Accounting for diversity in the treatment of suicide and self-injury: A systematic review of the past 50 years of randomized controlled trials.

PURPOSE: Patients receiving treatment for self-injurious thoughts and behaviors (SITBs) have diverse backgrounds, yet it remains unclear exactly who is represented in the current SITB treatment literature. METHODS: We conducted a systematic review of the past 50 years of randomized controlled trials (RCTs) testing SITB treatments to evaluate sampling practices and reporting of sample characteristics, as well as inclusion of global populations across the included 525 papers. We also assessed changes over the past five decades in these three domains. RESULTS: SITB RCTs frequently reported age and sex (98.6%-95.1%), less frequently reported race (83.4%-38.6%), socioeconomic status (48.1%-46.1%) and ethnicity (41.9%-8.1%), and rarely reported LGBTQ+ status (3.7%-1.6%). U.S.-based RCTs featured predominantly White, non-Hispanic, and non-LGBTQ+ samples. Most RCTs were conducted in high-income North American or European countries. Sample reporting practices, sample representativeness, and inclusion of global populations modestly and inconsistently improved over time. CONCLUSIONS: There has not been substantial improvement in reporting practices, sample representativeness, or inclusion of global populations in SITB RCTs over the past 50 years. Acknowledging who is being studied and representing diverse populations in SITB treatment research is key to connecting research advances with those who may need it most.

External Validation and Updating of a Statistical Civilian-Based Suicide Risk Model in US Naval Primary Care.

Importance: Suicide remains an ongoing concern in the US military. Statistical models have not been broadly disseminated for US Navy service members. Objective: To externally validate and update a statistical suicide risk model initially developed in a civilian setting with an emphasis on primary care. Design, Setting, and Participants: This retrospective cohort study used data collected from 2007 through 2017 among active-duty US Navy service members. The external civilian model was applied to every visit at Naval Medical Center Portsmouth (NMCP), its NMCP Naval Branch Health Clinics (NBHCs), and TRICARE Prime Clinics (TPCs) that fall within the NMCP area. The model was retrained and recalibrated using visits to NBHCs and TPCs and updated using Department of Defense (DoD)-specific billing codes and demographic characteristics, including expanded race and ethnicity categories. Domain and temporal analyses were performed with bootstrap validation. Data analysis was performed from September 2020 to December 2022. Exposure: Visit to US NMCP. Main Outcomes and Measures: Recorded suicidal behavior on the day of or within 30 days of a visit. Performance was assessed using area under the receiver operating curve (AUROC), area under the precision recall curve (AUPRC), Brier score, and Spiegelhalter z-test statistic. Results: Of the 260 583 service members, 6529 (2.5%) had a recorded suicidal behavior, 206 412 (79.2%) were male; 104 835 (40.2%) were aged 20 to 24 years; and 9458 (3.6%) were Asian, 56 715 (21.8%) were Black or African American, and 158 277 (60.7%) were White. Applying the civilian-trained model resulted in an AUROC of 0.77 (95% CI, 0.74-0.79) and an AUPRC of 0.004 (95% CI, 0.003-0.005) at NBHCs with poor calibration (Spiegelhalter P < .001). Retraining the algorithm improved AUROC to 0.92 (95% CI, 0.91-0.93) and AUPRC to 0.66 (95% CI, 0.63-0.68). Number needed to screen in the top risk tiers was 366 for the external model and 200 for the retrained model; the lower number indicates better performance. Domain validation showed AUROC of 0.90 (95% CI, 0.90-0.91) and AUPRC of 0.01 (95% CI, 0.01-0.01), and temporal validation showed AUROC of 0.75 (95% CI, 0.72-0.78) and AUPRC of 0.003 (95% CI, 0.003-0.005). Conclusions and Relevance: In this cohort study of active-duty Navy service members, a civilian suicide attempt risk model was externally validated. Retraining and updating with DoD-specific variables improved performance. Domain and temporal validation results were similar to external validation, suggesting that implementing an external model in US Navy primary care clinics may bypass the need for costly internal development and expedite the automation of suicide prevention in these clinics.

