Single-Session Intervention Targeting Self-Injurious Behavior in Adolescents (Project SAVE)

Participants (aged 13-16) will be randomized to: (1) an online, active control group program encouraging feelings disclosure (i.e. supportive therapy single session intervention [SSI]), or (2) an online program targeting nonsuicidal self-injury (NSSI; i.e. Project "SAVE"-Stop Adolescent Violence Everywhere-SSI). Investigators will test whether SAVE results in significantly greater: Reductions in NSSI 3 months post-treatment [primary outcome] Reductions in self-reported likelihood of future NSSI immediately post-treatment [secondary outcome] Reduction in suicide ideation frequency 3 months post-treatment [secondary outcome]

Most mental health problems emerge by age 14, often leading to chronic impairments and adverse impacts for individuals, families, and societies. Similarly, suicidal thoughts and behaviors are relatively rare during childhood but increase significantly during the transition to adolescence. Self-injurious thoughts and behaviors (SITBs), which include suicidal thoughts and behaviors and nonsuicidal self-injury (NSSI; intentional self-harm enacted without suicidal intent), are among the best predictors of future suicidal thoughts and attempts in the current state of literature on the topic. There is also significant escalation from suicidal thoughts to suicidal behaviors during this developmental period. Most youth who transition from suicidal thoughts to suicidal behaviors will do so within 1-2 years after the onset of suicide ideation, which tends to begin in adolescents. Taken together, adolescence is a key developmental period for effective depression and suicide intervention and prevention. The goal of these studies is to test Single Session Interventions (SSIs) (defined below) that are designed to boost coping and resilience in adolescents. The investigators hope that this research will be valuable in ensuring that these SSIs are accessible to more adolescents. Despite significant strides in the development of psychosocial treatments for youth mental health problems, up to 80% of youth in the United States with mental health needs receive no services at all. Even when services are accessed, the definition of "treatment" is incredibly broad, with evidence-based interventions being the exception, not the rule. More typically, "treatment" may involve a single, unstructured contact with a medical doctor, counselor, emergency hotline, religious leader, or other health professional (e.g., occupational therapist), the benefits of which are unknown. This reality poses a critical challenge to the field of clinical intervention science-one that has been articulated many times before. Eight years ago, Kazdin and Blase (2011) called for a "rebooting" of psychotherapy research and practice, stating that "mental health professionals are not likely to reduce the prevalence, incidence, and burden of mental illness without a major shift in intervention research and clinical practice" via a portfolio of novel approaches to service delivery. Yet again, Kazdin (2019) asked us to reconsider how psychotherapy can and should be delivered: He argues that without removing requirements of a face-to-face format, an 'expert' with years of training, and a physical office, the need-to-access gap will likely stay stagnant. Indeed, Kazdin (2019) asserts the need to think about therapeutic action through an entirely new lens, asserting that "interventions may or may not rely on the psychological treatments that continue to dominate research." Any action-focused path to reducing the need-to-access gap will require moving beyond the dominant settings, formats, and systems that have constrained intervention delivery to date. Indeed, as Chorpita (2019) asserts, "it may well be time we move past thinking about 'treatments' as our only form of solutions" (p. 475), with "treatments" referencing the dominant design of current psychosocial therapies: weekly, clinician-delivered interventions in brick-and mortar clinical settings, which are too often inaccessible to those they are designed to serve. Brief, online interventions: Even among youths who do access services, treatment is often brief: U.S. youths who begin therapy attend an average of ~4 sessions, and the modal number of sessions attended is one. This creates a need to quantify and capitalize on what can be accomplished, given appropriate targeting and structure, in a short period of time. Single-session interventions (SSIs) have been found to benefit youth and help reduce depression symptoms. SSIs are brief and could be easily accessed online, thus, potentially limiting barriers to accessing treatment. Therefore, SSIs could markedly increase the capacity to address teen mental illness and SITBs in a cost-effective manner (Barak & Grohol, 2011). Indeed, SSIs can successfully reduce mental health problems: In a meta-analysis of 50 randomized trials including 10,508 youth participants (ages 4-19) SSIs significantly reduced psychopathology of multiple types (mean g = 0.32), including self-administered SSIs-those that did not involve a therapist (e.g., online SSIs; mean g = 0.32). To date, two types of SSIs have been shown to reduce depressive symptoms in adolescents and/or young adults. However, only one of these two SSIs is designed for online administration (the other is administered by a therapist and therefore more challenging to disseminate). In the present series of studies, the investigators aim to test the efficacy of SSIs self-administered online to reduce depression and SITBs in a large sample of teenagers. The Current Research The goal of this study is to test single session interventions (SSIs) that are designed to boost coping and resilience in adolescents. Specifically, participants (aged 13-16) will be randomized to: (1) an online, active control group program encouraging feelings disclosure (i.e. supportive therapy SSI), or (2) an online program targeting nonsuicidal self-injurious behavior (i.e. Project "SAVE"-Stop Adolescent Violence Everywhere-SSI). The investigators will test whether participants randomized to the SAVE SSI report: Lower self-reported likelihood of future NSSI immediately post-intervention Lower self-reported frequency of non-suicidal self-injury (in the past 3 months) at 3 month follow-up Lower self-reported frequency of suicide ideation (in the past 3 months) at 3 month follow-up

We will use a between-subjects design. Adolescents (13-16) recruited online via social media who report elevated self-hate and past month engagement in nonsuicidal self-injury will be randomized to the SAVE (active) or psychotherapy support (control)-SSI.

Random assignment to condition will be conducted in a triple-masked manner. Participants will be masked to whether they received active treatment and investigators will be masked to which condition the participant is randomized to-as the randomization occurs automatically within the Qualtrics survey. Additionally, as the primary and secondary outcomes are self-report measures, we will not be utilizing an outcomes assessor (e.g. a clinical interviewer).

SAVE is a ~30-minute, self-administered, web-based program that uses components of cognitive behavior therapy and dialectical behavior therapy designed to decrease self-injurious behaviors in youth. The Project SAVE SSI has 4 general content sections: (1) explaining the science behind how changing your actions (i.e. decreasing self-injurious behaviors) can positively impact your emotions over time; (2) providing scientific evidence and testimonials from other teens that have successfully decreased their self-injurious behaviors and noticed positive change as a result; (3) evidence-based tips for overcoming common obstacles to decreasing self-injurious behaviors in day to day life; and (4) offering an opportunity for youth to share their own thoughts and advice on what they have learned with other teenagers who are facing similar challenges.

Supportive Therapy SSI (Schleider & Weisz, 2018): ~30-minute, self-administered, web-based program that uses components of supportive therapy to encourage feelings sharing. The supportive therapy SSI encourages participants in the control group to identify and express their feelings by (1) explaining why sharing feelings is natural, important, and helpful and (2) including testimonials from teens who have shared their feelings with close others.

Internet-based, 30-minute single session intervention targeting NSSI via decreasing urge to act on self-punishment/self-harm thoughts/urges.

Our primary outcome variable will compare past 3-month NSSI frequency at 3-month follow-up-for participants assigned to the Project SAVE SSI versus supportive therapy SSI (i.e. control condition). Participants will indicate "how many times they have purposely hurt themselves without wanting to die" in the past 3 months in an open-response text box. Notably, in the present study's survey flow, this question is displayed following display logic-such that participants endorsing zero NSSI in the past 3-months (via a previous question) skip to the end of the question block and are not prompted to answer our main frequency outcome question. For these individuals, we will impute a past 3-month frequency value equal to 0 (as they have endorsed zero NSSI in the previous 3 months in the earlier question).

Likelihood of future NSSI will be measured immediately pre- and post-intervention. We will evaluate self-reported likelihood of future non-suicidal self-injury immediately post-intervention-comparing participants in the Project SAVE SSI versus supportive therapy SSI control conditions, and controlling for pre-intervention likelihood score. An item from the Self-Injurious Thoughts and Behaviors Interview-R (Fox et al., 2020) will assess participants' beliefs about the likelihood of future non-suicidal self injury on a 5-point Likert Scale (range = 0 to 4; 0 = "not at all"; 4 = "extremely").

We will compare past 3-month SI frequency at 3-month follow-up-for participants assigned to the Project SAVE SSI versus supportive therapy SSI (i.e. control condition). At pre-intervention and 3-month follow-up time points, participants will indicate "how many days did they have thoughts about killing themselves for more than a few minutes" in the past 3 months in an open-response text box. Notably, in the present study's survey flow, this question is displayed following display logic-such that participants endorsing zero history of SI (via a previous question) skip to the end of the question block and are not prompted to answer our frequency outcome question. For these individuals, we will impute a past 3-month frequency value equal to 0 (as they have endorsed zero SI in the previous question). More information about how we will handle very large numbers (i.e. outliers) for this outcome is reported below (see "Outliers and Exclusions" section; "Winsorizing Outliers" subsection).

Inclusion Criteria: Must be 13 - 16 years old at the time of the baseline survey Must report comfort reading and writing in English Must endorse no learning disability, visual impairment, or other difficulty that makes it difficult to answer questions on a computer Must have access to a laptop or smartphone with internet access Must endorse purposefully hurting self without wanting to die within the past month Must endorse either "I do not like myself" or "I hate myself" on CDI-II item 7 in a screener survey Exclusion Criteria: Type 1 exclusions are exclusions that can be automated and do not require consensus among investigators. A research team member who will not be involved with conducting the primary study analyses will use type 1 exclusions to determine when participant recruitment is complete. Recruitment for the present study will end once 500 participants have been randomized who pass all type 1 exclusions tests (i.e. are still eligible for inclusion after excluding for type 1 criteria; see our section on sample size for more details). Type 1 exclusions include: Not meeting inclusion criteria Participants who exit the study prior to condition randomization for our listed analyses (note: the present study will use an intention-to-treat approach, where every participant who was randomized to a study condition-and who is not excluded for one of the specific reasons outlined in this section-will be included our our analyses) Type 2 Exclusions Type 2 exclusions are exclusions that cannot be automated and require consensus among investigators. Participants excluded for type 2 reasons will be subtracted from the N = 500 total individuals who were randomized and met criteria for inclusion following type 1 exclusions. Here, we will exclude the following participants based on failure to meet the following quality check criteria: Participants who respond with either copy/pasted responses from text earlier in the intervention (e.g. copy and pasting only text from a previous testimonial slide) to any of free response questions Participants demonstrating an obvious lack of English fluency in open response questions; these participants will be identified by consensus from all study team members Participants responding with random text in open response questions; these participants will be identified by consensus from all study team members Duplicate responses from the same individual in baseline or follow-up surveys (i.e. more than one response with an identical IP address). Where duplicate responses for the same individual are present in the same survey (e.g. more than 1 response for a single individual at follow-up), we will exclude the response that is less complete, retaining the more complete of the two responses. If both responses are 100% complete, we will retain the first of the two responses for that survey. Notably, if an individual completes the baseline survey more than once-and happens to be randomized to both conditions-we will exclude this individual's responses from our analysis altogether. We will also exclude for primary analyses (but may run sensitivity analyses including them) any participants who provide responses of 3 words or fewer to writing prompts that ask for at least 2 sentences or more. These exclusions are based on previous single session intervention research conducted online (for one example, see Schleider & Weisz, 2018: https://doi.org/10.1080/15374416.2017.1405353)

The data will be de-identified both by removing identifying information and any other procedures necessary to keep the data from being re-identified using known methods. The independent and dependent variables relevant to the primary outcomes will be shared

Upon submission of the manuscript to a peer-reviewed journal. The data will be available open access thereafter.

Dobias ML, Schleider JL, Jans L, Fox KR. An online, single-session intervention for adolescent self-injurious thoughts and behaviors: Results from a randomized trial. Behav Res Ther. 2021 Dec;147:103983. doi: 10.1016/j.brat.2021.103983. Epub 2021 Oct 7.