Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Oct 1.
Published in final edited form as: Suicide Life Threat Behav. 2023 Aug 21;53(5):870–879. doi: 10.1111/sltb.12989

Safety Plan Use in the Daily Lives of Adolescents after Psychiatric Hospitalization

Alexis M May 1,*, Nadia Al-Dajani 2,*, Elizabeth D Ballard 3, Ewa Czyz 4
PMCID: PMC10592210  NIHMSID: NIHMS1922737  PMID: 37605441

Abstract

Introduction:

Safety planning type interventions (SPTI’s) are brief suicide-specific interventions. Little is known about safety plan use during high-risk periods, and whether safety plan use is influenced by baseline characteristics. This study examined how adolescents recently hospitalized for suicide risk use their safety plans post-discharge, tested moderators of safety plan utilization and explored the relationship between changes in utilization and changes in suicidal ideation (SI) over time.

Methods:

Seventy-eight adolescents hospitalized for suicide risk who participated in a pilot trial of safety planning responded to one survey/day for 4 weeks post-discharge and completed a 1-month assessment.

Results:

Over 90% of adolescents reported having access to their safety plan during the month post-discharge. Safety plan use and SI declined over time. No baseline characteristics predicted safety plan use in the 4 weeks after discharge, or changes in safety plan use over time. However, the relationship between changes in safety plan use and changes in SI was moderated. For girls, SI and safety plan use rose and fell together; for boys, safety plan use declined regardless of changes in SI.

Conclusions:

High risk adolescents retain and use their safety plans. Results underscore the importance of looking at sex effects on SPTI utilization.

The prevalence of suicidal thinking, behavior and death continues to grow among adolescents in the United States. Over forty percent of high school students reported persistent feelings of sadness or hopelessness in 2021, an increase from 29% in 2011, while the percentage of students who seriously considered suicide increased from 16% to 22% over the last ten years (Centers for Disease Control and Prevention [CDC], 2023). The rate of suicide attempts for adolescents in grades 9–12 has increased from 8% in 2011 to 10% in 2021. In 2020, suicide was the second leading cause of death for children aged 10–14 within the United States and the third leading cause of death for youth 15–24 years (National Center for Health Statistics, 2020).

Brief, scalable, and effective interventions are needed to stem the tide. Safety plans, crisis response plans, and coping cards, subsequently referred to as safety planning type interventions (SPTIs), are one such type of intervention that have been increasingly deployed in clinical settings. SPTIs originated as a part of longer-term suicide-specific treatments (e.g., brief cognitive behavioral therapy, dialectical behavior therapy, cognitive therapy for suicide prevention) and were later modified into standalone interventions (Bryan et al., 2017; Stanley & Brown, 2012). While SPTIs were originally used with adults presenting with elevated risk of suicide, they have been increasingly used in a wider range of populations (e.g., adolescents), settings (e.g., emergency departments) and via different modalities (e.g., group therapy, self-guided web applications) (Goodman et al., 2021; Hill, Dodd, Gomez, Do, & Kaplow, 2020; Miller et al., 2017). In most SPTIs, individuals generate personal coping strategies for use when suicidal urges reemerge in the future. In addition, SPTIs typically include identification of suicide warning signs, social supports, as well as professional and emergency resources

Emerging evidence among adults suggests SPTIs are efficacious in reducing suicidal behavior. A systematic review of the Safety Planning Intervention noted decreases in suicidal ideation, depression, hopelessness, and attempts (Ferguson, Rhodes, Loughhead, McIntyre, & Procter, 2021). A meta-analysis of six controlled trials of SPTIs found reductions in suicidal behaviors, but no difference in suicidal ideation (Nuij et al., 2021). Moreover, SPTIs are widely recommended as best practice strategies by organizations such as the National Institute for Health and Care Excellence (NICE Public Health and Social Care Guidelines Team, 2018), the Suicide Prevention Resource Center (Suicide Prevention Resource Center, 2022), and National Action Alliance for Suicide Prevention (National Action Alliance for Suicide Prevention, 2019). Further, SPTIs are required for behavioral healthcare accreditation by the Joint Commission as a part of discharge planning for individuals identified as at risk of suicide (The Joint Commission, 2019). Recent work has supported the feasibility of deploying SPTIs in pediatric emergency departments (Bettis, Donise, MacPherson, Bagatelas, & Wolff, 2020) and in university counselling centers (Stewart et al., 2020), however examinations of the efficacy of SPTIs in youth are lacking.

Despite the early signs of efficacy and the wide adoption of SPTIs, little is understood about how SPTIs work in practice. For example, there are several outstanding questions about SPTI usage, such as whether the plans are retained after they are provided, how frequently they are used after the initial intervention, and whether different demographic or clinical characteristics predict their use. One investigation of SPTI use in service members reported that the majority (over 80%) retained their physical plan but most did not keep it on their person in the 6 months post intervention (Bryan et al., 2018). Participants were most likely to recall the social supports section of the plan and least likely to recall the warning signs section. Even less is known about how adolescents use their SPTIs, if they differ from adults in how they remember and perceive the initial safety planning session, whether they keep track of the safety plan itself, and how often use the plan. Given important differences between adolescents and adults in developmental stage and lifestyle, it is important to further understand characteristics of SPTI use in this population specifically.