Is physical pain causally related to suicidal behavior: An experimental test.

Existing evidence suggests a link between physical pain and suicide, but the nature of this relationship remains unknown. To address this critical gap in knowledge, the present study leveraged a validated virtual reality (VR) suicide paradigm to experimentally examine the causal effects of physical pain on subsequent virtual suicidal behaviors. Based on previous findings, we hypothesized that physical pain would causally drive virtual suicidal behavior only if suicide was conceptualized as having desirable anticipated consequences (e.g., a means of escaping from current pain; an opportunity to avoid future pain). We tested this by randomizing 326 participants across four different conditions: a physical pain condition, an anticipated escape condition, an anticipated avoidance condition, and a control condition. As predicted, physical pain alone did not result in statistically significant increases in VR suicide rates; however, the anticipation that virtual suicidal behavior would result in the avoidance of future physical pain had a large causal effect on VR suicide rates (B = 1.61, p < .001, IRR = 5.01). We failed to find evidence that anticipating that VR suicide would provide an escape from currently experienced physical pain increases the likelihood of VR suicide. Our findings add to a growing body of evidence suggesting that the anticipated consequences of suicide (e.g., avoidance of future physical pain) may serve as primary causes of suicidal behavior.

Efficacy of psychotropic medications on suicide and self-injury: a meta-analysis of randomized controlled trials.

Using psychotropic medications to treat and prevent self-injurious thoughts and behaviors (SITBs) has become increasingly popular, but conclusive evidence supporting the efficacy this approach remains elusive. To empirically examine whether psychotropic medications are efficacious treatments for SITBs, the present meta-analysis comprehensively summarizes all published randomized controlled trials (RCTs) that have reported the causal effects of psychotropic medications on suicide and self-injury. A total of 251 papers from 718 unique RCTs were included. A frequentist pairwise approach was adopted for meta-analyses. Potential effect modifiers were examined via met regressions and potential biases were evaluated through sensitivity analyses. On average, medications yielded an 8% reduction in SITB frequency and a reduction of 0.2 standard deviations in symptoms and severity. Findings were largely consistent across potential effect modifiers, and significant evidence of publication bias was not detected. Only one medication class (i.e., antipsychotics) and two specific medications (i.e., citalopram, ketamine) produced larger-than-average treatment effects. Psychostimulants and typical antipsychotics may produce iatrogenic effects. Less than 4% of included studies required individuals to exhibit SITBs, and nearly half of analyzed effects were drawn from studies that excluded individuals on the basis of SITB risk. Taken together, findings suggest that psychotropic medications produce small treatment effects on SITBs; however, these findings should be considered in light of the methodological constraints of the existing literature, including the lack of studies intentionally including individuals with SITBs. It is critical for future RCTs to prioritize including individuals with existing SITBs to further clarify treatment effects in self-injurious and suicidal populations. Additional research is needed to better understand the treatment mechanisms of psychotropic medications and identify the causal processes underlying SITBs.

Efficacy of interventions for suicide and self-injury in children and adolescents: a meta-analysis.

Despite increased numbers of children and adolescents seeking and receiving mental health treatment, rates of self-injurious thoughts and behaviors (SITBs) in youth are rising. In the hopes of aiding ongoing efforts to alleviate the burden of SITBs in this vulnerable population, the present study summarizes current knowledge on the efficacy of SITB interventions in children and adolescents. We conducted a meta-analysis of randomized controlled trials (RCTs) assessing treatment effects on SITBs in child and adolescent populations. A total of 112 articles comprising 558 effect sizes were included in analyses. Nearly all interventions produced nonsignificant reductions in SITBs. For binary SITB outcomes, a nonsignificant treatment effect was detected, with an RR of 1.06 (95% CIs [0.99, 1.14]). For continuous SITB outcomes, analyses also yielded a nonsignificant treatment effect (g = - 0.04 [- 0.12, 0.05]). These patterns were largely consistent across SITB outcomes, regardless of intervention type, treatment components, sample and study characteristics, and publication year. Our findings highlight opportunities for improving SITB intervention development and implementation in child and adolescent populations. The most efficacious interventions are likely to directly target the causes of SITBs; therefore, future research is needed to identify the causal processes underlying the onset and maintenance of SITBs in youth.

