Prediction Models for Suicide Attempts and Deaths: A Systematic Review and Simulation

Relations Between Suicide Risk and Patient Safety Attitudes Among the Nursing Team in a Brazilian Context

OBJECTIVES

To analyze the aspects of suicide risks and their relation to patient safety attitudes among the nursing staff of a large public university hospital in Brazil.

METHODS

A cross-sectional and descriptive study with 226 nursing workers from a large public university hospital in Brazil. Socioprofessional information, health conditions, and family history related to suicide risk were collected through the Mini International Neuropsychiatric Interview Plus and the Safety Attitudes Questionnaire. A 95% CI was considered, and statistical tests such as the Student t test, χ 2 , analysis of variance, and multiple linear regression were used.

RESULTS

Most of the safety attitudes were below the average score considered positive (mean Safety Attitudes Questionnaire value >0.75), and 41 (18.1%) workers were considered to have any degree of suicide risk. Participants with parents or siblings who had attempted suicide were 3.44 times more likely to have moderate or high suicide risk. Negative safety attitudes were associated with health conditions and family history, considered suicide risk factors. Participants with moderate or high suicide risk were 2.83 times more likely to have worse patient safety attitudes concerning job satisfaction.

CONCLUSIONS

This study reveals significant associations between patient safety attitudes and the mental health of nursing workers, expanding the view of worker health management actions and, consequently, patient safety culture.

Associations between NIH Toolbox Emotion Battery measures and previous suicide attempt in bipolar I disorder

Suicide attempts are more prevalent in people with bipolar I disorder (BD-I) than in the general population. Most prior studies of suicide in BD-I have focused on separate emotion-related assays or clinician-administered scales, whereas a single, brief, and multidimensional battery of self-report measures has not yet been explored. Here, we utilized the NIH Toolbox Emotion Battery (NIHTB-EB) to assess various emotional measures, determine which were cross-sectionally associated with prior suicide attempt in BD-I, evaluate whether the NIHTB-EB could be used to identify past suicide attempt in BD-I with machine learning, and compare model performance versus using clinical mood scales. The study included 39 participants with BD-I and history of suicide attempt, 48 with BD-I without history of suicide attempt, and 58 controls. We found that 9 of the 17 measures were associated with past suicide attempt in BD-I. The initial random forest model indicated that the most important distinguishing variables were perceived stress, emotional support, anger-hostility, anger-physical aggression, perceived rejection, loneliness, and self-efficacy. Overall, the models utilizing NIHTB-EB measures performed better (69.0 % to 70.1 % accuracy) than the model containing clinical mood scale information without the NIHTB-EB measures (57.5 % accuracy). These findings suggest the NIHTB-EB could be a useful and easy-to-deploy tool in understanding the role of emotion-related measures in suicide in BD-I. Furthermore, these results highlight specific emotional subdomains that could be promising targets for longitudinal studies or interventions aimed at reducing suicide in BD-I.

The Suicide Status Form-4 (SSF-IV) as a potentially therapeutic suicide risk assessment tool

BACKGROUND

Empirically supported suicide risk assessment and conceptualization is a central aim of the Zero Suicide model. The Suicide Status Form (SSF) is the essential document and scaffolding of the Collaborative Assessment and Management of Suicidality-Brief Intervention (CAMS-BI) and is hypothesized as an example of a psychological assessment as therapeutic intervention (PATI). However, this hypothesis has never been directly tested.

METHODS

N = 57 patients deemed at risk for outpatient suicidal behavior and treated as part of an inpatient psychiatric consultation and liaison service were recruited to participate in CAMS-BI at a Level 1 trauma center in the southeastern United States. During the CAMS-BI process, patients were asked to rate their subjective units of distress (SUDS) at five time points throughout the intervention (k = 285).

RESULTS

The omnibus random intercept multilevel model revealed a significant difference in pre- to post-session ratings of SUDS across patients. Post hoc pairwise comparisons revealed no significant differences between SSF sections (e.g., Section A, Section B, and Section C) and relative reductions in SUDS; however, there was an observable trend toward a favorable effect of Section A of the SSF.

CONCLUSIONS

The SSF may represent an example of PATI pending replication and extension of the current results.

Multi-modal risk factors differentiate suicide attempters from ideators in military veterans with major depressive disorder

BACKGROUND

The suicide rate for United States military veterans is 1.5× higher than that of non-veterans. To meaningfully advance suicide prevention efforts, research is needed to delineate factors that differentiate veterans with suicide attempt/s, particularly in high-risk groups, e.g., major depressive disorder (MDD), from those with suicidal ideation (no history of attempt/s). The current study aimed to identify clinical, neurocognitive, and neuroimaging variables that differentiate suicide-severity groups in veterans with MDD.

METHODS

Sixty-eight veterans with a DSM-5 diagnosis of MDD, including those with no ideation or suicide attempt (N = 21; MDD-SI/SA), ideation-only (N = 17; MDD + SI), and one-or-more suicide attempts (N = 30; MDD + SA; aborted, interrupted, actual attempts), participated in this study. Participants underwent a structured diagnostic interview, neurocognitive assessment, and 3 T-structural/diffusion tensor magnetic-resonance-imaging (MRI). Multinomial logistic regression models were conducted to identify variables that differentiated groups with respect to the severity of suicidal behavior.

RESULTS

Relative to veterans with MDD-SI/SA, those with MDD + SA had significantly higher left cingulum fractional anisotropy, decreased attentional control on emotional-Stroop, and faster response time with intact accuracy on Go/No-Go. Relative to MDD + SI, MDD + SA had higher left cingulum fractional anisotropy and faster response time with intact accuracy on Go/No-Go.

LIMITATIONS

Findings are based on retrospective, cross-sectional data and cannot identify causal relationships. Also, a healthy control group was not included given the study's focus on differentiating suicide profiles in MDD.

CONCLUSIONS

This study suggests that MRI and neurocognition differentiate veterans with MDD along the suicide-risk spectrum and could inform suicide-risk stratification and prevention efforts in veterans and other vulnerable populations.

Risk Model-Guided Clinical Decision Support for Suicide Screening: A Randomized Clinical Trial

Importance

Suicide prevention requires risk identification, intervention, and follow-up. Traditional risk identification relies on patient self-reporting, support network reporting, or face-to-face screening. Statistical risk models have been studied and some have been deployed to augment clinical judgment. Few have been tested in clinical practice via clinical decision support (CDS). Barriers to effective CDS include potential alert burden for a stigmatized clinical problem and lack of data on how best to integrate scalable risk models into clinical workflows.

Objective

To evaluate the effectiveness of risk model-driven CDS on suicide risk assessment.

Design, Setting, and Participants

This comparative effectiveness randomized clinical trial was performed from August 17, 2022, to February 16, 2023, in the Department of Neurology across the divisions of Neuro-Movement Disorders, Neuromuscular Disorders, and Behavioral and Cognitive Neurology at Vanderbilt University Medical Center, an academic medical center in the US Mid-South. Patients scheduled for routine care in those settings were randomized at visit check-in. Follow-up was completed March 16, 2023, and data were analyzed from April 11 to July 24, 2023. Analyses were based on intention to treat.

Interventions

Interruptive vs noninterruptive CDS to prompt further suicide risk assessment using a real-time, validated statistical suicide attempt risk model. In the interruptive CDS, an alert window via on-screen pop-up and a patient panel icon were visible simultaneously. Dismissing the alert hid it with no effect on the patient panel icon. The noninterruptive CDS showed the patient panel icon without the pop-up alert. When present, the noninterruptive CDS displayed "elevated suicide risk score" in the patient summarization panel. Hovering over this icon resulted in a pop-up identical to the interruptive CDS.

Main Outcomes and Measures

The main outcome was the decision to assess risk in person. Secondary outcomes included rates of suicidal ideation and attempts in both treatment arms and baseline rates of documented screening during the prior year. Manual medical record review of every trial encounter was used to determine whether suicide risk assessment was subsequently documented.

Results

A total of 561 patients with 596 encounters were randomized to interruptive or noninterruptive CDS in a 1:1 ratio (mean [SD] age, 59.3 [16.5] years; 292 [52%] women). Adjusting for clinician cluster effects, interruptive CDS led to significantly higher numbers of decisions to screen (121 of 289 encounters [42%]) compared with noninterruptive CDS (12 of 307 encounters [4%]) (odds ratio, 17.70; 95% CI, 6.42-48.79; P < .001) and compared with the baseline rate the prior year (64 of 832 encounters [8%]). No documented episodes of suicidal ideation or attempts occurred in either arm.