Variations in how suicidal thoughts and behaviors present may impact how interventions such as SPTIs are perceived, remembered and used. Suicidal thoughts and behaviors are heterogenous, with vastly different patterns based on demographic and clinical characteristics. For example, suicide deaths are more common in boys than girls (Rhodes et al., 2012), while suicidal thoughts and attempts are more common among girls (CDC, 2021). The role of suicidal ideation among boys may differ – for example, some findings suggest suicidal ideation severity does not predict future suicide attempts for boys, while it does for girls (King, Jiang, Czyz, & Kerr, 2014) though others find it is predictive of attempts across sexes (Horwitz, Czyz, & King, 2015; Thompson & Light, 2011). Across sex, individuals with multiple suicide attempts tend to have a different clinical course than those with no or one attempt, presenting with more psychiatric diagnoses and with a greater likelihood of a future attempt (Miranda et al., 2008). Finally, confidence in using coping strategies typically included as part of SPTIs could also impact use, particularly as lower self-efficacy in using these strategies is associated with higher risk for subsequent suicide attempts and related crises in youth (Czyz et al., 2016). Thus, sex, multiple attempt status, and coping self-efficacy are potentially important moderators of SPTI utilization.

This study leverages data from a pilot trial of safety planning and different post-discharge contacts among hospitalized youth. The primary analyses found that safety planning augmented with post-discharge text messages or a booster call was associated, preliminarily, with greater likelihood of safety plan use (Czyz et al., 2021). The aim of this secondary data analysis is to explore how adolescents use their safety plans in the month after discharge and how safety plan use changes week-to-week over this high-risk period. Specifically, this paper examines: 1) how many youth kept and used their safety plan; 2) whether changes in suicidal ideation and changes in safety plan use are related during the post-discharge period; and 3) whether key baseline youth characteristics (i.e. sex, multiple suicide attempt history, self-efficacy) predict safety plan use or moderate the relationship between suicidal ideation and safety plan use in the first month following hospitalization.

Methods

Participants and Procedures

Adolescents (age range 13–17) psychiatrically hospitalized due to suicide risk (i.e., suicidal ideation with method, intent, or plan in the past week and/or past month suicide attempt) were recruited for this study from a single hospital. Exclusion criteria included: (1) altered mental status (e.g., mania, psychosis); (2) severe cognitive impairment; (3) transfer to medical unit/residential treatment; (4) no legal guardian (e.g., ward of state); and (5) no access to cell phone. Eligibility was determined based on admission records and consultation with the treatment team. Participants were recruited between March 2019 – January 2020 as part of a pilot psychosocial intervention study incorporating safety planning prior to discharge with post-discharge contacts via text messages or booster calls (Czyz et al., 2021).

Ninety-four individuals were approached to participate, of which 82 (87.2%) provided both parental consent and adolescent assent. A total of 80 adolescents completed the baseline assessment prior to hospital discharge and initiated the daily survey protocol the day following discharge from hospital. Daily surveys were sent via text to participants every evening for four weeks. Participants could receive up to $222 in study compensation (i.e., $4/daily survey completed). Daily survey responses were monitored by on-call staff who contacted participants reporting current ideation with intent/plan or a suicide attempt in the last 24 hours. In addition, an automated message with crisis contact information was sent to participants if they endorsed any suicidal thoughts. The analytic sample includes 78 teens, as two individuals did not complete any daily surveys. At 1-month follow-up, 75 (96.2%) participants were retained. This study was approved by the Institutional Review Board at the University of Michigan and conforms to the Declaration of Helsinki standards.

Baseline and 1-month Follow-up Measures

Baseline multiple suicide attempt history.

Multiple suicide attempt history was assessed using the Columbia-Suicide Severity Rating Scale (C-SSRS; Posner et al., 2011), obtained via medical record review; the C-SSRS is used as part of clinical protocol prior to admission.

Baseline self-efficacy to manage suicidal thoughts.

The Efficacy to Cope with Suicidal Thoughts and Urges Scale (Czyz et al., 2016) a 12-item measure that was used to assess level of confidence in using 12 coping responses that map onto components of safety plans (e.g., use distraction or relaxation strategies, seek personal or professional supports) when experiencing suicidal ideation. Participants rated their level of confidence on a 10-point scale, with 0 = not at all confident and 10 = extremely confident. Internal consistency in this sample was 0.91.

Safety plan presence at follow-up.

At 1-month follow-up, participants were asked “Do you still have a copy of your plan?” Responses options were: 1 = Yes, 2 = No, 3 = Not sure, and 4 = Never got one.

