Teen Options for Change: An Intervention for Young Emergency Patients Who Screen Positive for Suicide Risk

Participants were 49 adolescents who met study inclusion criteria, which included being 14 to 19 years of age; having a positive suicide risk screen, defined as suicidal ideation, a recent suicide attempt, or positive screens for both depression and substance abuse (measures described below); and presenting with a nonpsychiatric chief complaint. Exclusion criteria were a level 1 trauma (critically ill or medically unstable), significant cognitive impairment (unable to complete the self-report screen), or disposition of psychiatric hospitalization. [Descriptive information for the full sample of adolescents screened and a CONSORT diagram are included in the online appendix.]

Suicidal ideation and history of recent suicide attempt were assessed with two questions adapted from the Columbia–Suicide Severity Rating Scale (6): recent (past two weeks) suicidal ideation or a recent (past month) suicide attempt. The 15-item Suicidal Ideation Questionnaire–Junior (SIQ-JR) (7) was used to measure the nature and frequency of suicidal thoughts over the past month. [Sample items and psychometric properties for the SIQ-JR and all study measures are included in the online appendix.]

Depression was assessed with the Reynolds Adolescent Depression Scale–2, Short Form (RADS-2:SF), a ten-item self-report scale that measures severity and frequency of depression symptoms (8). Hopelessness was assessed with the Beck Hopelessness Scale (BHS) (9). Alcohol consumption and at-risk drinking were assessed with the Alcohol Use Disorders Identification Test (AUDIT) (10). The first three items (AUDIT-C) were used for the suicide risk screen, and all ten items were used in baseline and follow-up measures. Substance use and its consequences were assessed with the six-item CRAFFT (11), which has demonstrated strong sensitivity and specificity for identifying substance-related problems among adolescent general medical patients (11). A positive screen was defined as a score of 26 or higher on the RADS-2:SF with comorbid substance misuse (CRAFFT score ≥2) or alcohol misuse (AUDIT-C score ≥3).

All recruitment took place during afternoon and evening emergency department shifts between November 2009 and October 2010. Adolescents who screened positive were required to provide at least two telephone numbers, verified by project staff, before randomization to the TOC intervention or to enhanced treatment as usual. To ensure the integrity of randomization, we used envelopes with randomly ordered group assignments, organized into four packets and stratified by gender and suicide attempt history.

Adolescents randomly assigned to enhanced treatment as usual were given a crisis card with phone numbers for suicidal emergency support, in addition to written information about depression, suicide risk, firearm safety, and local mental health services. Adolescents randomly assigned to the TOC intervention received these same resources plus personalized feedback about their screening responses. They also participated in an adapted motivational interview (approximately 35–45 minutes) with a mental health professional. Study therapists completed a minimum of 40 hours of training conducted by a member of the Motivational Interviewing Trainers’ Network. They were certified when their taped pilot sessions met preset criteria for adherence with motivational interviewing elements (for example, use of open-ended questions and affirmations). [Data on the adherence of TOC therapists with motivational interviewing principles are provided in the online appendix.]

A TOC therapist reviewed the personalized feedback with each adolescent assigned to the TOC condition. This feedback included normative data for depression and suicidal ideation, risk level for substance use, and functional impairment. TOC therapists used a guide that was culturally tailored, with input from adolescent focus groups, to clarify goals and values of participants and to facilitate identification of behavioral goals. Using motivational interviewing techniques, such as building a discrepancy between the adolescent’s values, goals, and actual fulfillment of these goals, therapists facilitated the adolescent’s development of a personalized action plan. This consisted of an objective (for example, to finish high school or obtain a GED) and a list of up to three steps to take toward this objective. Adolescents in the TOC group also received a handwritten follow-up note and a telephone check-in from their therapist two to five days after their emergency department visit to support and facilitate implementation of their action plan.