Suicide plan prevalence, recurrence, and longitudinal association with nonfatal suicide attempt.

BACKGROUND: Each year, millions of people develop suicide plans. These plans are assumed to indicate imminent suicide risk, yet this has rarely been tested. The present study seeks to address two questions: (1) how prevalent are specific thoughts of suicide plans among individuals with a history of suicidal thoughts and behaviors and (2) do suicide plans confer risk of future suicide attempts in the short term? METHODS: Secondary data analysis was performed on a longitudinal dataset (N = 1021). Prevalence and frequencies of suicide planning features (i.e., method, time, place) at baseline and 3, 14, and 28 days post-baseline were calculated. Logistic regressions were conducted to assess whether suicide plans confer risk of suicide attempts across a 28-day follow-up period. RESULTS: Suicide planning more commonly involved thoughts of method than place and/or time. High variability in suicide planning was evident and thoughts of suicide plans frequently recurred. Contrary to assumptions, suicide plans displayed weak associations with nonfatal suicide attempt across the 28-day follow-up period. CONCLUSIONS: Suicide plans appear heterogeneous in nature. They do not appear to play a strong role in predicting nonfatal suicide attempts. Re-evaluation of the central role that suicide plans occupy within clinical risk assessments may be warranted.

Does Body Mass Index Confer Risk for Future Suicidal Thoughts and Behaviors? A Meta-analysis of Longitudinal Studies.

PURPOSE OF REVIEW: Body mass index (BMI) outside of the "normal" range is commonly cited as a predictor of adverse health outcomes and has been identified as a potential risk factor for suicidal thoughts and behaviors (STBs). This meta-analysis provides a descriptive and quantitative summary of the literature evaluating the longitudinal relationship between BMI/weight status and STBs. RECENT FINDINGS: The longitudinal literature examining the relationship between BMI/weight status and STBs is small and methodologically constrained. Within the existing literature, BMI and weight status are generally weak or nonsignificant risk factors for STBs. It is possible that body weight has a complex relationship with physical and mental health, including STBs, which may not be possible to accurately capture with a singular metric such as BMI. BMI and weight status do not appear to robustly predict STBs, at least within the methodological constraints of the existing literature.

Raising doubt about the anticipated consequences of suicidal behavior: Evidence for a new approach from laboratory and real-world experiments.

OBJECTIVE: Converging evidence from basic science and experimental suicide research suggest that the anticipated consequences of suicide may have direct causal effects on suicidal behavior and accordingly represent a promising intervention target. Raising doubt about individuals' desirable anticipated consequences of suicide may be one means of disrupting this target. We tested this possibility across two complementary experimental studies. METHOD: Study 1 tested the effects of raising doubt about desirable anticipated consequences on virtual reality (VR) suicide in the lab, randomizing 413 participants across four conditions. In Study 2, 226 suicidal adults were randomized to an anticipated consequence manipulation or control condition then re-assessed at 2- and 8-weeks post-baseline. RESULTS: In Study 1, anticipating that engaging in VR suicide would guarantee a desirable outcome significantly increased the VR suicide rate; conversely, raising doubt about the desirable anticipated consequences significantly reduced the VR suicide rate. In Study 2, raising doubt about the anticipated consequences of attempting suicide by firearm significantly reduced the perceived lethality of firearms as well as self-predicted likelihood of future suicide attempts, with effects sustained at 2-week follow-up. CONCLUSIONS: Findings suggest that raising doubt about desirable anticipated consequences of suicide merits further research as one potential approach to inhibit suicidal behavior.