Conclusions and Relevance

In this randomized clinical trial of interruptive and noninterruptive CDS to prompt face-to-face suicide risk assessment, interruptive CDS led to higher numbers of decisions to screen with documented suicide risk assessments. Well-powered large-scale trials randomizing this type of CDS compared with standard of care are indicated to measure effectiveness in reducing suicidal self-harm.

Trial Registration

ClinicalTrials.gov Identifier: NCT05312437.

Are we Using the Right Evidence to Inform Suicide Prevention in Low- and Middle-Income Countries? An Umbrella Review

OBJECTIVE

Suicide disproportionately affects low- and middle-income countries and evidence regarding prevention approaches developed in high income countries may not be applicable in these settings. We conducted an umbrella review to assess whether the conclusions of suicide prevention systematic reviews accurately reflect the studies contained within those reviews in terms of setting generalizability.

METHODS

We conducted database searches in PubMed/Medline, Embase, PsycInfo, PsychExtra, OVID global health, and LILACS/BECS. We included systematic reviews with the outcome of suicide, including bereavement studies where suicide death was also the exposure.

RESULTS

Out of the 147 reviews assessed, we found that over 80% of systematic reviews on suicide deaths do not provide an accurate summary of review findings with relation to geographic relevance and ultimately generalizability.

CONCLUSION

Systematic reviews are often the resource used by practitioners and policymakers to guide services. Misleading reviews can detrimentally impact suicide prevention efforts in LMICs. We call for systematic reviewers to be responsible when generalizing the findings of their reviews particularly in the abstracts.

Predicting suicidal behavior outcomes: an analysis of key factors and machine learning models

BACKGROUND

Suicidal behaviors, which may lead to death (suicide) or survival (suicide attempt), are influenced by various factors. Identifying the specific risk factors for suicidal behavior mortality is critical for improving prevention strategies and clinical interventions. Predicting the outcomes of suicidal behaviors can help identify individuals at higher risk of death, enabling timely and targeted interventions. This study aimed to determine the critical risk factors associated with suicidal behavior mortality and identify an effective classification model for predicting suicidal behavior outcomes.

MATERIALS AND METHODS

This study utilized data recorded in the suicidal behavior registry system of hospitals in Ilam Province. In the first phase, duplicate records were removed, and the data was numerically encoded via Python version 3.11; then, the data was analyzed using chi-square and Fisher's exact tests in SPSS version 22 software to identify the factors influencing suicidal behavior mortality. In the second phase, missing data were removed, and the dataset was standardized. Five binary classification algorithms were utilized, including Random Forest, Logistic Regression, and Decision Trees, with hyperparameters optimized using the area under the receiver operating characteristic curve (AUC) and F1 score metrics. These models were compared based on accuracy, recall, precision, F1 score, and AUC.

RESULTS

Among 3833 cases of suicidal behavior in various hospitals in Ilam Province, the results indicated that the method of suicidal behavior (P < 0.001), reason for suicidal behavior (P < 0.001), age group (P < 0.001), education level (P < 0.001), marital status (P = 0.004), and employment status (P = 0.042) were significantly associated with suicide. Variables such as the season of suicidal behavior, gender, father's education, and mother's education were not significantly related to suicidal behavior mortality. Furthermore, the random forest model demonstrated the highest area under the ROC curve (0.79) and the highest classification accuracy and F1 score on both the training data (0.85 and 0.2, respectively) and test data (0.86 and 0.31, respectively) for predicting suicidal behaviors outcomes among the models tested.

CONCLUSION

This study identified key factors such as older age, lower education, divorce or widowhood, employment, physical methods, and socioeconomic issues as significant predictors of suicidal behavior outcomes. A combination of statistical models for feature selection and machine learning algorithms for prediction was used, with Random Forest showing the best performance. This approach highlights the potential of integrating statistical methods with machine learning to improve suicide risk prediction and intervention strategies.

Prospectively predicting 6-month risk for non-suicidal self-injury among adolescents after psychiatric hospitalization based on a predictive model

Background

It is challenging to predict the occurrence of non-suicidal self-injury (NSSI) among adolescents over short periods. Moreover, the predictive value of indices for NSSI remains elusive. Thus, this study aimed to identify predictors of NSSI within 6 months among adolescents after psychiatric hospitalization by establishing a risk assessment model.

Methods

A total of 632 high-risk participants were included in this study. The distribution characteristics of adolescent NSSI were initially assessed through a cross-sectional survey, following which risk factors were identified using logistic regression analysis. The risk score method was then used to construct a 6-month risk assessment model for NSSI. Lastly, the predictive effect of the model was evaluated by indicators such as the area under the receiver operating characteristic (ROC) curve and the positive predictive value.

Results

After 6 months, 412 cases of NSSI were identified. According to the logistic regression model, the frequency of relapses, medication status, and NSSI history were identified as influencing factors. Higher scores on the Impulsive Behavior Scale and Pittsburgh Sleep Quality Index were associated with a higher risk of NSSI. Conversely, higher scores on the Pain and Belief Perception Scale were correlated with a lower risk of NSSI. Moreover, the area under the ROC curve for the predictive model was 0.9989, with a 95% confidence interval of (0.9979, 0.9999), highlighting its high predictive ability and accuracy. The predictive model was validated using 78 patients, yielding an area under the ROC curve of 0.9703 and a 95% confidence interval of (0.9167, 0.9999), demonstrating outstanding predictability.

Conclusion

These results collectively showed that the predictive model could accurately predict adolescent NSSI. Thus, the model's primary variables may be applied to predict the risk of NSSI in the clinical setting.

Potential Harms of Responding to Youth Suicide Risk in Schools

The potential harms related to interventions for adults with suicide-related risk, particularly hospitalization, have been well documented. Much less work has focused on the potential harms related to interventions with youth struggling with suicidal thoughts and behaviors. Young people are most likely to receive mental health services in schools, which are recognized as meaningful sites for effective suicide prevention work. However, no overviews have conceptualized the potential harms to youth when schools engage in ineffective suicide prevention efforts. In this article, we discuss three prominent overlapping areas of potential harms: (1) privacy-related, (2) relationship-related, and (3) mental health-related. We then discuss key factors thought to influence the development and maintenance of these potential harms. We conclude by noting ways in which school-based mental health providers may attempt to reduce unintentional harms in this area, with an overarching goal of helping support school mental health providers and the youth they serve.

Developing a suicide risk model for use in the Indian Health Service

We developed and evaluated an electronic health record (EHR)-based model for suicide risk specific to an American Indian patient population. Using EHR data for all patients over 18 with a visit between 1/1/2017 and 10/2/2021, we developed a model for the risk of a suicide attempt or death in the 90 days following a visit. Features included demographics, medications, diagnoses, and scores from relevant screening tools. We compared the predictive performance of logistic regression and random forest models against existing suicide screening, which was augmented to include the history of previous attempts or ideation. During the study, 16,835 patients had 331,588 visits, with 490 attempts and 37 deaths by suicide. The logistic regression and random forest models (area under the ROC (AUROC) 0.83 [0.80-0.86]; both models) performed better than enhanced screening (AUROC 0.64 [0.61-0.67]). These results suggest that an EHR-based suicide risk model can add value to existing practices at Indian Health Service clinics.

Clinical characteristics and suicidal ideation as predictors of suicide: prospective study of 1000 referrals to general adult psychiatry

AIMS AND METHOD

Questions often follow the suicide of someone who presented to general adult psychiatry (GAP) when expressing suicidal thoughts: 'Why were they not admitted, or managed differently, when they said they were suicidal?' Answering these questions requires knowledge of the prevalence of suicidal ideation in patients presenting to GAP. Therefore, we determined the general clinical characteristics, including suicidal ideation, of a large sample of patients presenting to a GAP emergency assessment service or referred as non-emergencies to a GAP service.

RESULTS

Suicidal ideation was very common, being present in 76.4% of emergency presentations and 33.4% of non-emergency referrals. It was very weakly associated with suicide, varied between different diagnostic categories, and previous assessment by GAP did not appear to affect it. The suicide rate during the contingent episode of care was estimated as 66 per 100 000 episodes.

CLINICAL IMPLICATIONS

This, and other evidence, shows that suicide cannot be predicted with an accuracy that is useful for clinical decision-making. This is not widely appreciated but has serious consequences for patients and healthcare resources.

