting service use significantly increased over 90 days. For every 30 days of the intervention, the odds of receiving individual therapy increased by 54%, family therapy by 32%, medication management by 24%, non-mental health support by 54%, school services by 48%, and other services by 30%. Also, baseline depression and SI were associated with the use of some services. Youth with higher depressive symptoms at baseline were less likely to receive individual therapy, but more likely to receive medication management and school services. Youth with higher SI at baseline were more likely to receive individual therapy and medication management.
Mental Health and Suicide-Related Outcomes
Depressive symptoms and SI on average decreased by 65% and 86%, respectively. Youth engaging in SRB (last 30 days) also decreased by 84% (Table 3). Mixed-effects regression analyses indicated that such decreases were statistically significant, even adjusting for demographic and clinical characteristics, baseline depressive symptoms, and SI (Table 4). Specifically, for every 30 days, depressive symptoms decreased an average of 3.75 points, SI decreased an average of 0.84 points, and the odds of engaging in SRB decreased 54%. Depressive symptoms and SI at baseline were significantly associated with treatment outcomes. Youth with higher baseline depressive symptoms were likely to have elevated depressive symptoms and SI over time, but baseline depressive symptoms were not related to SRB during the intervention. Youth with higher baseline SI were more likely to have higher SI over time and engage in SRB. However, baseline SI was not associated with changes in depressive symptoms over time.
Discussion
Given that prior research has indicated the risk of SITBs and psychiatric rehospitalization remains elevated following discharge from inpatient hospitalization (Czyz & King, 2015; Mirkovic et al., 2020; Wolff et al., 2018), the aim of the present study was to examine whether participation in the LINC intervention was followed by improved engagement in services and mental health outcomes for at-risk youth. Overall, patients receiving the LINC intervention had sustained and large reductions in depressive symptoms and SITBs postdischarge. While higher baseline symptoms predicted later symptom levels, when controlled for, patients still had significant decreases across all symptom domains (compared to baseline), which suggests that improvements may not have been just regression to the mean. Of note, the most rapid changes in depressive symptoms and SITBs occurred in the first 30 days of the intervention. This finding has important clinical implications given suicide risk and mortality rates are elevated during the first few weeks and months following discharge from psychiatric care (Chung et al., 2019).
Additionally, youth had a low rate (16%) of inpatient readmission. This is notable given that compared to other studies (used as a benchmark) without post-discharge care, youth rehospitalization rates are significantly higher (28–43%; Adrian et al., 2020; Czyz & King, 2015; James et al., 2010). Furthermore, use of nonresidential services is quite low following discharge from inpatient or emergency department settings, with only 25–61% (Adrian et al., 2020; James et al., 2010; Sobolewski et al., 2013) receiving nonintensive services, far lower than nonresidential service use (78–90% at different time points) in the current sample. In fact, there was increased engagement in formal and informal services compared to baseline. The increased/sustained use of nonresidential services with the decreased use of restrictive services suggests that providers successfully built rapport and utilized strategies to motivate, encourage, and reinforce help-seeking behavior and engagement in supportive services, which are key components of the intervention.
Considering several elements of the intervention are particularly innovative compared to prior efforts, there may be multiple mechanisms of change that could explain the beneficial effects observed. Positive results could be due to the emphasis during provider training on rapport building, motivational/empowerment, and service navigation skills (Andersen & Davidson, 2007; Karver et al., 2006; Miller & Rollnick, 2002; Salzer et al., 1997). For example, numerous studies have found significant associations between treatment process variables and youth service utilization/participation and outcomes, with the establishment of the therapist–patient relationship being a salient factor (Karver et al., 2018). Given that providers developed relationships with patients prior to discharge (during a period of high emotional intensity) and maintained contact over time, they may have had the opportunity to develop genuine and trusting bonds while also establishing credibility. This may have facilitated other helping processes, such as increasing comfort with identifying triggers/stressors and modifying beliefs that services can be beneficial, which could have led to increased motivation to utilize appropriate coping strategies to keep themselves safe. Furthermore, being trained to identify and reassess suicide risk and communicate concerns of safety to patients (in an empathic/nonjudgmental manner) placed providers in a position to continuously enhance safety plans and reevaluate needs, service barriers, and the utility of coping resources, including service use referrals/linkages. Thus, youth in need of interventions may have been identified and responded to sooner than they would have been without such services, thereby decreasing the need for use of more restrictive levels of care.