Daily Survey Measures

Daily Suicidal Ideation Frequency.

Each evening, participants were asked to indicate the frequency of suicidal thoughts within the last 24 hours (“How many times did you have thoughts of killing yourself?”) using a 4-point Likert scale ranging from not at all (0) to all the time (4). This item was modeled after the C-SSRS (Posner et al., 2011) and was previously used to assess daily suicidal ideation (Czyz, King, & Nahum-Shani, 2018).

Daily Safety Plan Use.

Each evening, participants were asked the following: In the last 24 hours, did you think about, look at, or use your safety plan? Response options included “not at all,” “a little,” or “a lot.”

Data Analytic Plan

All statistical analyses were conducted using R version 4.03 (R Core Team, 2020). First, we provided characteristics of safety plan retention and use reported via daily surveys over the 4-week post-discharge period. Changes in safety plan use over this period were graphed using ggplot2 (Wickham, 2016). Second, we explored the relationship between baseline characteristics (i.e., multiple attempt history, self-efficacy, sex) and post-discharge safety plan use. We analyzed two types of models: (1) safety plan use across the 4-week period examined in terms of overall mean safety plan use (mean-level models) and; (2) stability of safety plan use defined as overall slope over this period (slope models). G*Power was used to calculate an a priori power analysis - identifying a sample size of 77 is needed to detect an effect size f2 of 0.15 or higher, which is equivalent to Cohen’d d of at least 0.30, at power = 0.80. This suggests these analyses were sufficiently powered to detect effects of this size or greater.

Mean-Level Models.

For each participant, an aggregate mean-level score for suicidal ideation frequency and for safety-plan use were calculated. A series of linear models were conducted to investigate if baseline variables (i.e., sex, multiple suicide attempt history, and self-efficacy score) predicted overall mean of safety plan use (outcome variable). Three separate models (one for each baseline variable) were examined. In each model, mean-level suicidal ideation frequency and intervention indicators (i.e., intervention group that each person was randomized to) were included as covariates. We additionally examined if an interaction between the baseline variables of interest and mean suicidal ideation frequency predicted the overall mean of safety plan use. If this interaction was non-significant, it was not included in the final model.

Slope Models.

For each participant, slope of suicidal ideation frequency and slope of safety-plan use were calculated using tidyverse (Wickham et al., 2019) and dplyr (Wickham, François, Henry, & Müller, 2021). Slope variables take into consideration the rate of change over the 4-week period, along with the direction of change. For instance, a negative slope for suicidal ideation frequency suggests that suicidal ideation frequency reduced over the 4-week period while a positive slope for suicidal ideation frequency suggests that it increased. The value of the slope indicates the magnitude of this change, with a slope of 0 indicating no change over the 4-week period. A series of linear models were conducted to investigate if baseline variables (i.e., multiple attempt history, self-efficacy, sex) predicted safety plan use slope (outcome variable). Three separate models (one for each baseline variable) were examined. In each model, mean-level safety plan use, mean-level suicidal ideation frequency, slope of suicidal ideation frequency, and intervention indicators were included as covariates. The mean-level covariates were included to account for differences in average safety plan use across individuals and to investigate the association between rate of change across the 4-week period above and beyond average levels (which were examined in prior models). We additionally considered if an interaction between each baseline variable of interest and slope of suicidal ideation frequency was associated with safety plan use slope. If this interaction was non-significant, it was not included in the final model.

Results

Mean age of participants was 15.2 years (SD = 1.4) and 69% reported that they were assigned female sex at birth. The sample’s ethnoracial breakdown was (could check all that apply): 83.3% (n=65) White, 6.4% (n=5) African-American/Black, 5.1% (n=4) Asian, 5.1% (n=4) American Indian or Alaska Native, and 1.3% (n=1) Native Hawaiian or Other Pacific Islander, 2.6% (n=2) Other. Nine participants (11.5%) self-identified as Hispanic. Additional sample details can be found in Czyz et al. (2021). Throughout the 4-week sampling period, 1621 daily observations were recorded (adherence rate: 72.4%). Of these observations, 631 (38.9%) captured instances of suicidal ideation. On average, adolescents responded to 20.78 surveys (SD = 6.92, range = 1–28 surveys/participant), with suicidal ideation instances ranging between 0–28 across the assessment period (M = 9.86, SD = 7.91). Of note, 82.1% of participants reported at least one instance of suicidal ideation. At baseline, 35% (n = 27) of the study sample had a history of multiple suicide attempts, and mean baseline self-efficacy was 67.51 (range = 6–120).