Longitudinal panel networks of risk and protective factors for early adolescent suicidality in the ABCD sample

Rates of youth suicidal thoughts and behaviors (STBs) are rising, and younger age at onset increases vulnerability to negative outcomes. However, few studies have investigated STBs in early adolescence (ages 10-13), and accurate prediction of youth STBs remains poor. Network analyses that can examine pairwise associations between many theoretically relevant variables may identify complex pathways of risk for early adolescent STBs. The present study applied longitudinal network analysis to examine interrelations between STBs and several previously identified risk and protective factors. Data came from 9,854 youth in the Adolescent Brain Cognitive Development Study cohort (Mage = 9.90 ± .62 years, 63% white, 53% female at baseline). Youth and their caregivers completed an annual measurement battery between ages 9-10 through 11-12 years. Panel Graphical Vector Autoregressive models evaluated associations between STBs and several mental health symptoms, socioenvironmental factors, life stressors, and substance use. In the contemporaneous and between-subjects networks, direct associations were observed between STBs and internalizing symptoms, substance use, family conflict, lower parental monitoring, and lower school protective factors. Potential indirect pathways of risk for STBs were also observed. Age-specific interventions may benefit from prioritizing internalizing symptoms and early substance use, as well as promoting positive school and family support.

First impressions count: Therapists' impression on patients' motivation and helping alliance predicts psychotherapy dropout

OBJECTIVE

With meta-analytically estimated rates of about 25%, dropout in psychotherapies is a major concern for individuals, clinicians, and the healthcare system at large. To be able to counteract dropout in psychotherapy, accurate insights about its predictors are needed.

METHOD

We compared logistic regression models with two machine learning algorithms (elastic net regressions and gradient boosting machines) in the prediction of therapy dropout in two large inpatient samples (N = 1,691 and N = 12,473) using baseline and initial process variables reported by patients and therapists.

RESULTS

Predictive accuracies of the two machine learning algorithms were similar and higher than for logistic regressions: Therapy dropout could be predicted with an AUC of .73 and .83 for Sample 1 and 2, respectively. The initial evaluation of patients' motivation and the therapeutic alliance rated by the respective therapist were the most important predictors of dropout.

CONCLUSIONS

Therapy dropout in naturalistic inpatient settings can be predicted to a considerable degree by using baseline indicators and therapists' first impressions. Feature selection via regularization leads to higher predictive performances whereas non-linear or interaction effects are dispensable. The most promising point of intervention to reduce therapy dropouts seems to be patients' motivation and the therapeutic alliance.

Association of veteran suicide risk with state-level firearm ownership rates and firearm laws in the USA

BACKGROUND

Veterans have higher suicide rates than matched non-veterans, with firearm suicides being especially prevalent among veterans. We examined whether state firearm laws and state firearm ownership rates are important risk factors for suicide among veterans.

METHODS

US veteran's and demographically matched non-veteran's suicide rates, 2002-2019, are modelled at the state level as a function of veteran status, lethal means, state firearm law restrictiveness, household firearm ownership rates and other covariates.

RESULTS

Marginal effects on expected suicide rates per 100 000 population were contrasted by setting household firearm ownership to its 75th versus 25th percentile values of 52.3% and 35.3%. Ownership was positively associated with suicide rates for both veterans (4.35; 95% credible interval (CrI): 1.90, 7.14) and matched non-veterans (3.31; 95% CrI: 1.11, 5.77). This association was due to ownership's strong positive association with firearms suicide, despite a weak negative association with non-firearm suicide. An IQR difference in firearm laws corresponding to three additional restrictive laws was negatively associated with suicide rates for both veterans (-2.49; 95% CrI: -4.64 to -0.21) and matched non-veterans (-3.19; 95% CrI: -5.22 to -1.16). Again, these differences were primarily due to associations with firearm suicide rates. Few differences between veterans and matched non-veterans were found in the associations of state firearm characteristics with suicide rates.

DISCUSSION

Veterans' and matched non-veterans' suicide risk, and specifically their firearm suicide risk, was strongly associated with state firearm characteristics.

CONCLUSIONS

These results suggest that changes to state firearm policies might be an effective primary prevention strategy for reducing suicide rates among veterans and non-veterans.

Predicting Suicides Among US Army Soldiers After Leaving Active Service

Importance

The suicide rate of military servicemembers increases sharply after returning to civilian life. Identifying high-risk servicemembers before they leave service could help target preventive interventions.

Objective

To develop a model based on administrative data for regular US Army soldiers that can predict suicides 1 to 120 months after leaving active service.

Design, Setting, and Participants

In this prognostic study, a consolidated administrative database was created for all regular US Army soldiers who left service from 2010 through 2019. Machine learning models were trained to predict suicides over the next 1 to 120 months in a random 70% training sample. Validation was implemented in the remaining 30%. Data were analyzed from March 2023 through March 2024.

Main outcome and measures

The outcome was suicide in the National Death Index. Predictors came from administrative records available before leaving service on sociodemographics, Army career characteristics, psychopathologic risk factors, indicators of physical health, social networks and supports, and stressors.

Results

Of the 800 579 soldiers in the cohort (84.9% male; median [IQR] age at discharge, 26 [23-33] years), 2084 suicides had occurred as of December 31, 2019 (51.6 per 100 000 person-years). A lasso model assuming consistent slopes over time discriminated as well over all but the shortest risk horizons as more complex stacked generalization ensemble machine learning models. Test sample area under the receiver operating characteristic curve ranged from 0.87 (SE = 0.06) for suicides in the first month after leaving service to 0.72 (SE = 0.003) for suicides over 120 months. The 10% of soldiers with highest predicted risk accounted for between 30.7% (SE = 1.8) and 46.6% (SE = 6.6) of all suicides across horizons. Calibration was for the most part better for the lasso model than the super learner model (both estimated over 120-month horizons.) Net benefit of a model-informed prevention strategy was positive compared with intervene-with-all or intervene-with-none strategies over a range of plausible intervention thresholds. Sociodemographics, Army career characteristics, and psychopathologic risk factors were the most important classes of predictors.

Conclusions and relevance

These results demonstrated that a model based on administrative variables available at the time of leaving active Army service can predict suicides with meaningful accuracy over the subsequent decade. However, final determination of cost-effectiveness would require information beyond the scope of this report about intervention content, costs, and effects over relevant horizons in relation to the monetary value placed on preventing suicides.

Late-life suicide: machine learning predictors from a large European longitudinal cohort

Background

People in late adulthood die by suicide at the highest rate worldwide. However, there are still no tools to help predict the risk of death from suicide in old age. Here, we leveraged the Survey of Health, Ageing, and Retirement in Europe (SHARE) prospective dataset to train and test a machine learning model to identify predictors for suicide in late life.

Methods

Of more than 16,000 deaths recorded, 74 were suicides. We matched 73 individuals who died by suicide with people who died by accident, according to sex (28.8% female in the total sample), age at death (67 ± 16.4 years), suicidal ideation (measured with the EURO-D scale), and the number of chronic illnesses. A random forest algorithm was trained on demographic data, physical health, depression, and cognitive functioning to extract essential variables for predicting death from suicide and then tested on the test set.

Results

The random forest algorithm had an accuracy of 79% (95% CI 0.60-0.92, p = 0.002), a sensitivity of.80, and a specificity of.78. Among the variables contributing to the model performance, the three most important factors were how long the participant was ill before death, the frequency of contact with the next of kin and the number of offspring still alive.

Conclusions

Prospective clinical and social information can predict death from suicide with good accuracy in late adulthood. Most of the variables that surfaced as risk factors can be attributed to the construct of social connectedness, which has been shown to play a decisive role in suicide in late life.

Detecting suicide risk among U.S. servicemembers and veterans: a deep learning approach using social media data

BACKGROUND

Military Servicemembers and Veterans are at elevated risk for suicide, but rarely self-identify to their leaders or clinicians regarding their experience of suicidal thoughts. We developed an algorithm to identify posts containing suicide-related content on a military-specific social media platform.

METHODS

Publicly-shared social media posts (n = 8449) from a military-specific social media platform were reviewed and labeled by our team for the presence/absence of suicidal thoughts and behaviors and used to train several machine learning models to identify such posts.

RESULTS

The best performing model was a deep learning (RoBERTa) model that incorporated post text and metadata and detected the presence of suicidal posts with relatively high sensitivity (0.85), specificity (0.96), precision (0.64), F1 score (0.73), and an area under the precision-recall curve of 0.84. Compared to non-suicidal posts, suicidal posts were more likely to contain explicit mentions of suicide, descriptions of risk factors (e.g. depression, PTSD) and help-seeking, and first-person singular pronouns.

CONCLUSIONS

Our results demonstrate the feasibility and potential promise of using social media posts to identify at-risk Servicemembers and Veterans. Future work will use this approach to deliver targeted interventions to social media users at risk for suicide.