Adolescents and their parents or guardians were offered dollar gift items as a token of appreciation for completing the screen. Adolescents were remunerated $20 for completion of the baseline assessment. At the two-month follow-up assessment, participants were remunerated $30, with an additional $20 incentive if they returned to the hospital for the assessment. Community-based assessments were also conducted. The clinicians conducting follow-up assessments were blind to intervention condition.

We obtained institutional review board (IRB) approvals from our university and the participating hospital. Written informed consent was obtained from parents or guardians when present (the IRB waived the parental consent requirement if a parent or guardian was not present), and all adolescents gave their assent to participate. The study team followed a detailed risk-management protocol with action steps for adolescents in either group who met criteria for “high risk” (such as active suicidal intent or a past-week attempt) at baseline and follow-up.

We used chi square and independent t tests to compare the baseline demographic and clinical characteristics of adolescents randomly assigned to TOC or enhanced treatment as usual. Our analysis of intervention effects followed the intent-to-treat strategy, which included all 46 adolescents who completed the follow-up assessment. Repeated-measures analyses of variance were used for the continuous outcome measures.

The ages of adolescent participants ranged from 14 to 19 years, and the mean±SD age was 17.7±1.7. Eighty percent (N=39) of the sample was female. The racial-ethnic distribution was 57% (N=28) African American, 39% (N=19) Caucasian, 4% (N=2) American Indian or Alaska Native, 2% (N=1) Native Hawaiian/Pacific Islander, 2% (N=1) Hispanic, and 2% (N=1) other. These percentages do not sum to 100% because some adolescents self-identified as biracial.

Of the 49 adolescents who screened positive for elevated suicide risk, 35% (N=17) had recent suicidal ideation or a recent suicide attempt, and 53% (N=26) reported current depressive symptoms with comorbid substance abuse. A total of six participants (12%) screened positive for both criteria. Seven of the participants (14%) had attempted suicide within the past month.

Of the 27 adolescents who were randomly assigned to TOC, 23 participated in the full intervention (personalized feedback and adapted motivational interview); four participants did not participate in the interview because they were discharged and chose to leave the hospital before the interview could be completed. Adolescents’ behavior change goals primarily concerned personal growth and improvement (70%, N=16), academic achievement or school attendance (15%, N=3), and increased independence (15%, N=3). Goals related to personal growth and improvement ranged from improving the use of supports, such as friends and family (20%, N=5), to engaging more in favored or new hobbies or activities (17%, N=4) or getting professional help (25%, N=6). Specific examples include “getting up at set time [8 a.m.] in the morning” and “[increase] physical activity.”

Forty-six of 49 participants (94%) completed the two-month follow-up measures. Retention analyses were not conducted because only three adolescents were lost to follow-up. These three adolescents (two African Americans and one Caucasian) had a mean age of 17.92±1.66. Mental health service utilization after the emergency department visit did not differ between groups. Seven adolescents in the TOC group (29%) and eight adolescents in the treatment-as-usual group (36%) received some type of mental health service during the two months after their visit.

The mean baseline and two-month follow-up scores for TOC and control groups are shown in Table 1. For depression (RADS-2:SF), we found a significant main effect for treatment group (F=10.84, df=1 and 44, p<.01) and a significant time × treatment group interaction (F=9.89, df=1 and 44, p<.01), indicating a positive effect for TOC. We did not find any significant treatment or group effects for hopelessness (BHS), suicidal ideation (SIQ-JR), or alcohol misuse (AUDIT). There was a significant effect for time for suicidal ideation (F=7.41, df=1 and 44, p<.01). Adolescents showed a decrease in suicidal ideation over the study period.

Because this preliminary RCT was not powered to identify modest effects as statistically significant, we also examined actual effect sizes for each outcome variable. The TOC intervention had large positive effects for depression, moderate positive effects for hopelessness, and small positive effects for suicidal ideation and alcohol misuse (Table 1).