Longitudinal studies support the safety and ethics of virtual reality suicide as a research method.

Many have expressed concerns about the safety and ethics of conducting suicide research, especially intense suicide research methods that expose participants to graphic depictions of suicidality. We conducted two studies to evaluate the effects of one such method called virtual reality (VR) suicide. Study 1 tested the effects of VR suicide exposure over the course of one month in participants with (n = 56) and without a history of suicidality (n = 50). Study 2 exposed some participants to VR suicide scenarios (n = 79) and others to control scenarios (n = 80). Participants were invited to complete a follow-up assessment after an average of 2 years. For both studies, the presence of suicidality post exposure was the primary outcome, with closely related constructs (e.g., capability for suicide, agitation) as secondary outcomes. Study 1 found no pre-post increases in suicidality or related variables, but revealed several significant decreases associated with small to medium effect sizes in suicide-related constructs. In Study 2, VR suicide exposure did not cause any significant increases in suicidality or related variables. Together with prior research, these findings suggest that methods involving intense suicide stimuli appear safe and consistent with utilitarian ethics.

Does fearlessness about death mediate the association between NSSI and suicide attempts? A longitudinal study of over 1,000 high-risk individuals.

OBJECTIVE: Nonsuicidal self-injury (NSSI) is often cited as a key risk factor for future suicidal behavior. Capability for suicide has been repeatedly cited as an important mechanism that can account for this association. Despite this, direct tests of this hypothesis have been rare and methodologically constrained. In the present study, we conducted a direct test of this hypothesis while addressing several constraints of prior literature. METHOD: In a large sample of suicidal and self-injuring adults (n = 1,020), we tested whether changes in fearlessness about death (FAD), a core facet of the capability for suicide, accounted for the relationship between NSSI and future suicide attempts at 28-day and 2-year follow-up. FAD was assessed using the gold-standard self-report form (ACSS-FAD), an implicit test of suicide-related affect (affect misattribution paradigm-Suicide), and explicit affective ratings of suicide-relevant images. Mediation with bootstrapping was implemented to test our main hypotheses. RESULTS: As anticipated, lifetime NSSI frequency was significantly associated with suicide attempt frequency at follow-up; however, FAD failed to consistently mediate this association. Results were largely consistent across all three measures of FAD. Post hoc power analyses indicated sufficient power to detect small effects. CONCLUSIONS: Taken together, these results fail to support the hypothesis that capability for suicide explains the link between NSSI and future suicidal behavior. We discuss the implications of our results for research and theory, situating our findings in the context of recent advances in the understanding of suicide risk more broadly. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Do suicidal desire and facets of capability for suicide predict future suicidal behavior? A longitudinal test of the desire-capability hypothesis.

Capability-based models propose that people die by suicide because they want to, and they can. Despite the intuitive appeal of this hypothesis, longitudinal evidence testing its predictive validity has been limited. This study tested the predictive validity of the desire-capability hypothesis. A total of 1,020 self-injuring and/or suicidal adults were recruited worldwide online from suicide, self-injury, and mental health web forums. After baseline assessment, participants completed follow-up assessments at 3, 14, and 28 days after baseline. Participant retention was high (>90%) across all follow-up assessments. Analyses examined the effect of the statistical interaction between suicidal desire and indices of capability for suicide on future nonfatal suicide attempts. Main analyses focused on the fearlessness about death facet of capability for suicide; exploratory analyses examined preparations for suicide. Logistic regression was used to predict suicide attempt status at follow-up; zero-inflated negative binomial models were implemented to predict the frequency of nonfatal suicide attempts at follow-up. Results were consistent across models, finding very little evidence of the desire-capability interaction as a significant predictor of suicide attempt status or frequency at follow-up. We close with a discussion of the limitations of this study as well as the implications of our findings for future suicide science. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Suicide Ideation and Thwarted Interpersonal Needs among Psychiatric Inpatients: A Network Approach.