ADHD and cognitive disengagement syndrome symptoms related to self-injurious thoughts and behaviors in early adolescents

The current study examined attention-deficit/hyperactivity disorder (ADHD) dimensions and cognitive disengagement syndrome (CDS) symptoms in relation to self-injurious thoughts and behaviors (SITBs) in an early adolescent sample. Participants were 341 adolescents ages 10-12 years (52.2% female; 37.8% people of color) recruited from the community. Caregivers reported on CDS and ADHD symptoms. Adolescents completed a rating scale and were administered an interview assessing SITBs. We estimated associations using logistic regression in a stepped fashion: (1) no adjustment, (2) adjustment for sex, race, family income, and psychotropic medication use, and (3) further adjustment for depressive symptoms. In this early adolescent community sample, 22.9% reported a history of suicidal ideation, 8.2% reported a history of a suicide plan, 6.2% reported a history of non-suicidal self-injury (NSSI), and 16.4% met a clinical cutoff for current suicide risk. Across most analyses using rating scale or interview methods, higher mean CDS scores were related to endorsement of suicidal ideation and planning. ADHD inattentive (IN) and hyperactive-impulsive (HI) symptoms were associated with endorsement of NSSI, and ADHD-IN symptoms were associated with thoughts of suicide and/or plan measured via questionnaire, though effects were less robust and not significant, potentially due to low base rates impacting statistical power. This study adds to a growing body of research highlighting the importance of screening for CDS symptoms among individuals with and without ADHD. More research, especially longitudinal work, is needed that examines possible differential pathways to SITBs by ADHD and CDS symptoms to advance SITB prevention, early detection, and intervention.

Systematic Review: The Measurement Properties of the Suicidal Ideation Questionnaire and Suicidal Ideation Questionnaire-Jr

OBJECTIVE

The Suicidal Ideation Questionnaire (SIQ) and the Suicidal Ideation Questionnaire-Junior (SIQ-Jr) were designed to capture suicidal ideation in adolescents and are often used in clinical trials. Our aim was to identify and appraise the published literature with respect to the validity, reliability, responsiveness, and interpretability of the SIQ and SIQ-Jr.

METHOD

We conducted a systematic review following COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN) guidelines to identify, appraise, and synthesize published literature on measurement properties and interpretability of the SIQ and SIQ-Jr. We searched MEDLINE, Embase, APA PsycINFO, CINAHL, Web of Science, and Scopus from inception to May 16, 2023, to identify sources relevant to our aim.

RESULTS

We identified 15 sources meeting our eligibility criteria. The body of literature did not meet COSMIN standards to make recommendations for use with regard to these measurement instruments.

CONCLUSION

Further research is needed, with a focus on content validity and structural validity, prior to recommending the SIQ and SIQ-Jr for use in clinical practice and in clinical trials. No specific grant funding was used for this review.

PLAIN LANGUAGE SUMMARY

In this systematic review, authors analyzed 15 sources examining measurement properties of the Suicidal Ideation Questionnaire and Suicidal Ideation Questionnaire-Jr. Both measures are designed to capture suicidal ideation in adolescents and are used in clinical practice and clinical trials. The authors identified sufficient evidence for convergent validity for both measures. Authors concluded that further research is needed to support content validity, structural validity as a unidimensional scale (as they are often used) as well as their internal consistency, test-retest reliability, discriminative validity, predictive validity, and interpretability of these measures. The authors also emphasize the need to consider the limitations of these measures for researchers studying suicidal ideation and clinicians using these measures in their assessments of young people.

Enhancing Aviation Safety through AI-Driven Mental Health Management for Pilots and Air Traffic Controllers

This article provides an overview of the mental health challenges faced by pilots and air traffic controllers (ATCs), whose stressful professional lives may negatively impact global flight safety and security. The adverse effects of mental health disorders on their flight performance pose a particular safety risk, especially in sudden unexpected startle situations. Therefore, the early detection, prediction and prevention of mental health deterioration in pilots and ATCs, particularly among those at high risk, are crucial to minimize potential air crash incidents caused by human factors. Recent research in artificial intelligence (AI) demonstrates the potential of machine and deep learning, edge and cloud computing, virtual reality and wearable multimodal physiological sensors for monitoring and predicting mental health disorders. Longitudinal monitoring and analysis of pilots' and ATCs physiological, cognitive and behavioral states could help predict individuals at risk of undisclosed or emerging mental health disorders. Utilizing AI tools and methodologies to identify and select these individuals for preventive mental health training and interventions could be a promising and effective approach to preventing potential air crash accidents attributed to human factors and related mental health problems. Based on these insights, the article advocates for the design of a multidisciplinary mental healthcare ecosystem in modern aviation using AI tools and technologies, to foster more efficient and effective mental health management, thereby enhancing flight safety and security standards. This proposed ecosystem requires the collaboration of multidisciplinary experts, including psychologists, neuroscientists, physiologists, psychiatrists, etc. to address these challenges in modern aviation.

Effectiveness of early and intensive intervention on suicide prevention: CARS programme

The aim of this study was to evaluate the outcomes of the Programme for Management of Suicidal Behaviour and Suicide Prevention (CARS). Individuals treated in the emergency department of University Hospital Marqués de Valdecilla for suicidal thoughts or attempts (N = 401) between 1-March-2016 and 31-December-2018 were considered. No randomization by patients or groups was performed. Student's t-test, chi-square and repeated measure analysis of variance were used. Kaplan-Meier survival function and Cox proportional hazard regression models were employed to estimate the risks of relapse. Outcome of those who voluntary enrol CARS were compared with treatment as usual (TAU) at 6- and 12-months follow-up. The results indicate a significant reduction and delayed occurrence of suicidal behaviour over a 12-month follow-up period with the CARS programme compared to TAU, along with a decreased frequency of hospital admissions. CARS programme demonstrates a substantial impact, significantly reducing the risk of recurrent suicidal behaviour by 35.5 % and the risk of repeated suicidal attempts by 47.2 % at the 12-month follow-up. The programme exhibits a dual protective effect, diminishing suicidal behaviour and fostering improved long-term outcomes. In conclusion, CARS effectively reduced suicidal behaviour recurrence, achieving significant decreases in suicidal thoughts, plans and attempts.

Alcohol use disorder and fitness to drive: Discrepancies between health professionals' evaluations and objective measures of alcohol use and cognitive functioning

Aims: In this study, we investigated if health professionals' evaluations of driving ability corresponded with measures of severity of alcohol use and measures of cognitive functions necessary for safely driving a car. Methods: A total of 90 participants from a multicentre study were included. Participants were categorised into three groups: (1) the group judged fit to drive (FIT); (2) the group judged not fit to drive (UNFIT); and (3) the group who had lost their driver's licence due to legal sanctions (LEGAL). The participants' AUDIT scores, earlier treatment episodes and results from neuropsychological tests of reaction time, attention and visuospatial ability were included in the analyses. Results: We found a significant difference in the severity of alcohol use disorder (AUD) and visuospatial abilities between the FIT and UNFIT groups. Half of the UNFIT group had at least mild visuospatial difficulties, compared to only a quarter in the FIT group. There were no group differences in reaction time or attentional measures. The LEGAL group had more severe AUD than the other groups. Conclusion: The FIT group did not perform differently from the UNFIT group on attention and reaction time measures. The UNFIT group had more visuospatial impairments, but even half of this group had normal scores. It is uncertain whether the differences between the two groups are of practical significance. The quality of health professionals' evaluations may be questioned, and the results highlight the need for more reliable and valid criteria for doing fitness to drive evaluations.

Development and external validation of a logistic and a penalized logistic model using machine-learning techniques to predict suicide attempts: A multicenter prospective cohort study in Korea

Despite previous efforts to build statistical models for predicting the risk of suicidal behavior using machine-learning analysis, a high-accuracy model can lead to overfitting. Furthermore, internal validation cannot completely address this problem. In this study, we created models for predicting the occurrence of suicide attempts among Koreans at high risk of suicide, and we verified these models in an independent cohort. We performed logistic and penalized regression for suicide attempts within 6 months among suicidal ideators and attempters in The Korean Cohort for the Model Predicting a Suicide and Suicide-related Behavior (K-COMPASS). We then validated the models in a test cohort. Our findings indicated that several factors significantly predicted suicide attempts in the models, including young age, suicidal ideation, previous suicidal attempts, anxiety, alcohol abuse, stress, and impulsivity. The area under the curve and positive predictive values were 0.941 and 0.484 after variable selection and 0.751 and 0.084 in the test cohort. The corresponding values for the penalized regression model were 0.943 and 0.524 in the original training cohort and 0.794 and 0.115 in the test cohort. The prediction model constructed through a prospective cohort study of the suicide high-risk group showed satisfactory accuracy even in the test cohort. The accuracy with penalized regression was greater than that with the "classical" logistic model.