Characteristics of Safety Plan Use

On average, participants reported using their safety plan on 48.0% of observation days across the 4-week period (778 out of 1621 possible observations). Seventy participants (90%) indicated safety plan use on at least one day across the sampling period. Of the 745 instances of safety plan use, in 408 instances (54.8%) participants reported no suicidal ideation and in 337 (45.2%) they reported the presence of ideation. Of the 631 instances of suicidal ideation, in 294 (46.6%) participants reported no safety plan use and in the 337 (53.4%) they reported using their safety plan. At the 1-month follow-up assessment, 90.7% (n = 68) of 75 respondents who completed the follow-up indicated that they still had access to their safety plan, 2.7% (n = 2) reported they did not have access to their safety plan, and 6.7% (n = 5) indicated they were unsure.

Figure 1 illustrates change in safety plan use over the 4-week post-hospitalization period, along with change in suicidal ideation frequency over the same period. As shown, safety plan use declined as time went on, as did suicidal ideation frequency. Of the total number of available observation days any given week, youth reported using their safety plan 54.8% of the days in week 1 (249 out of 454 observations), 47.5% of the days in week 2 (202 out of 425 observations), 43.4% of the days in week 3 (164 out of 378 observations), and 35.5% of the days in week 4 (130 out of 366 observations). Multilevel logistic regression models revealed that safety plan use decreased significantly between weeks 1 and 2 (p < .000), did not change between weeks 2 and 3 (p = .266), and dropped significantly between weeks 3 and 4 (p = .009). With regards to changes in suicidal ideation frequency across weeks, multilevel logistic regression models showed that suicidal ideation frequency followed a similar pattern in that it was significantly lower in week 2 than in week 1 (p < .000) and did not change between weeks 2 and 3 (p = .283). However, in contrast to a decreasing patten in safety plan use, suicidal ideation did not change significantly between weeks 3 and 4 (p = .625).

Figure 1.

Figure 1.

Open in a new tab

Change in Safety Plan Use and Suicidal Ideation Frequency Over Time.

Note. Scale range for safety plan use (1 = not at all, 3 = a lot). Scale range for suicidal ideation frequency (0 = not at all, 4 = all the time). Average safety plan use and average suicidal ideation frequency were calculated per day across participants. SI = suicidal ideation.

Mean and Slope Models

Table 1 illustrates findings for the three models with baseline variables predicting aggregate mean safety plan use. As observed, none of the baseline characteristics were associated with aggregate average safety plan use across the 4-week period, over and above covariates. Additionally, none of the baseline characteristics interacted with aggregate average suicidal ideation frequency to predict mean safety plan use. Table 2 illustrates findings for the three aggregate slope (direction and magnitude of change over the 4-week period) models. Baseline multiple attempt history1 and baseline self-efficacy ratings did not predict slope of safety plan use. Neither of these baseline variables interacted with direction/magnitude of change (slope) of suicidal ideation frequency to predict slope of safety plan use. However, as shown in Table 2, the interaction between baseline sex and slope of suicidal ideation frequency was significantly associated with slope of safety plan use (ΔR = .097).2

Table 1.

Associations between Baseline Variables and Aggregate Mean Safety Plan Use.

Aggregate Mean Models
Predictor variables b SE p Model R 2
 Step 1: Main Effects
Multiple Attempt History 0.08 0.12 .511 0.06
Self-efficacy 0.00 0.08 .362 0.07
Sex −0.12 0.13 .358 0.06
 Step 2: Interaction Terms
- - - - -
Open in a new tab

Note. Each row represents a separate model. If the interaction term was non-significant, it was removed from the model. Significant interaction terms are highlighted in Step 2 of the table. All predictor variables were measured at baseline. Outcome variable = mean safety plan use. Covariates included in all models: intervention indicators, mean suicidal ideation frequency.

Table 2.

Associations between Baseline Variables and Aggregate Slope of Safety Plan Use.

Aggregate Slope Models
Predictor variables b SE p Model R 2
 Step 1: Main Effects
Multiple Attempt History −0.01 0.01 .204 0.08
Self-efficacy −0.00 0.00 .746 0.06
Sex −0.00 0.00 .269 0.08
 Step 2: Interaction Terms
Sex x SI slope −0.03 0.09 .006 0.17
Open in a new tab

Note. Each row represents a separate model. If the interaction term was non-significant, it was removed from the model. Significant interaction terms are highlighted in Step 2 of the table. All predictor variables were measured at baseline. Outcome variable = slope safety plan use. Covariates included in all models: intervention indicators, mean suicidal ideation frequency, mean safety plan use, slope suicidal ideation frequency. SI = suicidal ideation.

Interaction of Sex by Change in Suicidal Ideation Frequency Over Time predicting Change in Safety Plan Use Over Time

Visual inspection of the significant sex by slope of suicidal ideation frequency interaction was conducted to further interpret this finding. As seen in Figure 23, decreasing safety plan use over time was associated with decreasing suicidal ideation frequency in females and increasing safety plan use over time was associated with increasing suicidal ideation frequency in females. This pattern was not evident in males, such that safety plan use over time was decreasing whether or not suicidal ideation frequency was decreasing. Findings for males demonstrate a general pattern of decreasing safety plan use irrespective of direction of change in suicidal ideation frequency. This interpretation is confirmed by a simple slopes test, finding that the relationship between suicidal ideation frequency slope and safety plan use slope is significant in female participants (B = 0.24, SE = 0.09, p = .007) but not in male participants (B = −0.01, SE = 0.02, p = .640).