Limitations and Future Research
The absence of a comparison group limits the ability to conclude the intervention caused observed changes. It is possible that other factors (e.g., supportive families and developed coping skills) contributed to the outcomes observed. Another limitation is more than 40% of parents did not consent to participate. It is unclear if parents possessed negative attitudes about the intervention and experienced conflicting demands on time and/or if their children exhibited poorer mental health, had increased suicide risk, or may have been more resistant to change compared to consenting parents. Therefore, it is unknown if changes in service utilization, depressive symptoms, and SITBs would generalize to all patients in these settings. Conversely, it is worth noting many of the youth with the highest levels of symptoms were the most likely to utilize a variety of services which is quite atypical relative to the prior literature (Hom et al., 2015). Nonetheless, future research should include randomized control trials to determine if results can be attributed to the LINC intervention. Future research could also examine adaptations to the mode of delivery (e.g., telehealth), include longer follow-up periods, and explore the potential differential impact of the intervention on diverse cultural groups and settings.
Conclusions
This study provides preliminary evidence to suggest that a care coordination intervention utilizing patient engagement, case management, suicide risk assessment and management, safety planning, and motivational and resiliency-building strategies can have a potential effect in engaging and retaining at-risk youth in services and reducing suicide risk during a high-risk period. The findings underscore the importance of providing immediate and intensive follow-up care in which patient symptoms, needs, and service barriers are monitored and assessed over time. Specifically, building a therapeutic alliance immediately following admission to inpatient care helps to establish creditability, rapport, and trust – qualities needed to engage, motivate, and empower youth and their families to work on a plan of care to keep youth safe. As health and behavioral health systems seek to improve suicide care, care coordination interventions utilizing warm hand-offs, active engagement, psychosocial education/advocacy, community asset mapping/networking, and assessment/monitoring of patient/family needs, service linkages, and mental health/suicide risk management may be an effective means to standardizing care.
Kim Gryglewicz, PhD, MSW, is an associate professor in the School of Social Work at the University of Central Florida, Orlando, FL, USA. Her research focuses on studying risk and resiliency factors for suicide, translation and dissemination prevention research, and evaluation of suicide prevention strategies.
Amanda Peterson is a clinical psychology doctoral graduate student at the University of South Florida, Tampa, FL, USA. Her primary research interests involve better understanding of suicide risk factors and improving suicide prevention programming for youth and adults.
Eunji Nam, PhD, is an assistant professor of social welfare at Incheon National University, Incheon, South Korea. Her areas of interest include young adults with mental health conditions, mental health service utilization, and program development and evaluation.
Michelle M. Vance, PhD, MSW, is an assistant professor in the Department of Social Work/Sociology at North Carolina Agricultural and Technical State University in Greensboro, NC, USA. She engages in health disparities research that centers on race and gender, and their impact on mental health outcomes for youth and adults.
Lisa Borntrager, MSW, is a research associate at the Center for Behavioral Health Research and Training at the University of Central Florida, Orlando, FL, USA. Her research interests include suicide prevention/intervention, program evaluation, health and mental health disparities, and evidenced-based practices/interventions.
Marc S. Karver, PhD, is an associate professor of clinical psychology at the University of South Florida, Tampa, FL, USA. He specializes in examining core processes (e.g., intervention engagement) in youth mental health services and improving prevention, assessment, and intervention services for suicidal youth and young adults.
The views expressed in this paper do not necessarily reflect the views, opinions, policies, or endorsement of SAMHSA or the US Department of Health and Human Services, nor does mention of trade names, commercial practices, or organizations imply endorsement by the US Government.
References
2020). Medically treated self-injury among children and adolescents: Repeated attempts and service use over 1 year. Psychiatric Services, 71(5), 447–455. 10.1176/appi.ps.201900152
(2007). Improving access to care in America: Individual and contextual indicators. In R. M. AndersenT. H. RiceG. F. Kominsji (Eds.), Changing the U.S. health care system: Key issues in health services policy and management (pp. 3–31). Jossey-Bass.
(2017). Promoting treatment engagement with specific populations: Depression. In W. O'DonohueL. JameC. Snipes (Eds.), Practical strategies and tools to promote treatment engagement (pp. 151–170). Springer.
(2019). Meta-analysis of suicide rates in the first week and the first month after psychiatric hospitalization. BMJ Open, 9(3), e023883. https://10.1136/bmjopen-2018-023883
(2019). Death rates due to suicide and homicide among persons aged 10–24: United States, 2000–2017. NCHS Data Brief (352), 1–8.