Using the Strategy of Genome-Wide Association Studies to Identify Genetic Markers of Suicidal Behavior: A Narrative Review

BACKGROUND

Several studies involving various suicidal phenotypes based on the strategy of the search of genome-wide associations with single nucleotide polymorphisms have been performed recently. These studies need to be generalized.

AIM

To systematize the findings of a number of genome-wide association studies (GWAS) for suicidal phenotypes, annotate the identified markers, analyze their functionality, and possibly substantiate the hypothesis holding that these phenotypes reflect a nonspecific set of gene variants that are relevant as relates to stress-vulnerability as a key endophenotype of suicidal behavior (SB).

METHODS

A search on the PubMed and related resources using the combinations "suicide AND GWAS" and "suicidal behavior AND GWAS" was performed. It yielded a total of 34 independent studies and meta-analyses.

RESULTS

For the 10 years since such studies emerged, they have undergone significant progress. Estimates of the SNP heritability of SB in some cases are comparable with estimates of heritability based on the twin method. Many studies show a high genetic correlation with the genomic markers of the most common mental disorders (depression, bipolar disorder, schizophrenia, post-traumatic stress disorder). At the same time, a genomic architecture specific to SB is also encountered. Studies utilizing the GWAS strategy have not revealed any associations of SB with candidate genes that had been previously studied in detail (different neurotransmitters, stress response system, polyamines, etc.). Frequently reported findings from various studies belong in three main groups: 1) genes involved in cell interactions, neurogenesis, the development of brain structures, inflammation, and the immune responses; 2) genes encoding receptors for neurotrophins and various components of the intracellular signaling systems involved in synaptic plasticity, embryonic development, and carcinogenesis; and 3) genes encoding various neuro-specific proteins and regulators.

CONCLUSION

In general, GWAS in the field of suicidology mainly serve the purpose of a deeper understanding of the pathophysiology of suicidal behavior. However, they also demonstrate growing capability in terms of predicting and preventing suicide, especially when calculating the polygenic risk score among certain populations (psychiatric patients) and in combination with tests of different modalities. From our point of view, there exists a set of markers revealed by the GWAS strategy that seems to point to a leading role played by stress vulnerability, an endophenotype that is formed during early development and which subsequently comes to play the role of key pathogenetic mechanism in SB.

Validation of a Multivariable Model to Predict Suicide Attempt in a Mental Health Intake Sample

Importance

Given that suicide rates have been increasing over the past decade and the demand for mental health care is at an all-time high, targeted prevention efforts are needed to identify individuals seeking to initiate mental health outpatient services who are at high risk for suicide. Suicide prediction models have been developed using outpatient mental health encounters, but their performance among intake appointments has not been directly examined.

Objective

To assess the performance of a predictive model of suicide attempts among individuals seeking to initiate an episode of outpatient mental health care.

Design, Setting, and Participants

This prognostic study tested the performance of a previously developed machine learning model designed to predict suicide attempts within 90 days of any mental health outpatient visit. All mental health intake appointments scheduled between January 1, 2012, and April 1, 2022, at Kaiser Permanente Northern California, a large integrated health care delivery system serving over 4.5 million patients, were included. Data were extracted and analyzed from August 9, 2022, to July 31, 2023.

Main Outcome and Measures

Suicide attempts (including completed suicides) within 90 days of the appointment, determined by diagnostic codes and government databases. All predictors were extracted from electronic health records.

Results

The study included 1 623 232 scheduled appointments from 835 616 unique patients. There were 2800 scheduled appointments (0.17%) followed by a suicide attempt within 90 days. The mean (SD) age across appointments was 39.7 (15.8) years, and most appointments were for women (1 103 184 [68.0%]). The model had an area under the receiver operating characteristic curve of 0.77 (95% CI, 0.76-0.78), an area under the precision-recall curve of 0.02 (95% CI, 0.02-0.02), an expected calibration error of 0.0012 (95% CI, 0.0011-0.0013), and sensitivities of 37.2% (95% CI, 35.5%-38.9%) and 18.8% (95% CI, 17.3%-20.2%) at specificities of 95% and 99%, respectively. The 10% of appointments at the highest risk level accounted for 48.8% (95% CI, 47.0%-50.6%) of the appointments followed by a suicide attempt.

Conclusions and Relevance

In this prognostic study involving mental health intakes, a previously developed machine learning model of suicide attempts showed good overall classification performance. Implementation research is needed to determine appropriate thresholds and interventions for applying the model in an intake setting to target high-risk cases in a manner that is acceptable to patients and clinicians.

Suicide Death Prediction Using the Maryland Suicide Data Warehouse: A Sensitivity Analysis

OBJECTIVE

Nearly 50,000 Americans die each year from suicide, despite suicide death being a rare event in the context of health risk assessment and modeling. Prior research has underscored the need for contextualizing suicide risk models in terms of their potential uses and generalizability. This sensitivity analysis makes use of the Maryland Suicide Data Warehouse (MSDW) and illustrates how results inform clinical decision support.

METHOD

A cohort of 1 million living control patients were extracted from the MSDW in addition to 1,667 patients who had died by suicide between the years 2016 and 2019 according to the Maryland Office of the Medical Examiner (OCME). Data were extracted and aggregated as part of a 4-year retrospective design. Binary logistic and two penalized regression models were deployed in a repeated fivefold cross-validation. Model performances were evaluated using sensitivity, positive predictive value (PPV), and F1, and model coefficients were ranked according to coefficient size.

RESULTS

Several features were significantly associated with patients having died by suicide, including male sex, depressive and anxiety disorder diagnoses, social needs, and prior suicidal ideation and suicide attempt. Cross-validated binary logistic regression outperformed either ridge or LASSO (least absolute shrinkage and selection operator) models but generally achieved low-to-moderate PPV and sensitivity across most thresholds and a peak F1 of 0.323.

CONCLUSIONS

Suicide death prediction is constrained by the context of use, which determines the best balance of precision and recall. Predictive models must be evaluated close to the level of intervention. They may not hold up to different needs at different levels of care.

Dynamic learning of individual-level suicidal ideation trajectories to enhance mental health care

There has recently been an increase in ongoing patient-report routine outcome monitoring for individuals within clinical care, which has corresponded to increased longitudinal information about an individual. However, many models that are aimed at clinical practice have difficulty fully incorporating this information. This is in part due to the difficulty in dealing with the irregularly time-spaced observations that are common in clinical data. Consequently, we built individual-level continuous-time trajectory models of suicidal ideation for a clinical population (N = 585) with data collected via a digital platform. We demonstrate how such models predict an individual's level and variability of future suicide ideation, with implications for the frequency that individuals may need to be observed. These individual-level predictions provide a more personalised understanding than other predictive methods and have implications for enhanced measurement-based care.

Absence of nonfatal suicidal behavior preceding suicide death reveals differences in clinical risks

Nonfatal suicidality is the most robust predictor of suicide death. However, only ~10% of those who survive an attempt go on to die by suicide. Moreover, ~50% of suicide deaths occur in the absence of prior known attempts, suggesting risks other than nonfatal suicide attempt need to be identified. We studied data from 4,000 population-ascertained suicide deaths and 26,191 population controls to improve understanding of risks leading to suicide death. This study included 2,253 suicide deaths and 3,375 controls with evidence of nonfatal suicidality (SUI_SI/SB and CTL_SI/SB) from diagnostic codes and natural language processing of electronic health records notes. Characteristics of these groups were compared to 1,669 suicides with no prior nonfatal SI/SB (SUI_None) and 22,816 controls with no lifetime suicidality (CTL_None). The SUI_None and CTL_None groups had fewer diagnoses and were older than SUI_SI/SB and CTL_SI/SB. Mental health diagnoses were far less common in both the SUI_None and CTL_None groups; mental health problems were less associated with suicide death than with presence of SI/SB. Physical health diagnoses were conversely more often associated with risk of suicide death than with presence of SI/SB. Pending replication, results indicate highly significant clinical differences among suicide deaths with versus without prior nonfatal SI/SB.

Clinical prediction models for knee pain in patients with knee osteoarthritis: a systematic review

OBJECTIVE

To identify and describe existing models for predicting knee pain in patients with knee osteoarthritis.

METHODS

The electronic databases PubMed, EMBASE, CINAHL, Web of Science, and Cochrane Library were searched from their inception to May 2023 for any studies to develop and validate a prediction model for predicting knee pain in patients with knee osteoarthritis. Two reviewers independently screened titles, abstracts, and full-text qualifications, and extracted data. Risk of bias was assessed using the PROBAST. Data extraction of eligible articles was extracted by a data extraction form based on CHARMS. The quality of evidence was graded according to GRADE. The results were summarized with descriptive statistics.

RESULTS

The search identified 2693 records. Sixteen articles reporting on 26 prediction models were included targeting occurrence (n = 9), others (n = 7), progression (n = 5), persistent (n = 2), incident (n = 1), frequent (n = 1), and flares (n = 1) of knee pain. Most of the studies (94%) were at high risk of bias. Model discrimination was assessed by the AUROC ranging from 0.62 to 0.81. The most common predictors were age, BMI, gender, baseline pain, and joint space width. Only frequent knee pain had a moderate quality of evidence; all other types of knee pain had a low quality of evidence.

CONCLUSION

There are many prediction models for knee pain in patients with knee osteoarthritis that do show promise. However, the clinical extensibility, applicability, and interpretability of predictive tools should be considered during model development.

Suicide Crisis Syndrome: A systematic review

BACKGROUND

The objective of this systematic review is to describe the scientific evidence for the Suicide Crisis Syndrome (SCS), a presuicidal cognitive and affective state consisting of five symptomatic dimensions: entrapment, affective disturbance, loss of cognitive control, hyperarousal, and social withdrawal. The aim of this article is to summarize the emerging literature on the SCS and to assess the extent to which a uniform syndrome can be assumed.

METHODS

A systematic literature search was conducted in three different databases (PubMed, PsycInfo, and Google Scholar) using the search terms "Suicide Crisis Inventory," "Suicide Crisis Syndrome," "Narrative Crisis Model of Suicide," and "Suicide Trigger State."

RESULTS

In total, 37 articles from 2010 to 2022 were identified by search criteria. Twenty-one articles published between 2017 and 2022 were included in the systematic review. All but three studies were conducted in the United States and examined clinical samples of adult high-risk psychiatric in- and outpatients. Sample sizes ranged from N = 170 to 4846. The findings confirm the unidimensional structure of the proposed disorder and support the predictive validity for short-term suicidal behavior above and beyond suicidal ideation.

CONCLUSION

Despite the promising predictive validity of the SCS, a precise prediction of future suicidal behavior remains difficult.

Role of machine learning algorithms in suicide risk prediction: a systematic review-meta analysis of clinical studies

OBJECTIVE

Suicide is a complex and multifactorial public health problem. Understanding and addressing the various factors associated with suicide is crucial for prevention and intervention efforts. Machine learning (ML) could enhance the prediction of suicide attempts.

METHOD

A systematic review was performed using PubMed, Scopus, Web of Science and SID databases. We aim to evaluate the performance of ML algorithms and summarize their effects, gather relevant and reliable information to synthesize existing evidence, identify knowledge gaps, and provide a comprehensive list of the suicide risk factors using mixed method approach.

RESULTS

Forty-one studies published between 2011 and 2022, which matched inclusion criteria, were chosen as suitable. We included studies aimed at predicting the suicide risk by machine learning algorithms except natural language processing (NLP) and image processing. The neural network (NN) algorithm exhibited the lowest accuracy at 0.70, whereas the random forest demonstrated the highest accuracy, reaching 0.94. The study assessed the COX and random forest models and observed a minimum area under the curve (AUC) value of 0.54. In contrast, the XGBoost classifier yielded the highest AUC value, reaching 0.97. These specific AUC values emphasize the algorithm-specific performance in capturing the trade-off between sensitivity and specificity for suicide risk prediction. Furthermore, our investigation identified several common suicide risk factors, including age, gender, substance abuse, depression, anxiety, alcohol consumption, marital status, income, education, and occupation. This comprehensive analysis contributes valuable insights into the multifaceted nature of suicide risk, providing a foundation for targeted preventive strategies and intervention efforts.

CONCLUSIONS

The effectiveness of ML algorithms and their application in predicting suicide risk has been controversial. There is a need for more studies on these algorithms in clinical settings, and the related ethical concerns require further clarification.

Predictive model of suicide risk in Colombian university students: quantitative analysis of associated factors

Introduction

The risk of suicide and completed suicides among young university students presents critical challenges to mental and public health in Colombia and worldwide. Employing a quantifiable approach to comprehend the factors associated with these challenges can aid in visualizing the path towards anticipating and controlling this phenomenon.

Objective

Develop a predictive model for suicidal behavior in university students, utilizing predictive analytics.

Method

We conducted an observational, retrospective, cross-sectional, and analytical research study at the University of Manizales, with a focus on predictive applicability. Data from 2,436 undergraduate students were obtained from the research initiative "Building the Future: World Mental Health Surveys International College Students."

Results

The top ten predictor variables that generated the highest scores (ranking coefficients) for the sum of factors were as follows: history of sexual abuse (13.21), family history of suicide (11.68), medication (8.39), type of student (7.4), origin other than Manizales (5.86), exposure to cannabis (4.27), exposure to alcohol (4.42), history of physical abuse (3.53), religiosity (2.9), and having someone in the family who makes you feel important (3.09).

Discussion

Suicide involves complex factors within psychiatric, medical, and societal contexts. Integrated detection and intervention systems involving individuals, families, and governments are crucial for addressing these factors. Universities also play a role in promoting coping strategies and raising awareness of risks. The predictive accuracy of over 80% in identifying suicide risk underscores its significance.

Conclusion

The risk factors related to suicidal behavior align with the findings in specialized literature and research in the field. Identifying variables with higher predictive value enables us to take appropriate actions for detecting cases and designing and implementing prevention strategies.

Their final words: An analysis of suicide notes from the United States

The present study sought to explore motivations (affective, relationships, life events, injury/medical diagnosis) in suicide notes (N = 49) from the U.S. Authors ranged in age from 18 to 74 years and were majority male (73.5%). Four raters analyzed the notes and, through a series of meetings, came to a consensus on the motives behind each note writers' suicide in terms of the broader motivational themes and the narrower second-level themes. All notes were primarily affectional in nature, with some gender and age differences. For example, suicide notes from males frequently refer to financial hardships whereas suicide notes from females were more focused on lowered self-worth and notes written by younger persons focused more on affect and relationships, while notes written by older adults focused on life events and marriage difficulties and separation. Findings illuminate the varied nature of suicide motivations but also highlight important patterns across groups.

Sexual and Gender Minority Status and Suicide Mortality: An Explainable Artificial Intelligence Analysis

Objectives: Suicide risk is elevated in lesbian, gay, bisexual, and transgender (LGBT) individuals. Limited data on LGBT status in healthcare systems hinder our understanding of this risk. This study used natural language processing to extract LGBT status and a deep neural network (DNN) to examine suicidal death risk factors among US Veterans. Methods: Data on 8.8 million veterans with visits between 2010 and 2017 was used. A case-control study was performed, and suicide death risk was analyzed by a DNN. Feature impacts and interactions on the outcome were evaluated. Results: The crude suicide mortality rate was higher in LGBT patients. However, after adjusting for over 200 risk and protective factors, known LGBT status was associated with reduced risk compared to LGBT-Unknown status. Among LGBT patients, black, female, married, and older Veterans have a higher risk, while Veterans of various religions have a lower risk. Conclusion: Our results suggest that disclosed LGBT status is not directly associated with an increase suicide death risk, however, other factors (e.g., depression and anxiety caused by stigma) are associated with suicide death risks.

Predicting Suicide Among US Veterans Using Natural Language Processing-enriched Social and Behavioral Determinants of Health

Despite recognizing the critical association between social and behavioral determinants of health (SBDH) and suicide risk, SBDHs from unstructured electronic health record (EHR) notes for suicide predictive modeling remain underutilized. This study investigates the impact of SBDH, identified from both structured and unstructured data utilizing a natural language processing (NLP) system, on suicide prediction within 7, 30, 90, and 180 days of discharge. Using EHR data of 2,987,006 Veterans between October 1, 2009, and September 30, 2015, from the US Veterans Health Administration (VHA), we designed a case-control study that demonstrates that incorporating structured and NLP-extracted SBDH significantly enhances the performance of three architecturally distinct suicide predictive models - elastic-net logistic regression, random forest (RF), and multilayer perceptron. For example, RF achieved notable improvements in suicide prediction within 180 days of discharge, with an increase in the area under the receiver operating characteristic curve from 83.57-84.25% (95% CI = 0.63%-0.98%, p-val < 0.001) and the area under the precision recall curve from 57.38-59.87% (95% CI = 3.86%-4.82%, p-val < 0.001) after integrating NLP-extracted SBDH. These findings underscore the potential of NLP-extracted SBDH in enhancing suicide prediction across various prediction timeframes, offering valuable insights for healthcare practitioners and policymakers.

Developing a risk prediction model for death at first suicide attempt-Identifying risk factors from Thailand's national suicide surveillance system data

More than 60% of suicides globally are estimated to take place in low- and middle-income nations. Prior research on suicide has indicated that over 50% of those who die by suicide do so on their first attempt. Nevertheless, there is a dearth of knowledge on the attributes of individuals who die on their first attempt and the factors that can predict mortality on the first attempt in these regions. The objective of this study was to create an individual-level risk-prediction model for mortality on the first suicide attempt. We analyzed records of individuals' first suicide attempts that occurred between May 1, 2017, and April 30, 2018, from the national suicide surveillance system, which includes all of the provinces of Thailand. Subsequently, a risk-prediction model for mortality on the first suicide attempt was constructed utilizing multivariable logistic regression and presented through a web-based application. The model's performance was assessed by calculating the area under the receiver operating curve (AUC), as well as measuring its sensitivity, specificity, and accuracy. Out of the 3,324 individuals who made their first suicide attempt, 50.5% of them died as a result of that effort. Nine out of the 21 potential predictors demonstrated the greatest predictive capability. These included male sex, age over 50 years old, unemployment, having a depressive disorder, having a psychotic illness, experiencing interpersonal problems such as being aggressively criticized or desiring plentiful attention, having suicidal intent, and displaying suicidal warning signals. The model demonstrated a good predictive capability, with an AUC of 0.902, a sensitivity of 84.65%, a specificity of 82.66%, and an accuracy of 83.63%. The implementation of this predictive model can assist physicians in conducting comprehensive evaluations of suicide risk in clinical settings and devising treatment plans for preventive intervention.

Predicting adolescent suicidal behavior following inpatient discharge using structured and unstructured data

BACKGROUND

The objective was to develop and assess performance of an algorithm predicting suicide-related ICD codes within three months of psychiatric discharge.

METHODS

This prognostic study used a retrospective cohort of EHR data from 2789 youth (12 to 20 years old) hospitalized in a safety net institution in the Northeastern United States. The dataset combined structured data with unstructured data obtained through natural language processing of clinical notes. Machine learning approaches compared gradient boosting to random forest analyses.

RESULTS

Area under the ROC and precision-recall curve were 0.88 and 0.17, respectively, for the final Gradient Boosting model. The cutoff point of the model-generated predicted probabilities of suicide that optimally classified the individual as high risk or not was 0.009. When applying the chosen cutoff (0.009) to the hold-out testing set, the model correctly identified 8 positive cases out of 10, and 418 negative cases out 548. The corresponding performance metrics showed 80 % sensitivity, 76 % specificity, 6 % PPV, 99 % NPV, F-1 score of 0.11, and an accuracy of 76 %.

LIMITATIONS

The data in this study comes from a single health system, possibly introducing bias in the model's algorithm. Thus, the model may have underestimated the incidence of suicidal behavior in the study population. Further research should include multiple system EHRs.

CONCLUSIONS

These performance metrics suggest a benefit to including both unstructured and structured data in design of predictive algorithms for suicidal behavior, which can be integrated into psychiatric services to help assess risk.

Predicting Lifetime Suicide Attempts in a Community Sample of Adolescents Using Machine Learning Algorithms

Suicide is a major global health concern and a prominent cause of death in adolescents. Previous research on suicide prediction has mainly focused on clinical or adult samples. To prevent suicides at an early stage, however, it is important to screen for risk factors in a community sample of adolescents. We compared the accuracy of logistic regressions, elastic net regressions, and gradient boosting machines in predicting suicide attempts by 17-year-olds in the Millennium Cohort Study (N = 7,347), combining a large set of self- and other-reported variables from different categories. Both machine learning algorithms outperformed logistic regressions and achieved similar balanced accuracies (.76 when using data 3 years before the self-reported lifetime suicide attempts and .85 when using data from the same measurement wave). We identified essential variables that should be considered when screening for suicidal behavior. Finally, we discuss the usefulness of complex machine learning models in suicide prediction.

Randomized Controlled Comparative Effectiveness Trial of Risk Model-Guided Clinical Decision Support for Suicide Screening

Suicide prevention requires risk identification, appropriate intervention, and follow-up. Traditional risk identification relies on patient self-reporting, support network reporting, or face-to-face screening with validated instruments or history and physical exam. In the last decade, statistical risk models have been studied and more recently deployed to augment clinical judgment. Models have generally been found to be low precision or problematic at scale due to low incidence. Few have been tested in clinical practice, and none have been tested in clinical trials to our knowledge.

Methods

We report the results of a pragmatic randomized controlled trial (RCT) in three outpatient adult Neurology clinic settings. This two-arm trial compared the effectiveness of Interruptive and Non-Interruptive Clinical Decision Support (CDS) to prompt further screening of suicidal ideation for those predicted to be high risk using a real-time, validated statistical risk model of suicide attempt risk, with the decision to screen as the primary end point. Secondary outcomes included rates of suicidal ideation and attempts in both arms. Manual chart review of every trial encounter was used to determine if suicide risk assessment was subsequently documented.

Results

From August 16, 2022, through February 16, 2023, our study randomized 596 patient encounters across 561 patients for providers to receive either Interruptive or Non-Interruptive CDS in a 1:1 ratio. Adjusting for provider cluster effects, Interruptive CDS led to significantly higher numbers of decisions to screen (42%=121/289 encounters) compared to Non-Interruptive CDS (4%=12/307) (odds ratio=17.7, p-value <0.001). Secondarily, no documented episodes of suicidal ideation or attempts occurred in either arm. While the proportion of documented assessments among those noting the decision to screen was higher for providers in the Non-Interruptive arm (92%=11/12) than in the Interruptive arm (52%=63/121), the interruptive CDS was associated with more frequent documentation of suicide risk assessment (63/289 encounters compared to 11/307, p-value<0.001).

Conclusions

In this pragmatic RCT of real-time predictive CDS to guide suicide risk assessment, Interruptive CDS led to higher numbers of decisions to screen and documented suicide risk assessments. Well-powered large-scale trials randomizing this type of CDS compared to standard of care are indicated to measure effectiveness in reducing suicidal self-harm. ClinicalTrials.gov Identifier: NCT05312437.

Neurological, Metabolic, and Psychopathological Correlates of Lifetime Suicidal Behaviour in Major Depressive Disorder without Current Suicide Ideation

INTRODUCTION

Suicidal behaviour (SB) has a complex aetiology. Although suicidal ideation (SI) is considered the most important risk factor for future attempts, many people who engage in SB do not report it.

METHODS

We investigated neurological, metabolic, and psychopathological correlates of lifetime SB in two independent groups of patients with major depression (sample 1: n = 230; age: 18-65 years; sample 2: n = 258; age >60 years) who did not report SI during an index episode.

RESULTS

Among adults (sample 1), SB was reported by 141 subjects (58.7%) and severe SB by 33 (15%). After controlling for interactions, four risk factors for SB emerged: male gender (OR 2.55; 95% CI: 1.06-6.12), negative self-perception (OR 1.76; 95% CI: 1.08-2.87), subthreshold hypomania (OR 4.50; 95% CI: 1.57-12.85), and sexual abuse (OR 3.09; 95% CI: 1.28-7.48). The presence of at least two of these factors had the best accuracy in predicting SB: sensitivity = 57.6% (39.2-74.5); specificity = 75.1% (68.5-82.0); PPV = 27.9% (20.9-37.2); NPV = 91.4% (87.6-94.1). In older patients (sample 2), 23 subjects (9%) reported previous suicide attempts, which were characterized by earlier onset (25 years: OR 0.95: 0.92-0.98), impaired verbal performance (verbal fluency: OR 0.95: 0.89-0.99), higher HDL cholesterol levels (OR 1.04: 1.00-1.07) and more dyskinesias (OR 2.86: 1.22-6.70).

CONCLUSION

Our findings suggest that SB is common in major depressive disorder, even when SI is not reported. In these individuals it is feasible and recommended to investigate both psychiatric and organic risk factors. The predictive power of models excluding SI is comparable to that of models including SI.

Analysis and evaluation of explainable artificial intelligence on suicide risk assessment

This study explores the effectiveness of Explainable Artificial Intelligence (XAI) for predicting suicide risk from medical tabular data. Given the common challenge of limited datasets in health-related Machine Learning (ML) applications, we use data augmentation in tandem with ML to enhance the identification of individuals at high risk of suicide. We use SHapley Additive exPlanations (SHAP) for XAI and traditional correlation analysis to rank feature importance, pinpointing primary factors influencing suicide risk and preventive measures. Experimental results show the Random Forest (RF) model is excelling in accuracy, F1 score, and AUC (>97% across metrics). According to SHAP, anger issues, depression, and social isolation emerge as top predictors of suicide risk, while individuals with high incomes, esteemed professions, and higher education present the lowest risk. Our findings underscore the effectiveness of ML and XAI in suicide risk assessment, offering valuable insights for psychiatrists and facilitating informed clinical decisions.

The use of machine learning on administrative and survey data to predict suicidal thoughts and behaviors: a systematic review

Background

Machine learning is a promising tool in the area of suicide prevention due to its ability to combine the effects of multiple risk factors and complex interactions. The power of machine learning has led to an influx of studies on suicide prediction, as well as a few recent reviews. Our study distinguished between data sources and reported the most important predictors of suicide outcomes identified in the literature.

Objective

Our study aimed to identify studies that applied machine learning techniques to administrative and survey data, summarize performance metrics reported in those studies, and enumerate the important risk factors of suicidal thoughts and behaviors identified.

Methods

A systematic literature search of PubMed, Medline, Embase, PsycINFO, Web of Science, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Allied and Complementary Medicine Database (AMED) to identify all studies that have used machine learning to predict suicidal thoughts and behaviors using administrative and survey data was performed. The search was conducted for articles published between January 1, 2019 and May 11, 2022. In addition, all articles identified in three recently published systematic reviews (the last of which included studies up until January 1, 2019) were retained if they met our inclusion criteria. The predictive power of machine learning methods in predicting suicidal thoughts and behaviors was explored using box plots to summarize the distribution of the area under the receiver operating characteristic curve (AUC) values by machine learning method and suicide outcome (i.e., suicidal thoughts, suicide attempt, and death by suicide). Mean AUCs with 95% confidence intervals (CIs) were computed for each suicide outcome by study design, data source, total sample size, sample size of cases, and machine learning methods employed. The most important risk factors were listed.

Results

The search strategy identified 2,200 unique records, of which 104 articles met the inclusion criteria. Machine learning algorithms achieved good prediction of suicidal thoughts and behaviors (i.e., an AUC between 0.80 and 0.89); however, their predictive power appears to differ across suicide outcomes. The boosting algorithms achieved good prediction of suicidal thoughts, death by suicide, and all suicide outcomes combined, while neural network algorithms achieved good prediction of suicide attempts. The risk factors for suicidal thoughts and behaviors differed depending on the data source and the population under study.

Conclusion

The predictive utility of machine learning for suicidal thoughts and behaviors largely depends on the approach used. The findings of the current review should prove helpful in preparing future machine learning models using administrative and survey data.

Systematic review registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022333454 identifier CRD42022333454.

The Emergency Psychiatric Interview

A quality clinical interview helps establish a good collaborative relationship with the patient. This is especially important when emergency physicians conduct a psychiatric interview. Familiarity with interview techniques, empathic listening, and observation of nonverbal cues, behavior, and appearance enhance diagnostic excellence.

A Randomized Control Trial of a Digital Health Tool for Safer Firearm and Medication Storage for Patients with Suicide Risk

Most patients with suicide risk do not receive recommendations to reduce access to lethal means due to a variety of barriers (e.g., lack of provider time, training). Determine if highly efficient population-based EHR messaging to visit the Lock to Live (L2L) decision aid impacts patient-reported storage behaviors. Randomized trial. Integrated health care system serving Denver, CO. Served by primary care or mental health specialty clinic in the 75-99.5th risk percentile on a suicide attempt or death prediction model. Lock to Live (L2L) is a web-based decision aid that incorporates patients' values into recommendations for safe storage of lethal means, including firearms and medications. Anonymous survey that determined readiness to change: pre-contemplative (do not believe in safe storage), contemplative (believe in safe storage but not doing it), preparation (planning storage changes) or action (safely storing). There were 21,131 patients randomized over a 6-month period with a 27% survey response rate. Many (44%) had access to a firearm, but most of these (81%) did not use any safe firearm storage behaviors. Intervention patients were more likely to be categorized as preparation or action compared to controls for firearm storage (OR = 1.30 (1.07-1.58)). When examining action alone, there were no group differences. There were no statistically significant differences for any medication storage behaviors. Selection bias in those who responded to survey. Efficiently sending an EHR invitation message to visit L2L encouraged patients with suicide risk to consider safer firearm storage practices, but a stronger intervention is needed to change storage behaviors. Future studies should evaluate whether combining EHR messaging with provider nudges (e.g., brief clinician counseling) changes storage behavior.ClinicalTrials.gov: NCT05288517.

Can clinician's risk assessments distinguish those who disclose suicidal ideation from those who attempt suicide?

Participants were 85 individuals who made suicide attempts within two years of their Improving Access to Psychological Therapies (IAPT) assessment, identified using record linkage. Two comparison groups, non-suicidal controls (n = 1416) and (ideators, n = 743) were compared on variables extracted from the standardized IAPT risk assessment interview. Disclosure of a historical suicide attempt or non-suicidal self-injury (NSSI) distinguished those making an attempt from those with suicidal ideation only, but suicidal intent did not. A third of the participants concealed a historical suicide attempt. The IAPT Phobia Scale classified 49.30% of attempters with 100% specificity. The IAPT Phobia Scale may have clinical value in assessing risk but requires validation. Past suicide attempt and NSSI have better clinical risk assessment utility than current suicidal ideation intensity. Risk assessment relying on disclosure is likely to be flawed and risks support being withheld from those assumed to be at lower risk.

Scoping Review of Postvention for Mental Health Providers Following Patient Suicide

INTRODUCTION

As suicides among military personnel continue to climb, we sought to determine best practices for supporting military mental health clinicians following patient suicide loss (i.e., postvention).

MATERIALS AND METHODS

We conducted a scoping review of the literature using Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines. Our initial search of academic databases generated 2,374 studies, of which 122 were included in our final review. We categorized postvention recommendations based on the socioecological model (i.e., recommendations at the individual provider, supervisory/managerial, organizational, and discipline levels) and analyzed them using a narrative synthesizing approach.

RESULTS

Extracted recommendations (N = 358) comprised those at the provider (n = 94), supervisory/managerial (n = 90), organization (n = 105), and discipline (n = 69) levels.

CONCLUSIONS

The literature converges on the need for formal postvention protocols that prioritize (1) training and education and (2) emotional and instrumental support for the clinician. Based on the scoped literature, we propose a simple postvention model for military mental health clinicians and recommend a controlled trial testing of its effectiveness.

ASSIP vs. Crisis Counseling for Preventing Suicide Re-attempts: Outcome Predictor Analysis of a Randomized Clinical Trial Data

OBJECTIVE

Knowledge of the effectiveness and limits of the suitability of brief interventions in suicide prevention is greatly needed. We investigated subgroup differences and predictors for suicide re-attempts within a clinical trial population recruited for a brief intervention to prevent re-attempts.

METHODS

Consenting adult patients receiving treatment for a suicide attempt in Helsinki City general hospital emergency rooms in 2016-2017 (n = 239) were randomly allocated to (a) the Attempted Suicide Short Intervention Program (ASSIP) or (b) Crisis Counseling (CC). Participants also received their usual treatment. Information on primary outcome repeat attempts and secondary outcomes was collected via telephone and from medical and psychiatric records for 2 years. As proportions of re-attempts did not differ significantly between ASSIP and CC (29.2 vs. 35.2%), patients were pooled and predictors for suicide re-attempts were analyzed using Kaplan-Meier and logistic regression analyses.

RESULTS

Re-attempts were predicted by participants' younger age (OR 0.965 [0.933-0.998]), previous suicide attempts (OR 2.437 [1.106-5.370]), psychiatric hospitalization in the year preceding baseline (OR 3.256 [1.422-7.458]), and clinical diagnosis of a personality disorder (OR 4.244 [1.923-9.370]), especially borderline personality disorder (OR 5.922 [2.558-13.709]).

CONCLUSIONS

Within a population of suicide attempters consenting to a brief intervention trial, the risk of re-attempt was strongly predicted by subjects' young age, history of previous attempts, psychiatric hospitalizations, and personality disorder, particularly borderline personality disorder. The composition of treated populations with regard to these characteristics may strongly influence the observed success of brief interventions. Their potential as moderators of treatment effectiveness and as indicators of the utility of brief interventions warrants further investigation.HIGHLIGHTSDuring the 2-year follow-up, 32% of trial participants reattempted suicide.Rates of reattempts varied and were strongly predicted by clinical subgroup.Subgroup composition may strongly influence brief interventions' observed outcome.