Figure 2.

Figure 2.

Open in a new tab

Interaction between Sex and Slope of Suicidal Ideation Frequency Predicting Slope of Safety Plan Use.

Note. SD = standard deviation. −1 SD indicates decreasing suicidal ideation frequency over time, while +1 SD indicates increasing suicidal ideation frequency over time.

Discussion

Safety plans are a commonly used brief intervention in suicide prevention and have been widely disseminated across mental health care settings. However, very little is known about how SPTIs are actually used in practice by at-risk youth, and their correspondence with suicidal thoughts over time. This study sought to describe how psychiatrically hospitalized adolescents participating in a safety plan trial used their plans after discharge. Overall, results showed that the vast majority (>90%) of adolescents had access to their safety plan during the first month after discharge and that plans were often used on days without suicidal ideation. Further, overall safety plan usage gradually declined over time, as did thoughts of suicide. Notably, there were differences between boys and girls in the relationship between suicidal thoughts and safety plan usage.

Nearly all participants reported having access to their safety plan one month after discharge from the hospital. Given the challenges adolescents often face after a hospitalization, it is notable that so many plans were maintained. Retention may have been higher than in general clinical settings, as all participants received a safety planning intervention enhanced with motivational interviewing strategies. Nevertheless, this finding suggests that providing a physical copy of the safety plan did not appear to lead to its loss, even among adolescents. While digital plans may be valuable because of their portable and ubiquitous nature, plans that can be visually displayed (e.g., on a bulletin board, bathroom mirror, etc.) may have particular salience. Additionally, particularly for youth, it may be important to keep the plan physically separate from the phone, due to a propensity for phones to be lost or subject to disciplinary restrictions, incidents which themselves may indicate times of heightened distress. As safety planning interventions are increasingly offered online or via mobile apps, further research is needed to understand the impact of a physical versus a digital plan across age groups.

While perhaps counter to expectations, approximately half of the instances of safety plan use were on days without suicidal ideation. This suggests that adolescents may be using their safety plans very early in the suicidal process (e.g., using the plan at the first warning sign or reviewing the plan prior to an anticipated stressor) and avoiding the development of suicidal ideation. Alternatively, they may be using the safety plan to cope with other distressing emotions or avoid other unwanted behaviors (e.g., substance use). Future work can help clarify what prompts adolescents to use their safety plans and whether safety plans help support effective coping more generally.

We also observed a trend in which, on average, safety plan use dropped approximately 10 percentage points in the first week following discharge, then leveled out in weeks 2 and 3 before decreasing again in the fourth week. In contrast, while suicidal ideation similarly decreased in the first week following discharge, it then stayed consistent across weeks 2–4. This suggests that there might be room for improvement in increasing the “stickiness” of safety plan use. While plans may be used more frequently after their initial provision, use wanes over time, despite suicide risk remaining high throughout the months after discharge. Refresher sessions, digital reminders, as well as ongoing review and revision of the plan with a therapist might help mitigate the reduction in its use over time. Revisiting safety plans as part of ongoing treatment is already an important element of comprehensive interventions for individuals at risk for suicide, and pilot work has shown that providing either a booster call or text-based support focused on coping and safety plan use yielded greater safety plan use in the weeks after discharge (Czyz et al., 2021). Another possibility is that youth may have internalized the safety plan skills and resources– in which case, when suicidal ideation re-emerged several weeks after discharge, they did not explicitly report using the safety plan but were in fact using learned skills or resources. Understanding the specific reasons for decreasing safety plan use over time and the impact on future suicidal behavior represents an important area for future study.

In terms of baseline predictors of safety plan use across 4 weeks post discharge, no significant relationships were detected in the mean-level models, suggesting that average safety plan use itself was not associated with sex, multiple suicide attempt history, or beliefs in the ability to manage suicidal thoughts. However, a notable relationship between suicidal ideation and safety plan use was observed when examining slope-level models. Specifically, the relationship between the change in suicidal ideation frequency and the change in safety plan use was moderated by sex, such that for girls, suicidal ideation frequency and safety plan use rose (and fell) in tandem, while for boys there was no such relationship. For boys, regardless of the direction or rate of change of suicidal ideation frequency, safety plan use stayed the same. The difference observed between boys and girls is consistent with longstanding sex differences in suicidal thoughts and behaviors during adolescence. For example, adolescent girls are at higher risk of suicide attempts, while adolescent boys are at higher risk of suicide death (Miranda-Mendizabal et al., 2019). These findings suggest that boys may also respond to suicide prevention interventions differently than girls. Beyond biological sex differences, additional research is needed to understand patterns of safety plan use among gender minority youth.