(2015). Longitudinal trajectories of suicidal ideation and subsequent suicide attempts among adolescent inpatients. Journal of Clinical Child and Adolescent Psychology, 44(1), 181–193. http://10.1080/15374416.2013.836454
(2020). Association of timely outpatient mental health services for youths after psychiatric hospitalization with risk of death by suicide. JAMA Network Open, 3(8), e2012887. http://10.1001/jamanetworkopen.2020.12887
(2016). A strengths-based case management service for people with serious mental illness in Israel: A randomized control trial. Psychiatric Research, 241, 182–189. 10.1016/j.psychres.2016.04.106
(2017). Cancer care coordination: A systematic review and meta-analysis of over 30 years of empirical studies. Annals of Behavioral Medicine, 51(4), 532–546. 10.1007/s12160-017-9876-2
(2019). Effect of a motivational interviewing-based intervention on initiation of mental health treatment and mental health after an emergency department visit among suicidal adolescents: A randomized clinical trial. JAMA Network Open, 2(12), e1917941. 10.1001/jamanetworkopen.2019.17941
(2015). Florida youth suicide prevention project. Substance Abuse and Mental Health Service Administration.
(2018). Care monitoring form [Unpublished LINC care monitoring form]. University of Central Florida and University of South Florida.
(2015). Evaluating factors and interventions that influence help-seeking and mental health service utilization among suicidal individuals: A review of the literature. Clinical Psychology Review, 40, 28–39. 10.1016/j.cpr.2015.05.006
(2010). Post-discharge services and psychiatric rehospitalization among children and youth. Administration and Policy in Mental Health and Mental Health Services Research, 37(5), 433–445. 10.1007/s10488-009-0263-6
(2006). Meta-analysis of therapeutic relationship variables in youth and family therapy: The evidence for different relationship variables in the child and adolescent treatment outcome literature. Clinical Psychology Review, 26(1), 50–65. 10.1016/j.cpr.2005.09.001
(2018). Meta-analysis of the prospective relation between alliance and outcome in child and adolescent psychotherapy. Psychotherapy, 55(4), 341–355. 10.1037/pst0000176
(1995). A prospective study of adolescent suicidal behavior following hospitalization. Suicide and Life-Threatening Behavior, 25(3), 327–338. 10.1111/j.1943-278X.1995.tb00955.x
(2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. 10.1046/j.1525-1497.2001.016009606.x
(2007). Closing the quality gap: A critical analysis of quality improvement strategies – Care coordination. Agency for Healthcare Research and Quality.
(2002). Motivational interviewing: Preparing people to change (2nd ed.). Guilford Press.
(2020). Repeating a suicide attempt during adolescence: Risk and protective factors 12 months after hospitalization. European Child and Adolescent Psychiatry, 29, 1729–1740. 10.1007/s00787-020-01491-x
(2001). A randomized controlled trial of postcrisis suicide prevention. Psychiatric Services, 52(6), 828–833. 10.1176/appi.ps.52.6.828
(2018). Hospitalization for suicide ideation or attempt: 2008–2015. Pediatrics, 141(6), 1–10. 10.1542/peds.2017-2426
(2011). The Columbia-Suicide Severity Rating Scale: Initial validity and internal consistency findings from three multisite studies with adolescents and adults. The American Journal of Psychiatry, 168(12), 1266–1277. 10.1176/appi.ajp.2011.10111704
(2010). Evaluation of the Patient Health Questionnaire-9 Item for detecting major depression among adolescents. Pediatrics, 126(6), 1117–1123. 10.1542/peds.2010-0852
(1997). Validating quality indicators: Quality as relationship between structure, process, and outcome. Evaluation Review, 21(3), 292–309. 10.1177/0193841X9702100302
(2020). Depressed mood prior to death: Implications for precipitating factors of youth suicide. Crisis, 41(6), 445–452. 10.1027/0227-5910/a000660
(2013). Mental health follow-up among adolescents with suicidal behaviors after emergency department discharge. Archives of Suicide Research, 17(4), 323–334. 10.1080/13811118.2013.801807
(2017). Stata statistical software: Release 15. StataCorp LLC.
. (2014). Effectiveness of quality improvement strategies for coordination of care to reduce use of health care services: A systematic review and meta-analysis. Canadian Medical Association Journal, 186(15), E568–E578. 10.1503/cmaj.140289
(2016). Coping card usage can further reduce suicide reattempt in suicide attempter case management within 3-month intervention. Suicide Life Threat Behav, 46(1), 106–120. 10.1111/sltb.12177
(2018). Trajectories of suicidal ideation among adolescents following psychiatric hospitalization. Journal of Abnormal Psychology, 46(2), 355–363. 10.1007/s10802-017-0293-6
(