Known unknowns and unknown unknowns in suicide risk assessment: evidence from meta-analyses of aleatory and epistemic uncertainty

Practical strategies for care of individuals impacted by suicide in the emergency department: A narrative review

Introduction Suicide is among the leading causes of death and nurses care for survivors of suicide at many points in their grief journey. Every individual who dies by suicide leaves behind multiple affected survivors and how they are cared for immediately following the suicide influences how they cope with the death. The purpose of this article is to make recommendations for the care of survivors of suicide loss in the emergency department.

METHODS

A narrative review of the current literature was conducted using the databases CINAHL and ProQuest. Articles focused on survivors of suicide loss and their care in the immediate period after the death were utilized.

RESULTS

Four hundred and twenty-nine articles were found. Applying established inclusion and exclusion criteria and quality assessment using the SANSA guideline, 29 were included in the review.

DISCUSSION

Three themes were identified: 1. Risks for suicide in survivors of suicide loss; 2. Interventions in the immediate period after suicide loss; and 3. Active suicide postvention as suicide prevention. Emergency department nurses need to have the ability to readily assess and recognize the survivors of suicide loss who are at higher risk for complicated grieving, and providing rapid and immediate services and resources will help promote coping and positive mental health outcomes in survivors.

One-Week Suicide Risk Prediction Using Real-Time Smartphone Monitoring: Prospective Cohort Study

BACKGROUND

Suicide is a major global public health issue that is becoming increasingly common despite preventive efforts. Though current methods for predicting suicide risk are not sufficiently accurate, technological advances provide invaluable tools with which we may evolve toward a personalized, predictive approach.

OBJECTIVE

We aim to predict the short-term (1-week) risk of suicide by identifying changes in behavioral patterns characterized through real-time smartphone monitoring in a cohort of patients with suicidal ideation.

METHODS

We recruited 225 patients between February 2018 and March 2020 with a history of suicidal thoughts and behavior as part of the multicenter SmartCrisis study. Throughout 6 months of follow-up, we collected information on the risk of suicide or mental health crises. All participants underwent voluntary passive monitoring using data generated by their own smartphones, including distance walked and steps taken, time spent at home, and app usage. The algorithm constructs daily activity profiles for each patient according to these data and detects changes in the distribution of these profiles over time. Such changes are considered critical periods, and their relationship with suicide-risk events was tested.

RESULTS

During follow-up, 18 (8%) participants attempted suicide, and 14 (6.2%) presented to the emergency department for psychiatric care. The behavioral changes identified by the algorithm predicted suicide risk in a time frame of 1 week with an area under the curve of 0.78, indicating good accuracy.

CONCLUSIONS

We describe an innovative method to identify mental health crises based on passively collected information from patients' smartphones. This technology could be applied to homogeneous groups of patients to identify different types of crises.

Calibrating violence risk assessments for uncertainty

Psychiatrists and other mental health clinicians are often tasked with assessing patients' risk of violence. Approaches to this vary and include both unstructured (based on individual clinicians' judgement) and structured methods (based on formalised scoring and algorithms with varying scope for clinicians' judgement). The end result is usually a categorisation of risk, which may, in turn, reference a probability estimate of violence over a certain time period. Research over recent decades has made considerable improvements in refining structured approaches and categorising patients' risk classifications at a group level. The ability, however, to apply these findings clinically to predict the outcomes of individual patients remains contested. In this article, we review methods of assessing violence risk and empirical findings on their predictive validity. We note, in particular, limitations in calibration (accuracy at predicting absolute risk) as distinct from discrimination (accuracy at separating patients by outcome). We also consider clinical applications of these findings, including challenges applying statistics to individual patients, and broader conceptual issues in distinguishing risk and uncertainty. Based on this, we argue that there remain significant limits to assessing violence risk for individuals and that this requires careful consideration in clinical and legal contexts.

Newtonian science, complexity science and suicide-critically analysing the philosophical basis for suicide research: A discussion paper

AIM

A critical discussion comparing Newtonian science and complexity science as the philosophical basis for suicide research and its impact on suicide knowledge development and clinical practice.

DESIGN

Discussion paper.

DATA SOURCES

A review of literature on suicide research and complexity science ranging from 2000 to 2022.