Further, this study’s findings are consistent with prior work identifying the value of looking at patterns of change in suicidal thoughts and behaviors rather than simply looking at single-point severity. For example, findings emerging from dynamic systems theory support the nonlinearity by which suicidal thoughts change and suicidal behavior emerges (Bryan, 2022), emphasizing the importance of measuring fluctuations in suicidal ideation over time. In recently hospitalized adolescent specifically, suicidal ideation is not uniform and instead follows discrete longitudinal profiles of varying severity and chronicity (Czyz & King, 2015; Czyz et al., 2022; Wolff et al., 2018). A growing body of work has also identified subtypes of suicide experiences, one characterized by fluctuating suicidal ideation that leads to more unpredictable suicide attempts and another characterized by sustained suicidal ideation that can lead to planned attempts (Bernanke, Stanley, & Oquendo, 2017). Together, these findings all highlight the importance of evaluating not just static moments in time, but also patterns of change over time. While we focused on direction and rate of change through slopes, there are other metrics whereby fluctuations in SI could be investigated (e.g., standard deviation, RMSSD). Future work should explore distinct dynamics of ideation and safety plan use.

These findings should be viewed in context of several limitations. The sample is predominantly White and non-Hispanic, limiting generalizability to other racial and ethnic groups. This is a particularly important caveat in a time when rates of suicide death are rising precipitously among Black children and youth (Sheftall et al., 2022). The sample was also homogenous in that it consists of adolescents recently discharged from a psychiatric hospital. While this captures a particularly high-risk time for a high-risk group of adolescents, it restricts the generalizability of these finding to other youth with suicidal crises that did not result in hospitalization. Another important limitation is that this work does not describe a directional relationship between suicidal ideation and safety plan use. While our design is strengthened by the daily measurements of suicidal ideation and safety plan use, both were assessed concurrently and thus precluding testing the directionality of the relationship between safety planning use and suicidal ideation frequency. Future studies that assess suicidal ideation and safety plan use multiple times each day or with event-contingent reporting could further explore their temporal relationship. Finally, while group assignment was included as a covariate, it is possible that relationships among the predictors would be distinct across these groups. While we are unable to address these potential differences in our current study, future studies should explore if these findings can be replicated in a distinct sample.

This study also had several strengths. First, it addresses a novel question - how is a recommended brief intervention used after it is provided? Second, it focuses on a high-risk group during a particularly sensitive time when safety plan use is likely the most important. Third, the impact of recall bias is reduced by using a repeated measures design in which suicidal ideation frequency and safety plan use were recorded day by day. Finally, these initial findings suggest important future directions in understanding the use of SPTIs more generally and in youth specifically.

More work is needed to pinpoint the mechanism by which SPTIs confer benefit – for example is the intervention session itself most impactful, or is the quality of safety plans or the frequency of safety plan use key? Additionally, are SPTIs most beneficial prior to the experience of suicidal ideation (i.e., as a preventative approach)? Or during/following suicidal ideation? As provision of SPTI’s becomes more widespread and a larger proportion of the mental health care workforce is expected to provide them, understanding the active ingredient(s) of SPTIs could be beneficial in focusing training and innovation on the most impactful elements. Specifically, future studies could examine which elements of safety plans are used most often and which are perceived as most helpful by individuals at risk for suicide and what mechanisms account for the reduction in safety plan use, particularly among certain subgroups (e.g., males). It would be especially helpful to disentangle safety plan use per se from gradual internalizing of warning signs and coping skills. More broadly, replication of this work is necessary in light of aforementioned limitations, as well as extensions to examine the temporal relationship between real-time suicidal ideation and safety plan use across different populations of high-risk individuals and settings.

Acknowledgments:

This research was funded by NIMH K23MH113776 (Czyz, EK).

Footnotes

The authors have no conflicts of interest to disclose.

1

Mean-level models or slope models were run with a history of 1+ suicide attempts versus none. Results were similar to those of baseline multiple attempt history.

2

All models were also examined using generalized least squares with package RRPP (Collyer & Adams, 2019), as some variables violated assumptions of parametric tests. Findings remained the same across all models and only findings from parametric tests were thus reported.

3

To improve visual inspection of the graph, a male participant outlier was removed for having a slope of −1. This outlier was retained in analyses. Analyses were conducted both with and without the outlier, and results remained consistent.