We aimed to demonstrate the utility of an item-level network analysis approach to suicide risk by testing the interpersonal psychological theory of suicide (IPTS) among 402 psychiatric inpatients. We hypothesized specific thwarted belongingness (TB) or perceived burdensomeness (PB; Interpersonal Needs Questionnaire items) facets would positively relate to passive or active suicide ideation, and these facets would positively relate to each other and form distinct clusters. We also tested TB and PB facets central to the networks as predictors of suicide ideation compared to the full TB and PB subscales. Face-valid items congruent with latent constructs proposed by the IPTS (i.e., feelings of burden on society, feeling that one does not belong) were the only two facets uniquely predictive of passive and active suicide ideation. Facets of TB and PB did not form distinct clusters. Item-level network analysis may have important conceptual, assessment, predictive, and clinical implications for understanding suicide risk.

Interventions for suicide and self-injury: A meta-analysis of randomized controlled trials across nearly 50 years of research.

Self-injurious thoughts and behaviors (SITBs) are major public health concerns impacting a wide range of individuals and communities. Despite major efforts to develop and refine treatments to reduce SITBs, the efficacy of SITB interventions remains unclear. To provide a comprehensive summary of SITB treatment efficacy, we conducted a meta-analysis of published randomized controlled trials (RCTs) that have attempted to reduce SITBs. A total of 591 published articles from 1,125 unique RCTs with 3,458 effect sizes from the past 50 years were included. The random-effects meta-analysis yielded surprising findings: The overall intervention effects were small across all SITB outcomes; despite a near-exponential increase in the number of RCTs across five decades, intervention efficacy has not improved; all SITB interventions produced similarly small effects, and no intervention appeared significantly and consistently stronger than others; the overall small intervention effects were largely maintained at follow-up assessments; efficacy was similar across age groups, though effects were slightly weaker for child/adolescent populations and few studies focused on older adults; and major sample and study characteristics (e.g., control group type, treatment target, sample size, intervention length) did not consistently moderate treatment efficacy. This meta-analysis suggests that fundamental changes are needed to facilitate progress in SITB intervention efficacy. In particular, powerful interventions target the necessary causes of pathology, but little is known about SITB causes (vs. SITB correlates and risk factors). The field would accordingly benefit from the prioritization of research that aims to identify and target common necessary causes of SITBs. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Sleep disturbances as risk factors for suicidal thoughts and behaviours: a meta-analysis of longitudinal studies.

In recent years, there has been a growing interest in understanding the relationship between sleep and suicide. Although sleep disturbances are commonly cited as critical risk factors for suicidal thoughts and behaviours, it is unclear to what degree sleep disturbances confer risk for suicide. The aim of this meta-analysis was to clarify the extent to which sleep disturbances serve as risk factors (i.e., longitudinal correlates) for suicidal thoughts and behaviours. Our analyses included 156 total effects drawn from 42 studies published between 1982 and 2019. We used a random effects model to analyse the overall effects of sleep disturbances on suicidal ideation, attempts, and death. We additionally explored potential moderators of these associations. Our results indicated that sleep disturbances are statistically significant, yet weak, risk factors for suicidal thoughts and behaviours. The strongest associations were found for insomnia, which significantly predicted suicide ideation (OR 2.10 [95% CI 1.83-2.41]), and nightmares, which significantly predicted suicide attempt (OR 1.81 [95% CI 1.12-2.92]). Given the low base rate of suicidal behaviours, our findings raise questions about the practicality of relying on sleep disturbances as warning signs for imminent suicide risk. Future research is necessary to uncover the causal mechanisms underlying the relationship between sleep disturbances and suicide.

The Differences Between Individuals Engaging in Nonsuicidal Self-Injury and Suicide Attempt Are Complex (vs. Complicated or Simple).