IMPLICATIONS FOR NURSING

Suicide research based on a Newtonian worldview can have negative consequences for suicide knowledge development and can permeate nursing practice in ways that take away from addressing the complex needs of patients, their families and healthcare teams.

CONCLUSION

A Newtonian worldview as a philosophical basis for research is insufficient for the study of a phenomenon as complex as suicide. A complexity science approach is better suited to the study of suicide given the multiple, interrelated, emerging factors that can contribute to a person's decision to end their own life.

IMPACT

Suggestions are provided as to how a complexity science approach to the research of suicide can inform useful knowledge development that better meets the needs of individuals facing suicidality and their families. Researchers, healthcare administrators and nurses providing care to those struggling with suicidality can benefit from adopting a complexity science worldview in addressing this multifaceted phenomenon.

Ethical concerns relating to genetic risk scores for suicide

Genome-wide association studies (GWAS) provide valuable information in research contexts regarding genomic changes that contribute to risks for complex psychiatric conditions like major depressive disorder. GWAS results can be used to calculate polygenic risk scores (PRS) for psychiatric conditions, such as bipolar disorder or schizophrenia, as well as for other traits, such as obesity or hypertension. Private companies that provide direct-to-consumer (DTC) genetic testing sometimes report PRS for a variety of traits. Recently, the first well-powered GWAS study for suicide death was published. PRS reports that claim to assess suicide risk are therefore likely to appear soon in the DTC setting. We describe ethical concerns regarding the commercial use of GWAS results related to suicide. We identify several issues that must be addressed before PRS for suicide risk is made available to the public through DTC: (a) the potential for misinterpretation of results, (b) consumers' perceptions about determinism and behavior change, (c) potential contributions to stigma, discrimination, and health disparities; and (d) ethical problems regarding the testing of children and vulnerable adults. Tests for genetic prediction of suicidality may eventually have clinical significance, but until then, the potential for individual and public harm significantly outweighs any potential benefit. Even if genetic prediction of suicidality improves significantly, information about genetic risk scores must be distributed cautiously, with genetic counseling, and with adequate safeguards.

Problemi della previsione in psichiatria

Sono esaminati vari problemi relativi alla previsione in psichiatria. I dati disponibili mostrano, in modo simile alle scienze sociali, ampi limiti nella capacità previsionale, specie per quanto riguarda il suicidio, la violenza e altri aspetti comportamentali. Vengono esaminate le difficoltà che nascono dal cercare di derivare il futuro della persona dal suo passato, la mancata coerenza fra aspetti di personalità e possibili comportamenti e il privilegio dato a strumenti psicopatologici incentrati sul singolo caso, rispetto a quelli attuariali con valutazioni testistiche e statistiche. Vengono anche evidenziati i numerosi bias cognitivi che distorcono le previsioni, in particolare l'errore fondamentale di attribuzione, che privilegia aspetti personologici rispetto a quelli situazionali. Ma altri bias hanno una importante azione distorsiva, da quelli della rappresentatività a quelli della disponibilità, da quelli statistici, al framing o al priming. Emerge una psichiatria molto legata nelle pratiche ancora al senso comune e alla folk psychology, con la ricchezza ma anche i molti errori che la caratterizzano. Di fatto esiste una modesta capacità previsionale riconosciuta alla psicologia popolare e alla psichiatria, ma è legata più a vincoli situazionali che a modelli personologici e psicopatologici e in ogni caso scarsamente affidabile per la previsione clinica in psichiatria.

Meta-analysis of the strength of exploratory suicide prediction models; from clinicians to computers

BACKGROUND

Suicide prediction models have been formulated in a variety of ways and are heterogeneous in the strength of their predictions. Machine learning has been a proposed as a way of improving suicide predictions by incorporating more suicide risk factors.

AIMS

To determine whether machine learning and the number of suicide risk factors included in suicide prediction models are associated with the strength of the resulting predictions.

METHOD

Random-effect meta-analysis of exploratory suicide prediction models constructed by combining two or more suicide risk factors or using clinical judgement (Prospero Registration CRD42017059665). Studies were located by searching for papers indexed in PubMed before 15 August 2020 with the term suicid* in the title.

RESULTS

In total, 86 papers reported 102 suicide prediction models and included 20 210 411 people and 106 902 suicides. The pooled odds ratio was 7.7 (95% CI 6.7-8.8) with high between-study heterogeneity (I2 = 99.5). Machine learning was associated with a non-significantly higher odds ratio of 11.6 (95% CI 6.0-22.3) and clinical judgement with a non-significantly lower odds ratio of 4.7 (95% CI 2.1-10.9). Models including a larger number of suicide risk factors had a higher odds ratio when machine-learning studies were included (P = 0.02). Among non-machine-learning studies, suicide prediction models including fewer risk factors performed just as well as those including more risk factors.

CONCLUSIONS

Machine learning might have the potential to improve the performance of suicide prediction models by increasing the number of included suicide risk factors but its superiority over other methods is unproven.

Can machine-learning methods really help predict suicide?

PURPOSE OF REVIEW

In recent years there has been interest in the use of machine learning in suicide research in reaction to the failure of traditional statistical methods to produce clinically useful models of future suicide. The current review summarizes recent prediction studies in the suicide literature including those using machine learning approaches to understand what value these novel approaches add.

RECENT FINDINGS

Studies using machine learning to predict suicide deaths report area under the curve that are only modestly greater than, and sensitivities that are equal to, those reported in studies using more conventional predictive methods. Positive predictive value remains around 1% among the cohort studies with a base rate that was not inflated by case-control methodology.

SUMMARY

Machine learning or artificial intelligence may afford opportunities in mental health research and in the clinical care of suicidal patients. However, application of such techniques should be carefully considered to avoid repeating the mistakes of existing methodologies. Prediction studies using machine-learning methods have yet to make a major contribution to our understanding of the field and are unproven as clinically useful tools.

Inconvenient truths in suicide prevention: Why a Restorative Just Culture should be implemented alongside a Zero Suicide Framework

OBJECTIVE

The prevailing paradigm in suicide prevention continues to contribute to the nihilism regarding the ability to prevent suicides in healthcare settings and a sense of blame following adverse incidents. In this paper, these issues are discussed through the lens of clinicians' experiences as second victims following a loss of a consumer to suicide, and the lens of health care organisations.

METHOD

We discuss challenges related to the fallacy of risk prediction (erroneous belief that risk screening can be used to predict risk or allocate resources), and incident reviews that maintain a retrospective linear focus on errors and are highly influenced by hindsight and outcome biases.

RESULTS

An argument that a Restorative Just Culture should be implemented alongside a Zero Suicide Framework is developed.

CONCLUSIONS

The current use of algorithms to determine culpability following adverse incidents, and a linear approach to learning ignores the complexity of the healthcare settings and can have devastating effects on staff and the broader healthcare community. These issues represent 'inconvenient truths' that must be identified, reconciled and integrated into our future pathways towards reducing suicides in health care. The introduction of Zero Suicide Framework can support the much-needed transition from relying on a retrospective focus on errors (Safety I) to a more prospective focus which acknowledges the complexities of healthcare (Safety II), when based on the Restorative Just Culture principles. Restorative Just Culture replaces backward-looking accountability with a focus on the hurts, needs and obligations of all who are affected by the event. In this paper, we argue that the implementation of Zero Suicide Framework may be compromised if not supported by a substantial workplace cultural change. The process of responding to critical incidents implemented at the Gold Coast Mental Health and Specialist Services is provided as an example of a successful implementation of Restorative Just Culture-based principles that has achieved a culture change required to support learning, improving and healing for our consumers, their families, our staff and broader communities.

Suicide prevention in community mental health: changing systems

Purpose

This paper aims to describe a performance improvement process related to suicide assessment in a community mental health center. As suicide rates rise in the USA, it is crucial that community mental health providers are capable and comfortable to assess for suicide risk among individuals with mental illness. Support for healthcare providers is emphasized in the quadruple aim model of enhancing healthcare delivery and patient experience. The quadruple aim model is applied in the present performance improvement project in a community mental health center.

Design/methodology/approach

An interprofessional team used provider survey responses, critical incident data and other stakeholder input to implement a new assessment mechanism and education plan to support direct care staff to address suicide risk.

Findings

Although the rate of patient death by suicide at the community mental health center is low, managing risk is a frequent provider concern. Providers’ comfort assessing and managing suicide risk varied widely based on survey responses. A structured suicide assessment process was implemented to offer clarity and direction for providers. Education to address assessment and management was designed and implemented.

Research limitations/implications

Suicide data were retrospective and limited to known deaths, thus there may have been higher numbers of deaths by suicide historically. Providers’ comfort with suicide risk management was based on self-report and future work should also integrate skills-based assessment.

Originality/value

Improving the provider experience in mental health care must be explored. Focusing on provider input and voice in suicide-related efforts in community settings is a step toward integrating the quadruple aim ideals into mental health care.

On Anticipatory Accounts

Engaging an account of a judicial decision made in the Los Angeles Mental Health Court, this article interrogates the role of anticipation in the lived negotiation of moral, social and institutional orders. As Judge Samuel Benton recounts his attempt to let himself ‘emotionally off the hook’ in the wake of a patient’s suicide, anticipation emerges as: 1) an ordered, linear sequencing of events towards logical ends; 2) unsettled, temporally disjunctive engagements with the past in order to make sense of present experience and ambiguous futures; 3) existential negotiations of one’s potential morality and social belonging; and 4) distributed organization of information between people and across objects in order to elaborate present and future experience. These manifestations of anticipation reveal the social and temporal contingency and deep intersubjectivity of our negotiations with uncertainty in the unsettling process of becoming moral.

The prediction of suicide in severe mental illness: development and validation of a clinical prediction rule (OxMIS)

Assessment of suicide risk in individuals with severe mental illness is currently inconsistent, and based on clinical decision-making with or without tools developed for other purposes. We aimed to develop and validate a predictive model for suicide using data from linked population-based registers in individuals with severe mental illness. A national cohort of 75,158 Swedish individuals aged 15-65 with a diagnosis of severe mental illness (schizophrenia-spectrum disorders, and bipolar disorder) with 574,018 clinical patient episodes between 2001 and 2008, split into development (58,771 patients, 494 suicides) and external validation (16,387 patients, 139 suicides) samples. A multivariable derivation model was developed to determine the strength of pre-specified routinely collected socio-demographic and clinical risk factors, and then tested in external validation. We measured discrimination and calibration for prediction of suicide at 1 year using specified risk cut-offs. A 17-item clinical risk prediction model for suicide was developed and showed moderately good measures of discrimination (c-index 0.71) and calibration. For risk of suicide at 1 year, using a pre-specified 1% cut-off, sensitivity was 55% (95% confidence interval [CI] 47-63%) and specificity was 75% (95% CI 74-75%). Positive and negative predictive values were 2% and 99%, respectively. The model was used to generate a simple freely available web-based probability-based risk calculator (Oxford Mental Illness and Suicide tool or OxMIS) without categorical cut-offs. A scalable prediction score for suicide in individuals with severe mental illness is feasible. If validated in other samples and linked to effective interventions, using a probability score may assist clinical decision-making.

Community pharmacists' experiences and people at risk of suicide in Canada and Australia: a thematic analysis

PURPOSE

To explore Canadian and Australian community pharmacists' practice experiences in caring for people at risk of suicide.

METHODS

We conducted a thematic analysis of 176 responses to an open-ended extension question in an online survey.

RESULTS

Four themes were identified and include referrals and triage, accessibility for confiding, emotional toll, and stigma. Subthemes included gatekeeping the medication supply, sole disclosure, planning for end of life, concerns of support people, assessing the validity of suicidality, gaps in the system, not directly asking, ill-equipped, resources in the pharmacy, relying on others to continue care, and attention seeking.

CONCLUSIONS

Community pharmacists are caring for patients at risk of suicide frequently, and often with patients seeking the help of pharmacists directly. Pharmacists engage in activities and actions that would be considered outside of the traditional dispensing roles and provide support and intervention to people at risk of suicide through collaboration and other mechanisms. Further research to determine appropriate education and training and postvention supports is required.

Short term risk of non-fatal and fatal suicidal behaviours: the predictive validity of the Columbia-Suicide Severity Rating Scale in a Swedish adult psychiatric population with a recent episode of self-harm

BACKGROUND

The Columbia-Suicide Severity Rating Scale (C-SSRS) is a relatively new instrument for the assessment of suicidal ideation and behaviour that is widely used in clinical and research settings. The predictive properties of the C-SSRS have mainly been evaluated in young US populations. We wanted to examine the instrument's predictive validity in a Swedish cohort of adults seeking psychiatric emergency services after an episode of self-harm.

METHODS

Prospective cohort study of patients (n = 804) presenting for psychiatric emergency assessment after an episode of self-harm with or without suicidal intent. Suicidal ideation and behaviours at baseline were rated with the C-SSRS and subsequent non-fatal and fatal suicide attempts within 6 months were identified by record review. Logistic regression was used to evaluate separate ideation items and total scores as predictors of non-fatal and fatal suicide attempts. Receiver operating characteristics (ROC) curves were constructed for the suicidal ideation (SI) intensity score and the C-SSRS total score.

RESULTS

In this cohort, the median age at baseline was 33 years, 67% were women and 68% had made at least one suicide attempt prior to the index attempt. At least one non-fatal or fatal suicide attempt was recorded during follow-up for 165 persons (20.5%). The single C-SSRS items frequency, duration and deterrents were associated with this composite outcome; controllability and reasons were not. In a logistic regression model adjusted for previous history of suicide attempt, SI intensity score was a significant predictor of a non-fatal or fatal suicide attempt (OR 1.08; 95% CI 1.03-1.12). ROC analysis showed that the SI intensity score was somewhat better than chance in correctly classifying the outcome (AUC 0.62, 95% CI 0.57-0.66). The corresponding figures for the C-SSRS total score were 0.65, 95% CI 0.60-0.69.

CONCLUSIONS

The C-SSRS items frequency, duration and deterrents were associated with elevated short term risk in this adult psychiatric cohort, as were both the SI intensity score and the C-SSRS total score. However, the ability to correctly predict future suicidal behaviour was limited for both scores.

The androgen model of suicide completion

Suicide is a devastating public health issue that imposes severe psychological, social, and economic burdens not only for the individuals but also for their relatives, friends, clinicians, and the general public. Among the different suicidal behaviors, suicide completion is the worst and the most relevant outcome. The knowledge of biological etiopathological mechanisms involved in suicide completion is limited. Hitherto, no objective markers, either alone or in combination, can reliably predict who will complete a suicide. However, such parameters are strongly needed to establish and optimize prediction and prevention. We introduce here a novel ideation-to-completion framework in suicide research and discuss the problems of studies aiming at identifying and validating clinically useful markers. The male gender is a specific risk factor for suicide, which suggests that androgen effects are implicated in the transition from suicidal ideation to suicide completion. We present multiple lines of direct and indirect evidence showing that both an increased prenatal androgen load (with subsequent permanent neuroadaptations) and increased adult androgen activity are involved in suicide completion. We also review data arguing that modifiable maternal behavioral traits during pregnancy contribute to the offspring's prenatal androgen load and increase the risk for suicide completion later in life. We conclude that in utero androgen exposure and adult androgen levels facilitate suicide completion in an synergistic manner. The androgen model of suicide completion provides the basis for the development of novel predictive and preventive strategies in the future.

Can risk assessment predict suicide in secondary mental healthcare? Findings from the South London and Maudsley NHS Foundation Trust Biomedical Research Centre (SLaM BRC) Case Register

PURPOSE

The predictive value of suicide risk assessment in secondary mental healthcare remains unclear. This study aimed to investigate the extent to which clinical risk assessment ratings can predict suicide among people receiving secondary mental healthcare.

METHODS

Retrospective inception cohort study (n = 13,758) from the South London and Maudsley NHS Foundation Trust (SLaM) (London, UK) linked with national mortality data (n = 81 suicides). Cox regression models assessed survival from the last suicide risk assessment and ROC curves evaluated the performance of risk assessment total scores.

RESULTS

Hopelessness (RR = 2.24, 95% CI 1.05-4.80, p = 0.037) and having a significant loss (RR = 1.91, 95% CI 1.03-3.55, p = 0.041) were significantly associated with suicide in the multivariable Cox regression models. However, screening statistics for the best cut-off point (4-5) of the risk assessment total score were: sensitivity 0.65 (95% CI 0.54-0.76), specificity 0.62 (95% CI 0.62-0.63), positive predictive value 0.01 (95% CI 0.01-0.01) and negative predictive value 0.99 (95% CI 0.99-1.00).

CONCLUSIONS

Although suicide was linked with hopelessness and having a significant loss, risk assessment performed poorly to predict such an uncommon outcome in a large case register of patients receiving secondary mental healthcare.