References

  1. Bernanke JA, Stanley BH, & Oquendo MA (2017). Toward fine-grained phenotyping of suicidal behavior: The role of suicidal subtypes. Molecular Psychiatry. 10.1038/mp.2017.123 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bettis AH, Donise KR, MacPherson HA, Bagatelas P, & Wolff JC (2020). Safety Planning Intervention for Adolescents: Provider Attitudes and Response to Training in the Emergency Services Setting. Psychiatric Services (Washington, D.C.), 71(11), 1136–1142. 10.1176/appi.ps.201900563 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bryan C (2022). Rethinking suicide: Why prevention fails and how we can do it better. Oxford University Press. [Google Scholar]
  4. Bryan CJ, May AM, Rozek DC, Williams SR, Clemans TA, Mintz J, … Burch TS (2018). Use of crisis management interventions among suicidal patients: Results of a randomized controlled trial. Depression and Anxiety, 35(7). 10.1002/da.22753 [DOI] [PubMed] [Google Scholar]
  5. Bryan CJ, Mintz J, Clemans TA, Leeson B, Burch T, Williams S, … Rudd D (2017). Effect of crisis response planning vs. contracts for safety on suicide risk in U.S. Army Soldiers: A randomized clinical trial. Journal of Affective Disorders, 212(October 2016), 64–72. 10.1016/j.jad.2017.01.028 [DOI] [PubMed] [Google Scholar]
  6. Centers for Disease Control. (2020). Youth Risk Behavior Surveillance Data Summary & Trends Report: 2009–2019 Retrieved from https://www.cdc.gov/healthyyouth/data/yrbs/yrbs_data_summary_and_trends.htm
  7. Centers for Disease Control and Prevention. (2023). Youth Risk Behavior Survey data & summary.
  8. Czyz EK, King CA, & Nahum-Shani I (2018). Ecological assessment of daily suicidal thoughts and attempts among suicidal teens after psychiatric hospitalization: Lessons about feasibility and acceptability. Psychiatry Research, 267(June), 566–574. 10.1016/j.psychres.2018.06.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Czyz EK, King CA, Prouty D, Micol VJ, Walton M, & Nahum-Shani I (2021). Adaptive intervention for prevention of adolescent suicidal behavior after hospitalization: a pilot sequential multiple assignment randomized trial. Journal of Child Psychology and Psychiatry and Allied Disciplines. 10.1111/jcpp.13383 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Czyz Ewa K., Horwitz AG, Arango A, Cole-Lewis Y, Berona J, & King CA (2016). Coping with suicidal urges among youth seen in a psychiatric emergency department. Psychiatry Research. 10.1016/j.psychres.2016.04.118 [DOI] [PubMed] [Google Scholar]
  11. Czyz Ewa K., & King CA (2015). Longitudinal Trajectories of Suicidal Ideation and Subsequent Suicide Attempts Among Adolescent Inpatients. Journal of Clinical Child and Adolescent Psychology, 44(1), 181–193. 10.1080/15374416.2013.836454 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Czyz Ewa K, Koo HJ, Al-Dajani N, Kentopp SD, Jiang A, & King CA (2022). Temporal profiles of suicidal thoughts in daily life: Results from two mobile-based monitoring studies with high-risk adolescents. Journal of Psychiatric Research, 153, 56–63. 10.1016/j.jpsychires.2022.06.050 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ferguson M, Rhodes K, Loughhead M, McIntyre H, & Procter N (2021). The Effectiveness of the Safety Planning Intervention for Adults Experiencing Suicide-Related Distress: A Systematic Review.Archives of Suicide Research, 0(0), 1. 10.1080/13811118.2021.1915217 [DOI] [PubMed] [Google Scholar]
  14. Goodman M, Sullivan SR, Spears AP, Dixon L, Sokol Y, Kapil-Pair KN, … Stanley B (2021). An Open Trial of a Suicide Safety Planning Group Treatment: “Project Life Force.” Archives of Suicide Research, 25(3), 690–703. 10.1080/13811118.2020.1746940 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hill RM, Dodd CG, Gomez M, Do C, & Kaplow JB (2020). The Safety Planning Assistant: Feasibility and Acceptability of a Web-Based Suicide Safety Planning Tool for At-Risk Adolescents and Their Parents. Evidence-Based Practice in Child and Adolescent Mental Health, 5(2), 164–172. 10.1080/23794925.2020.1759469 [DOI] [Google Scholar]
  16. Horwitz AG, Czyz EK, & King CA (2015). Predicting Future Suicide Attempts Among Adolescent and Emerging Adult Psychiatric Emergency Patients. Journal of Clinical Child & Adolescent Psychology, 44(5), 751–761. 10.1080/15374416.2014.910789 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ivey-Stephenson AZ, Demissie Z, Crosby AE, Stone DM, Gaylor E, Wilkins N, … Brown M (2020). Suicidal Ideation and Behaviors Among High School Students - Youth Risk Behavior Survey, United States, 2019. MMWR Supplements. 10.15585/mmwr.su6901a6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. King CA, Jiang Q, Czyz EK, & Kerr DCR (2014). Suicidal ideation of psychiatrically hospitalized adolescents has one-year predictive validity for suicide attempts in girls only. Journal of Abnormal Child Psychology. 