Current psychosocial interventions

Psychosocial interventions for suicidal behavior mostly target symptoms of psychiatric illness, often using a cognitive-behavioral therapy (CBT) framework. However, results from two meta-analyses indicate a lack of efficacy in suicide interventions for adolescents [15, 16]. One meta-analysis of CBT to reduce suicidal behaviors identified 6 adolescent and 18 adult studies and found that a significant treatment effect was observed for adult samples but not for adolescents [16]. A more recent meta-analysis which included 9 adolescent studies similarly found that psychotherapy reduced suicide attempts in high-risk adults but not adolescents [15]. One recently published study not represented in these meta-analyses is a multisite Dialectical Behavior Therapy (DBT) study of 173 adolescents [17], which found advantages of DBT over the comparison group at the end of the 6-month treatment period, but not at the 12-month follow-up, in reducing suicide attempts. The DBT protocol, although promising, was highly intensive. The majority of extant studies focus on crisis intervention, psychiatric symptomatology, regulating negative affect, and reducing cognitive distortions. Adjunctive and alternative approaches to reducing suicidal behaviors, particularly in adolescents, are warranted.

Low positive affect as a potential mechanism of risk for suicidal behavior

Numerous studies suggest that low positive affect is a risk factor for suicidal behavior independent of negative affect and may be another mechanism that leads to suicidal behavior. Low positive affect has been found to prospectively predict time to suicidal events (i.e., either a suicide attempt or an emergency intervention for an acute suicidal crisis) in adolescents over six months of follow-up, even after controlling for a number of other predictors, including depression severity and anhedonia [10]. In a cross-sectional study of over 1000 adolescents, the positive affect experience of gratitude, was inversely associated with suicidal ideation and suicide attempts [18]. The effect of positive affect on suicidal ideation has also been demonstrated in older patients. In a study of 462 primary care patients, age 65 and over, positive affect distinguished people with suicide ideation from those without, even after statistically controlling for the influence of age, gender, depression, negative affect, illness burden, activity, sociability, cognitive functioning and physical functioning [19].

The Broaden and Build theory [20], offers a clear conceptual model for the association between positive affect and suicidal behavior. This model asserts that one function of positive affect may be to broaden attentional scope to help individuals be more open to novel stimuli and social supports, which in turn, broadens and builds psychological and social resources necessary for survival [20–25]. Consistent with this theory, positive affect has been shown to broaden attentional scope [21], motivate thought-action tendencies (i.e., increase behavioral activation) [20], and counter the deleterious effects of negative emotions [26].

Most individuals, particularly those with depressed mood, process emotions through a negativity bias [27], in which individuals pay more attention to and react more strongly to negative stimuli than positive stimuli. Thus, it is likely that positive affect and experiences are frequently being discounted. Understanding the functions of positive affect and negative affect, and intentionally bringing greater attention and awareness to positive affect, may be an important counterweight to this negativity bias, and in turn, may be a mechanism reducing suicide risk.

Positive psychology interventions

Aside from our pilot studies [28, 29], few Positive Psychology Interventions (PPIs) have specifically been applied to patients with suicidal behavior. Two meta-analyses of PPIs for depression and well-being found that these interventions have small to moderate effects in enhancing well-being and decreasing depressive symptoms and that effects are stronger for individual vs. group formats [30, 31]. However, the vast majority of these interventions were with nonclinical populations, with nearly half of them conducted with college students. Only a handful of studies have been recruited from clinical or hospital settings [32–35]. The few studies conducted with suicidal populations (not included in the meta-analyses) have utilized gratitude exercises and have yielded mixed findings with one study showing favorable effects on suicidal ideation, compared to a food diary, in inpatient adults [36]. Another study, however, found that those randomized to a cognition-focused intervention had significantly greater improvement in depression, suicidal ideation, optimism, and gratitude, compared to the positive psychology intervention [37].

Skills to Enhance Positivity (STEP)

We developed the Skills to Enhance Positivity (STEP) intervention for adolescents admitted to an inpatient psychiatric unit due to suicide risk. STEP is a multi-modal, adjunctive intervention involving four individual in-person sessions delivered on the inpatient unit, followed by text messaging post-discharge to assess mood and “push” positive affect strategies for up to three months. While STEP borrows heavily from preceding positive psychology interventions such as Seligman’s positive psychotherapy (PPT) [34], it is also distinct. STEP does not address happiness, optimism, or meaning, which may feel unattainable or invalidating to patients in acute distress. To mitigate perceptions of “toxic positivity” [38], STEP is grounded in: psychoeducation on the functions of both negative and positive emotions, the negativity bias that leads to an overemphasis on perceived threats [27], the broaden and build theory of positive emotions [20], and the concept of a positivity ratio [23]. The premise of STEP is that positive emotions and experiences may be easily discounted by the cognitive constriction that often accompanies suicidal ideation and behavior, yet they serve important protective functions. Furthermore, rather than replacing negative with positive, positive emotions can have an additive effect to counterbalance negative emotions. Consequently, specific exercises designed to increase attentional awareness (i.e., psychoeducation on functions of emotions, and specific exercises to cultivate mindfulness, gratitude, and savoring) are practiced, and, in this sense, STEP more closely aligns with acceptance-based approaches to treating depression, which have become increasingly common [39].

STEP was piloted in an open development trial (N = 20) [28] and a pilot RCT (N = 52) [29]; both trials indicated excellent feasibility and acceptability. Preliminary clinical findings of the pilot RCT were promising. Over six months of follow-up, five of 26 participants in the STEP condition (19%) vs. 10 of 26 in the Enhanced Treatment as Usual (ETAU; healthy habits texts) condition (38%) had a suicidal event. Six suicidal events were reported among STEP participants and 13 events were reported among ETAU participants. Thus, those randomized to STEP had 50% fewer suicidal events, in total, and 50% fewer participants engaging in suicidal events, compared to ETAU, the latter corresponding to a medium to large effect size h of .43. Those in the STEP condition also saw a more dramatic decrease in active suicidal ideation from baseline compared to their worst week during the follow-up period (STEP 49% vs. ETAU 19% decrease) (Cohen’s d = -0.74). Furthermore, compared to ETAU, parents of STEP participants reported significantly greater improvement in their child’s depression at post-treatment (d = 0.76; 95% CI 0.21–1.30) [29].

Given the strength of our pilot findings, we are conducting a larger, Hybrid Type 1 effectiveness-implementation multi-site study [40] to examine the effectiveness of STEP as an adjunctive intervention to reduce suicide events, active suicidal ideation, and depression, in youth, 12 to 18 years old, admitted to a psychiatric inpatient unit due to suicide risk, over twelve months of post-discharge follow-up. One significant difference between our pilot studies and the current trial is the selection of interventionists. While the initial pilot studies utilized research personnel to deliver STEP, as is appropriate for intervention development studies, the current protocol is delivered by non-doctoral, clinical staff on the unit. The decision was made to use unit staff as the interventionists because decreasing the length of time from pipeline to clinical implementation is paramount. We also created video trainings for STEP interventionists, and brief videos of STEP content which can be used to help the interventionists deliver STEP and also serve as exercise reminders which can be “pushed” remotely after discharge. We will examine the implementation context of STEP in this Hybrid Type 1 trial and to use these data to inform selection of implementation strategies to scale-up, spread, and sustain STEP in additional inpatient units if it proves efficacious.

Study design

This study uses a two-group RCT enrolling 216 adolescents to test the primary hypothesis that STEP, delivered to adolescents admitted to an inpatient psychiatric unit, will result in better clinical outcomes than an ETAU condition at the end of the Active Phase of treatment (3-months post-baseline) and at 6- and 12-month follow-up intervals. STEP is an adjunctive, multi-modal intervention that consists of an in-person phase of four sessions, and a remote delivery phase of daily mood ratings and positive affect exercise practice messages delivered via text messaging, which also contains links to videos and mental health resources, for the first month following discharge, followed by tapered messages in the following two months. The ETAU comparison condition is enhanced by healthy habits text messages, as well as reminders to check the participant’s safety plan, matched in frequency to the STEP text messaging protocol. Healthy habits text messages do not overlap with STEP content but are ecologically valid for mental health interventions. Participants in both conditions receive standard care while hospitalized as well as referral to standard mental health care after discharge. (For a timeline of interventions and assessments in each condition, see Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Timeline of interventions and assessments.

* Anchored to discharge date.

https://doi.org/10.1371/journal.pone.0287285.g001

Recruitment for this multi-site trial occurs at three adolescent inpatient units from different hospital systems in the Northeast U.S. (first participant recruited on 19 July 2022). The Principal Investigators (PIs) and Site PIs comprise the Steering Committee, and data management is centralized and overseen by a project data coordinator. Data from all sites are stored in REDCap, and when applicable, are double entered to check for accuracy. Oversight for this study, including any modifications to the protocol and safety of study participants, is conducted by a Single IRB (with all other sites ceding) and an NIMH Data and Safety Monitoring Board (DSMB). For continued approval, annual reports are submitted to the IRB and three times a year to the NIMH DSMB.

Participants

Interventions/Group assignment

Participants in both conditions receive standard inpatient care on each unit. Standard care on the three psychiatric units serving as sites for this study share several commonalities such as: 1) focused psychiatric assessment; 2) a multi-dimensional approach that includes combining individual sessions, group therapies, and medications; 3) inpatient community meetings, occupational therapy support and activities, tutoring; 4) family involvement through meetings and visits; and 5) a secure environment. All units share the goal of stabilizing the child’s behavioral health crisis and, upon discharge, offer recommendations for safety, including a safety plan. All families receive psychoeducation regarding suicidal behavior during their hospitalization, which includes: removing potential suicide means (e.g., guns); being alert to mood changes and suicidal statements; and learning how to access crisis and emergency services. All patients discharged from the inpatient unit are also referred to follow-up psychiatric care (typically an outpatient program, or short-term day treatment program, as well as an appointment with a psychiatrist for medication monitoring), and leave with a copy of their safety plan.

In-person phase.

STEP begins with four individual treatment sessions that include psychoeducation on functions of emotions, mindfulness meditation, savoring, and gratitude. Specific session content is described in Table 1. These practices were specifically selected as they are purported to increase eudemonic positive affect and are consistent with our theoretical framework of accepting the presence and function of negative emotions concomitant with a focus on attending to and cultivating positive emotions as a counterbalance. The psychoeducation session focuses on building rapport and explaining the scientific rationale for our program (e.g., negativity bias, function of positive emotions, counterbalancing negative and positive emotions), to provide context for these practices and obtain participant buy-in. The following sessions which focus on mindfulness mediation, savoring and gratitude respectively, present multiple ways of practicing, asking participants to choose the practice that is most feasible and comfortable for them. Sessions are designed to maximize flexibility such that if not all of the content is delivered in one session, it can be delivered in the following session. When sessions need to be condensed, the brief STEP videos can be utilized to reduce time and burden on the interventionists. Each session lasts approximately 25–30 minutes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. STEP in-person sessions.

https://doi.org/10.1371/journal.pone.0287285.t001

Therapist training

Therapists for this study are clinical staff based on the unit (e.g., milieu therapists, social work interns) who receive training to administer STEP. Training materials include a written treatment manual and five audio- and video-recordings describing the premise of STEP and illustrating each of the sessions. To enhance feasibility, training is self-paced and available on-demand. After reviewing the training materials, interventionists are asked to record role plays that are reviewed by trained interventionists, and then feedback is provided. Periodic booster trainings are conducted as needed.

Treatment integrity

As STEP is being deployed in a clinical setting by on-duty clinicians, recording actual sessions is not feasible. Thus, we developed self-administered adherence checklists that cover the most important elements of the intervention for each of the four sessions. Interventionists are asked to also rate their audio-recorded role plays to further reinforce training. These ratings will be compared against the ratings performed by a STEP trainer, and the interventionist will receive feedback on any differences in ratings; adherence over 80% will be considered acceptable. After the first three study participants, interventionists will be asked to do another rated role play to ensure that they are administering STEP as intended. Treatment integrity assessments also provide implementation outcome data for the real-world delivery of STEP in a future trial, to determine the extent to which frontline clinicians are able to deliver STEP with fidelity [42], and without a voltage drop or program drift, two of the main concerns with future sustainability of any evidence-based intervention in usual care settings [43].

Assessments

The assessment battery is administered at baseline (prior to randomization) by the research personnel. At the end of the active intervention phase (3 months post-baseline), and at the 6- and 12-month follow-up intervals, assessments are conducted by a masked outcome assessor who does not know participants’ condition assignment. (See Fig 2 for a schedule of assessments at each timepoint).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Schedule of enrolment, interventions, and assessments.

**t1 = 4-week intervention period; t2 = 1-month remote delivery phase via text; t3 = 2-month remote delivery phase (tapered); t4 = 3-month follow-up; t5 = 6-month follow-up; t6 = 12-month follow-up.

https://doi.org/10.1371/journal.pone.0287285.g002

Measures

Data analyses

Data from all randomized participants will be included in analyses (intent-to-treat). Preliminary analyses will include descriptive statistics to examine the distributional and psychometric properties of the variables (e.g., normality, internal consistency). Variables will be transformed to better approximate normality if this is necessary given the observed distribution and the assumptions of the analysis model. We will also examine post-inclusion attrition by comparing study completers to dropouts on sociodemographic, baseline data, and psychotropic medication usage to determine if they differ systematically. To accommodate attrition and other sources of missingness, we will utilize the method of multiple imputation by chained equations. We will examine the success of the stratified randomization by examining between-group differences at baseline. Preliminary analyses may also include analyses of adverse events, the progress of recruitment and retention, and data processing and quality markers as the study progresses. Daily mood monitoring data is not included in outcome analyses because it is conceptualized as part of this adjunctive intervention and only collected in one condition.

To examine the effectiveness of STEP in reducing suicidal events, active SI, and depression at 3- and 6-month follow-up (primary) and 12 month follow-up (secondary), we will use a repeated measures design and a multilevel or mixed effect modeling framework for continuous outcomes (e.g., proportion of weeks with active SI, depression) and a mixed effect logistic regression model will be used for binary outcomes (e.g., suicide events). For all hypotheses, we will treat the individual as the random effect. Treatment effects will be estimated as the weighted average of the differences between ETAU and STEP groups over all follow-up times, adjusting for follow-up time.

Ethics approval and consent to participate

This study was approved by the Butler Hospital Institutional Review Board (Study # 2104–002) serving as the Single IRB for the other institutions in this multisite study including Boston Children’s Hospital, Bradley Hospital, and Brown University. All study materials and protocols have been ethically reviewed, both in English and Spanish where appropriate. All substantial amendments will be reviewed by the Butler Hospital Institutional Review Board. Informed consent for the parents and informed assent for the children will be obtained from all participants.