In this pilot RCT, adolescents in the TOC intervention reported significantly greater reductions in depressive symptoms than adolescents in the enhanced treatment-as-usual group. This finding indicates that a brief emergency department–based intervention that incorporates personalized feedback and an adaptive motivational interview may be beneficial to adolescents who screen positive for suicide risk. Although further research is needed, particularly because of the limited sample size in this study, such a reduction in depression, in addition to the medium effect size for reducing hopelessness, suggests that a focus on positive behavioral activation and change has a beneficial impact on adolescents at risk, even within a short period.

TOC is a brief intervention that takes advantage of the emergency setting, in which a teen’s parent or guardian is often present and wait times provide an opportunity for assessment and intervention (12). Feasibility of TOC implementation in this setting was demonstrated in this study by the high levels of patient participation and retention and by the IRB waiver of required parental consent for adolescents’ participation in the emergency setting. Moreover, adolescents who use emergency services as their usual source of health care often have fewer financial resources and report higher rates of substance use, general health problems, and mental health problems (13). Therefore, compared with other settings, the emergency department setting may allow for a greater number of at-risk adolescents to be reached.

It is important to note that TOC had a nonsignificant impact on suicidal ideation. Adolescents in both groups—TOC and treatment as usual—showed significant reductions in suicidal ideation across the two months. Furthermore, TOC was not associated with increased treatment seeking. Evidence regarding effective interventions aimed at reducing suicidal thoughts and behaviors among youths is very limited. In a meta-analysis by Corcoran and colleagues (14), evidence pointed toward slight decreases in suicidal and self-harm events at posttest for intervention participants compared with a control group for various interventions; however, this positive trend was not evident at later follow-ups. This evidence suggests that the minimal impact on suicidal ideation found in our study may not be maintained over time. For adolescents at elevated suicide risk who report high levels of suicidal ideation, it may be important to expand TOC to include therapeutic approaches that specifically target the suicidal ideation.

The promising findings for TOC in this preliminary study warrant further study with a larger sample and longer follow-up period to determine the extent to which findings replicate, to examine the mediators of any positive effects (for example, in regard to behavioral change, self-efficacy, hopefulness, and more reasons for living), and to determine whether positive changes translate into reduced suicidal behavior over time.

There were several limitations to this study. The sample was recruited from one hospital’s emergency department in a relatively underserved, low-income community, and it is unclear whether these findings would apply to a larger, more representative population of adolescents. Similarly, all recruitment occurred during afternoon and evening emergency shifts, and adolescents who seek emergency services at other times may differ in some way that may have an impact on intervention effectiveness. Other limitations include the preliminary nature of this study, its relatively small sample size of 49 adolescents, and the absence of patient satisfaction data. Finally, although the consent rate was very high for follow-up assessments (N=46; 94%), two months is a relatively short follow-up period, and the sustained or delayed effects of TOC are unknown. We also did not have information to disaggregate the effective components of TOC, which included personalized feedback, an adapted motivational interview, and a follow-up note.

Our findings indicate that TOC may be a promising brief intervention for adolescents who screen positive for suicide risk in medical emergency departments but are not at high acute risk and in need of psychiatric hospitalization. As suicide risk screening becomes increasingly prevalent, triage options warrant further research and consideration, and TOC may be a useful triage option. This is particularly important given recent findings of an absence of improved outcomes even when adolescents follow through with outpatient recommendations after being discharged from the emergency department (15).

The authors thank the National Institutes of Health for funding this study (grant number K24 MH077705). The authors acknowledge Alan Hackett, M.S.W.; Tasha M. Kelley-Stiles, M.S.W.; Bianca Burch, M.S.W.; Henry Wynne; and Ryan Hill for assistance with study implementation. The authors also thank the emergency department staff; Rebecca Cunningham, M.D., for her administrative support; Ken Resnicow, Ph.D., for motivational interview training and consultation; and participating adolescents, parents, and other guardians.

The authors report no financial relationships with commercial interests.