BACKGROUND: Why do some people engage in nonsuicidal self-injury (NSSI) while others attempt suicide? One way to advance knowledge about this question is to shed light on the differences between people who engage in NSSI and people who attempt suicide. These groups could differ in three broad ways. First, these two groups may differ in a simple way, such that one or a small set of factors is both necessary and sufficient to accurately distinguish the two groups. Second, they might differ in a complicated way, meaning that a specific set of a large number of factors is both necessary and sufficient to accurately classify them. Third, they might differ in a complex way, with no necessary factor combinations and potentially no sufficient factor combinations. In this scenario, at the group level, complicated algorithms would either be insufficient (i.e., no complicated algorithm produces good accuracy) or unnecessary (i.e., many complicated algorithms produce good accuracy) to distinguish between groups. This study directly tested these three possibilities in a sample of people with a history of NSSI and/or suicide attempt. METHOD: A total of 954 participants who have either engaged in NSSI and/or suicide attempt in their lifetime were recruited from online forums. Participants completed a series of measures on factors commonly associated with NSSI and suicide attempt. To test for simple differences, univariate logistic regressions were conducted. One theoretically informed multiple logistic regression model with suicidal desire, capability for suicide, and their interaction term was considered as well. To examine complicated and complex differences, multiple logistic regression and machine learning analyses were conducted. RESULTS: No simple algorithm (i.e., single factor or small set of factors) accurately distinguished between groups. Complicated algorithms constructed with cross-validation methods produced fair accuracy; complicated algorithms constructed with bootstrap optimism methods produced good accuracy, but multiple different algorithms with this method produced similar results. CONCLUSIONS: Findings were consistent with complex differences between people who engage in NSSI and suicide attempts. Specific complicated algorithms were either insufficient (cross-validation results) or unnecessary (bootstrap optimism results) to distinguish between these groups with high accuracy.

The differences between suicide ideators and suicide attempters: Simple, complicated, or complex?

OBJECTIVE: Suicide ideators and suicide attempters might differ in 3 possible ways. First, they might differ in a simple way such that one or a small set of factors are both necessary and sufficient to distinguish between the 2 groups. Second, ideators and attempters might differ in a complicated way such that a specific combination of a large set of factors is necessary and sufficient for the distinction. Third, complex differences might exist: many possible combinations of a large set of factors may be sufficient to distinguish the 2 groups, but no combination may be necessary. This study empirically examined these possibilities. METHOD: Across 5 samples (total N = 3,869), univariate logistic regressions were conducted to test for simple differences. To test for complicated and complex differences, machine learning (ML) methods were used to identify the optimized algorithm with all variables. Subsequently, the same methods were repeated after removing the top 5 most important or discriminative variables, and a randomly selected 10% subset of variables. Multiple logistic regressions were conducted with all variables. RESULTS: Results were consistent across samples. Univariate logistic regressions on average yielded chance-level accuracy. ML algorithms with all variables showed good accuracy; substantial deviation from the optimized algorithms through the removal of variables did not result in significantly poorer performance. Multiple logistic regressions produced poor to fair accuracy. CONCLUSIONS: Differences between suicide ideators and attempters are complex. Findings suggest that their differences may be better understood on a psychological primitive level than a biopsychosocial factor level. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

User-Centered Design of a Machine Learning Intervention for Suicide Risk Prediction in a Military Setting.

Primary care represents a major opportunity for suicide prevention in the military. Significant advances have been made in using electronic health record data to predict suicide attempts in patient populations. With a user-centered design approach, we are developing an intervention that uses predictive analytics to inform care teams about their patients' risk of suicide attempt. We present our experience working with clinicians and staff in a military primary care setting to create preliminary designs and a context-specific usability testing plan for the deployment of the suicide risk indicator.

Significant shared heritability underlies suicide attempt and clinically predicted probability of attempting suicide.

Suicide accounts for nearly 800,000 deaths per year worldwide with rates of both deaths and attempts rising. Family studies have estimated substantial heritability of suicidal behavior; however, collecting the sample sizes necessary for successful genetic studies has remained a challenge. We utilized two different approaches in independent datasets to characterize the contribution of common genetic variation to suicide attempt. The first is a patient reported suicide attempt phenotype asked as part of an online mental health survey taken by a subset of participants (n = 157,366) in the UK Biobank. After quality control, we leveraged a genotyped set of unrelated, white British ancestry participants including 2433 cases and 334,766 controls that included those that did not participate in the survey or were not explicitly asked about attempting suicide. The second leveraged electronic health record (EHR) data from the Vanderbilt University Medical Center (VUMC, 2.8 million patients, 3250 cases) and machine learning to derive probabilities of attempting suicide in 24,546 genotyped patients. We identified significant and comparable heritability estimates of suicide attempt from both the patient reported phenotype in the UK Biobank (h2SNP = 0.035, p = 7.12 × 10-4) and the clinically predicted phenotype from VUMC (h2SNP = 0.046, p = 1.51 × 10-2). A significant genetic overlap was demonstrated between the two measures of suicide attempt in these independent samples through polygenic risk score analysis (t = 4.02, p = 5.75 × 10-5) and genetic correlation (rg = 1.073, SE = 0.36, p = 0.003). Finally, we show significant but incomplete genetic correlation of suicide attempt with insomnia (rg = 0.34-0.81) as well as several psychiatric disorders (rg = 0.26-0.79). This work demonstrates the contribution of common genetic variation to suicide attempt. It points to a genetic underpinning to clinically predicted risk of attempting suicide that is similar to the genetic profile from a patient reported outcome. Lastly, it presents an approach for using EHR data and clinical prediction to generate quantitative measures from binary phenotypes that can improve power for genetic studies.

Model complexity improves the prediction of nonsuicidal self-injury.

OBJECTIVE: Efforts to predict nonsuicidal self-injury (NSSI; intentional self-injury enacted without suicidal intent) to date have resulted in near-chance accuracy. Incongruence between theoretical understanding of NSSI and the traditional statistical methods to predict these behaviors may explain this poor prediction. Whereas theoretical models of NSSI assume that the decision to engage in NSSI is relatively complex, statistical models used in NSSI prediction tend to involve simple models with only a few theoretically informed variables. The present study tested whether more complex statistical models would improve NSSI prediction. METHOD: Within a sample of 1,021 high-risk self-injurious and/or suicidal individuals, we examined the accuracy of three different model types, of increasing complexity, in predicting NSSI across 3, 14, and 28 days. Univariate logistic regressions of each predictor and multiple logistic regression with all predictors were conducted for each timepoint and compared with machine learning algorithms derived from all predictors. RESULTS: Results demonstrated that model complexity was associated with predictive accuracy. Multiple logistic regression models (AUCs 0.70-0.72) outperformed univariate logistic models (average AUCs 0.56). Machine learning models that produced algorithms modeling complex associations across variables produced the strongest NSSI prediction across all time points (AUCs 0.87-0.90). These models outperformed all multiple logistic regression models, including those involving identical study variables. Machine learning algorithm performance remained strong even after the most important factor across algorithms was removed. CONCLUSIONS: Results parallel recent findings in suicide research and highlight the complexity that underlies NSSI. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Machine learning in suicide science: Applications and ethics.

For decades, our ability to predict suicide has remained at near-chance levels. Machine learning has recently emerged as a promising tool for advancing suicide science, particularly in the domain of suicide prediction. The present review provides an introduction to machine learning and its potential application to open questions in suicide research. Although only a few studies have implemented machine learning for suicide prediction, results to date indicate considerable improvement in accuracy and positive predictive value. Potential barriers to algorithm integration into clinical practice are discussed, as well as attendant ethical issues. Overall, machine learning approaches hold promise for accurate, scalable, and effective suicide risk detection; however, many critical questions and issues remain unexplored.