10.1007/s10802-013-9794-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Miller IW, Camargo CA, Arias SA, Sullivan AF, Allen MH, Goldstein AB, … Boudreaux ED (2017). Suicide prevention in an emergency department population: The ED-safe study. JAMA Psychiatry, 74(6), 563–570. 10.1001/jamapsychiatry.2017.0678 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Miranda-Mendizabal A, Castellví P, Parés-Badell O, Alayo I, Almenara J, Alonso I, … Alonso J (2019). Gender differences in suicidal behavior in adolescents and young adults: systematic review and meta-analysis of longitudinal studies. International Journal of Public Health. 10.1007/s00038-018-1196-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Miranda R, Scott M, Hicks R, Wilcox HC, Harris Munfakh J Lou, & Shaffer D (2008). Suicide attempt characteristics, diagnoses, and future attempts: Comparing multiple attempters to single attempters and ideators. Journal of the American Academy of Child and Adolescent Psychiatry, 47(1), 32–40. 10.1097/chi.0b013e31815a56cb [DOI] [PubMed] [Google Scholar]
  22. National Action Alliance for Suicide Prevention. (2019). Best practices in care transitions for individuals with suicide risk: Inpatient care to outpatient care. Washington, DC. [Google Scholar]
  23. National Center for Health Statistics. (2020). About multiple cause of death, 1999–2020. Hyattsville, MD: US Department of Health and Human Services, CDC, National Center for Health Statistics; 2021. [Google Scholar]
  24. NICE Public Health and Social Care Guidelines Team. (2018). Preventing suicide in community and custodial settings |. National Institute for Health and Care Excellence . [Google Scholar]
  25. Nuij C, Ballegooijen W Van Beurs D. De, Juniar D, Erlangsen A, Portzky G, … Riper H (2021). Safety planning-type interventions for suicide prevention : meta-analysis. The British Journal of Psychiatry, (419), 419–426. 10.1192/bjp.2021.50 [DOI] [PubMed] [Google Scholar]
  26. Posner K, Brown GK, Stanley B, Brent DA, Yershova KV, Oquendo MA, … Mann JJ (2011). The Columbia-suicide severity rating scale: Initial validity and internal consistency findings from three multisite studies with adolescents and adults. American Journal of Psychiatry, 168(12), 1266–1277. 10.1176/appi.ajp.2011.10111704 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Team RC. (2020). R: A Language and Environment for statistical computing. pbil.univ-lyon1.fr.
  28. Rhodes AE, Khan S, Boyle MH, Wekerle C, Goodman D, Tonmyr L, … Manion I (2012). Sex differences in suicides among children and youth: The potential impact of misclassification. Canadian Journal of Public Health. 10.1007/bf03403815 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Sheftall AH, Vakil F, Ruch DA, Boyd RC, Lindsey MA, & Bridge JA (2022). Black Youth Suicide: Investigation of Current Trends and Precipitating Circumstances. Journal of the American Academy of Child and Adolescent Psychiatry. 10.1016/j.jaac.2021.08.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Stanley B, & Brown GK (2012). Safety Planning Intervention: A Brief Intervention to Mitigate Suicide Risk. Cognitive and Behavioral Practice, 19(2), 256–264. 10.1016/j.cbpra.2011.01.001 [DOI] [Google Scholar]
  31. Stewart KL, Darling EV, Yen S, Stanley B, Brown GK, & Weinstock LM (2020). Dissemination of the safety planning intervention (SPI) to university counseling center clinicians to reduce suicide risk among college students. Archives of Suicide Research, 24(sup1), 75–85. [DOI] [PubMed] [Google Scholar]
  32. Suicide Prevention Resource Center. (2022). A Comprehensive Approach to Suicide Prevention.
  33. The Joint Commission. (2019). R3 Report National Patient Safety Goal for Suicide Prevention. [DOI] [PubMed]
  34. Thompson MP, & Light LS (2011). Examining Gender Differences in Risk Factors for Suicide Attempts Made 1 and 7 Years Later in a Nationally Representative Sample. Journal of Adolescent Health, 48(4), 391–397. 10.1016/j.jadohealth.2010.07.018 [DOI] [PubMed] [Google Scholar]
  35. Wickham H (2016). ggplot2 -Positioning Elegant Graphics for Data Analysis. Springer. [Google Scholar]
  36. Wickham H, Averick M, Bryan J, Chang W, McGowan L, François R, … Yutani H (2019). Welcome to the Tidyverse. Journal of Open Source Software. 10.21105/joss.01686 [DOI] [Google Scholar]
  37. Wickham H, François R, Henry L, & Müller K (2021). dplyr: A Grammar of Data Manipulation. R package version. Media. [Google Scholar]
  38. Wolff JC, Davis S, Liu RT, Cha CB, Cheek SM, Nestor BA, … Spirito A (2018). Trajectories of Suicidal Ideation among Adolescents Following Psychiatric Hospitalization. Journal of Abnormal Child Psychology, 46(2), 355–363. 10.1007/s10802-017-0293-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES