The DOI matches its Crossref record; 64 of the 65 references it cites were found in a registry or an index; 34 of the 58 sources its claims cite were checked for support.
Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior
Open accessAbstract
Risk assessment in the context of suicidal ideation and behavior cannot be conducted with certainty. Recent meta-analyses demonstrate that neither individual risk factors, composite risk scores, clinical judgment, nor adherence to theoretical models or artificial intelligence enables sufficiently accurate prediction of suicidal behavior. This raises the question of how clinicians should respond to the well-documented limitations in the precision of risk assessment. This article first reviews the current state of empirical evidence and subsequently reflects on the implications of these findings for clinical practice. Three alternative approaches to risk assessment are presented. Building on these, an integrated model for risk assessment is introduced in a practice-oriented manner.
Overview
CoChat Summary
TL;DR
This narrative review argues that suicide risk prediction is not clinically reliable at the individual level: neither risk factors, composite scores, clinical judgment, theoretical models, nor AI can predict suicidal behavior with sufficient accuracy. It proposes shifting practice from prediction-based stratification to collaborative, needs-based, severity-based, and capacity-based assessment.
Key findings
- Franklin et al.’s meta-analysis synthesized 365 longitudinal studies and 16 types of risk factors; it found that prediction of suicide attempts and suicides has not improved in the past 50 years and is only marginally better than random chance.
- Large et al.’s meta-analysis of 37 longitudinal studies found that among psychiatric patients categorized as high risk, 5.5% died by suicide over an average 63-month follow-up, but 95% of those classified high risk did not die by suicide; sensitivity analyses showed 44% of suicide decedents were not assigned to the high-risk group.
- Woodford et al. found that 70% of patients who engaged in self-injurious behavior after discharge had previously been judged not at risk, and about 80% of those judged high risk did not engage in self-injurious behavior.
- The paper states that in one cited cohort of more than 80,000 people, suicidal thoughts “almost every day” were associated with a tenfold higher risk, but absolute risk remained low: 0.03% vs 0.3%; it also notes that 99.7% of those with daily suicidal thoughts did not die by suicide.
Methods
- Narrative review and theoretical synthesis of meta-analyses, clinical guidelines, and influential longitudinal studies on suicide prediction and suicidality assessment; the authors explicitly state this was not a formal systematic review.
- The paper integrates evidence across individual risk factors, clinical judgment, theoretical models, AI, and observational studies to argue that prediction-based assessment fails at the individual level.
- It also proposes a practice-oriented integrated framework combining needs-based, severity-based, and capacity-based approaches.
Limitations
- The paper states its own analysis is based on a narrative and theoretical synthesis rather than a formal systematic review.
- It states that the proposed clinical alternatives are supported by varying levels of empirical evidence.
- It notes that the associated legal implications may differ across jurisdictions.
- Evident from the text: some quantitative examples are explicitly modeled rather than empirical results (for example, the time-horizon illustration), so those estimates should not be treated as direct study findings.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
24 Valid · 0 Not valid · 16 Not checked · 33 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“At the same time, however, this implies that 95% of those classified as high risk did not die by suicide”
§ Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior > Empirical findings
What the sources say
- Supportedcites [17] Large, 2016, Meta-analysis of longitudinal cohort studies of suicide risk assessment among psychiatric patients, PloS One source
The Discussion explicitly says the meta-analysis suggests that 95% of high-risk patients will not die by suicide. The Results report a pooled suicide rate of 5.5% among high-risk patients, consistent with that rounded figure. (LLM, conf=0.95)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Decades of epidemiological and clinical research have indeed established a reasonably consistent inventory of variables associated with elevated suicide risk at the group level.”
§ Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior > Introduction
What the sources say
- Not checked — no full text of the sourcecites [4] Crump, 2014, Sociodemographic, psychiatric and somatic risk factors for suicide: a Swedish national cohort study, Psychol Med source
- Supportedcites [5] Favril, 2022, Risk factors for suicide in adults: systematic review and meta-analysis of psychological autopsy studies, Evid Based Ment Health source
The Findings report a synthesis of 37 case–control studies identifying 40 suicide risk factors, and the Clinical implications section describes associations across clinical, life-event, family-history and sociodemographic domains. This supports the claim about an established inventory of group-level risk factors; the paper also cautions that association strengths vary substantially. (LLM, conf=0.88)
- Partly supportedcites [6] Grover, 2023, Meta-analysis of clinical risk factors for suicide among people presenting to emergency departments and general hospitals with suicidal thoughts and behaviors, Acta Psychiatr Scand sourceExperimental — accuracy not yet measured
Study selection (§3.1; page not available) reports studies spanning 1965–2022 and identifies clinical risk factors examined across multiple studies. Statistical strength (§3.3; page not available) lists factors associated with later suicide at the group level. But the evidence concerns people presenting to emergency departments and general hospitals with suicidal thoughts or behaviours, and the paper reports high between-study heterogeneity; it does not establish a reasonably consistent inventory for populations generally. (LLM, conf=0.82)
- Partly supportedcites [7] Ribeiro, 2018, Depression and hopelessness as risk factors for suicide ideation, attempts and death, Br J Psychiatry sourceExperimental — accuracy not yet measured
The Background (page not supplied) says depression and hopelessness are commonly cited risk factors, and the Main findings report that both confer risk for suicidal ideation, attempts and death at the group level. But the paper examines those factors, not a broader inventory of risk variables, and its Background notes inconsistent findings across studies. (LLM, conf=0.84)
- Partly supportedcites [8] Wiebenga, 2022, Associations of physiological stress systems with suicidality, Brain Behav Immun sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) identifies depression, anxiety, and comorbidity as group-level suicide-risk factors; the Clinical Characteristics section also reports differences between suicidal and non-suicidal groups. The paper does not establish that decades of research have produced a reasonably consistent inventory of risk variables, so that broader characterization is unsupported. (LLM, conf=0.82)
boostedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“and, in particular, they do not allow for prediction of the time at which suicidal behavior will occur.”
§ Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior > Empirical findings > Attempts to improve suicide risk assessment
What the sources say
- Partly supportedcites [26] Kirtley, 2022, Translating promise into practice: a review of machine learning in suicide research and prevention, Lancet Psychiatry sourceExperimental — accuracy not yet measured
The section “Scalability of capacity for detection, but not intervention” (page unavailable) says machine-learning research has focused on predicting outcomes months or years ahead and calls for more short-term prediction. That supports a limitation in current research, but not the categorical claim that AI models cannot predict when suicidal behavior will occur; the paper also describes models aiming to predict risk within hours. (LLM, conf=0.84)
- Not checked — no full text of the sourcecites [27] Belsher, 2019, Prediction models for suicide attempts and deaths: a systematic review and simulation, JAMA Psychiatry source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“and that around 80% of those who were assessed as being at high risk did not engage in self-injurious behavior.”
§ Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior > Empirical findings
What the sources say
- Supportedcites [18] Woodford, 2019, Accuracy of clinician predictions of future self-harm, Suicide Life Threat Behav source
The Conclusions (p. 15) report a pooled positive predictive value of 0.22 for clinicians’ high-risk classification for repeat self-harm. Thus, nearly 80% of those classified as high risk did not repeat self-harm. (LLM, conf=0.94)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Taken together, the predictions made by practitioners in this meta-analysis were only slightly better than chance.”
§ Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior > Empirical findings
What the sources say
- Supportedcites [18] Woodford, 2019, Accuracy of clinician predictions of future self-harm, Suicide Life Threat Behav source
The Discussion (p. 10) describes clinician predictions of future self-harm as a “minimal improvement in prediction over the expected rate.” The Discussion (p. 13) likewise says high-risk classifications offer no meaningful improvement over patients’ pre-test probability. These passages support the citing sentence’s characterization of the predictions as only slightly better than chance. (LLM, conf=0.90)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
46 more claims · 14 with their cited sources checked
Create a free account to see all 73 claims and every source verdict.
- It may state two separate claims as one. 13
- Probably not a substantial claim. 10
Citations
Cites
Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research
Franklin, J. C., Ribeiro, J. D., Fox, K. R., Bentley, K. H., Kleiman, E. M., Huang, X., Musacchio, K. M., Jaroszewski, A. C., Chang, B. P., & Nock, M. K. (2017) · Psychological Bulletin · doi:10.1037/bul0000084 · Cited by 3,784 · cited 2×
“Various studies have demonstrated that the reliability of risk assessments is moderate at best (9), and that the prediction of suicidal behavior, is currently insuf fi cient for clinically useful individual-level prediction (10).”
The integrated motivational–volitional model of suicidal behaviour
O'Connor, R. C., & Kirtley, O. J. (2018) · Philosophical Transactions of the Royal Society B: Biological Sciences · doi:10.1098/rstb.2017.0268 · Cited by 1,781
Assessing patients' capacities to consent to treatment
Appelbaum, P. S., & Grisso, T. (1988) · New England Journal of Medicine · doi:10.1056/nejm198812223192504 · Cited by 1,399 · cited 1×
“Finally, expressing a choice refers to the ability to communicate a clear and consistent decision regarding treatment and safety (58).”
Chu, C., Buchman-Schmitt, J. M., Stanley, I. H., Hom, M. A., Tucker, R. P., Hagan, C. R., Rogers, M. L., Podlogar, M. C., Chiurliza, B., Ringer, F. B., et al. (2017) · Psychological Bulletin · doi:10.1037/bul0000123 · Cited by 1,231 · cited 1×
“The effects are, however, comparatively small (22, 23) and do not exceed the predictive power found in the other meta-analyses described above.”
Ribeiro, J. D., Huang, X., Fox, K. R., & Franklin, J. C. (2018) · British Journal of Psychiatry · doi:10.1192/bjp.2018.27 · Cited by 894
School-based suicide prevention programmes: the SEYLE cluster-randomised, controlled trial
Wasserman, D., Hoven, C. W., Wasserman, C., Wall, M., Eisenberg, R., Hadlaczky, G., Kelleher, I., Sarchiapone, M., Apter, A., Balazs, J., et al. (2015) · The Lancet · doi:10.1016/s0140-6736(14)61213-7 · Cited by 781
Prediction of Suicide in Psychiatric Patients: Report of a Prospective Study
Pokorny, A. D. (1983) · Archives of General Psychiatry · doi:10.1001/archpsyc.1983.01790030019002 · Cited by 759
Suicide Rates After Discharge From Psychiatric Facilities: A Systematic Review and Meta-analysis
Chung, D. T., Ryan, C. J., Hadzi-Pavlovic, D., Singh, S. P., Stanton, C., & Large, M. M. (2017) · JAMA Psychiatry · doi:10.1001/jamapsychiatry.2017.1044 · Cited by 651 · cited 1×
“The fi gures shown are not empirical results, but a modeled example based on published post-discharge suicide rates (30, 31) combined with the sensitivity and speci fi city estimates reported by Large et al. (17); the underlying calculations are summarized in the table note.”
Stanley, B., Brown, G. K., Brenner, L. A., Galfalvy, H. C., Currier, G. W., Knox, K. L., Chaudhury, S. R., Bush, A. L., & Green, K. L. (2018) · JAMA Psychiatry · doi:10.1001/jamapsychiatry.2018.1776 · Cited by 648
Simon, G. E., Rutter, C. M., Peterson, D., Oliver, M., Whiteside, U., Operskalski, B., & Ludman, E. J. (2013) · Psychiatric Services · doi:10.1176/appi.ps.201200587 · Cited by 575 · cited 1×
“Bryan (11) illustrates this with a model calculation based on a study in which more than 80,000 people were examined over a period of four years (32).”
Research works printed but not linked: 10
- Teismann, 2022, Therapie-tools: suizidalität und krisenintervention
- Rudd, 2001, Treating suicidal behavior: an effective time-limited approach
- Meehl, 1954, Clinical versus statistical prediction
- Bryan, 2018, Brief cognitive-behavioral therapy for suicide prevention
- Gysin-Maillart, 2021, ASSIP
- Teismann, 2024, Suizid-screenings: eine kritik, Suizidprophylaxe, 51, 34
- Jobes, 2023, Managing suicidal risk: a collaborative approach
- 2022, Unipolare depression. S3-leitlinie/nationale versorgungsLeitlinie unipolare depression – kurzfassung
- Shea, 2011, The practical art of suicide assessment
- Bryan, 2010, Managing suicide risk in primary care
Cited by
E., D. F. (2026) · African Journal of Health Nursing and Midwifery · doi:10.52589/ajhnm-9sj6e7te
Ciobanu, A. M., Tebeanu, A. V., Tau, V., Chendea, A. A., Franti, E. D., Niculae, C., Vasile, C. I., Macarie, G. F., Neagu, L., Dascălu, M., et al. (2026) · Psychiatry International · doi:10.3390/psychiatryint7040174 · Cited by 1
Metadata
- Authors
- Teismann, T., Janssen, W. C., & Heering, H. D.
- Affiliations
- Deutsches Zentrum für Psychische Gesundheit (DE)
- Amsterdam UMC Location University of Amsterdam (NL)
- GGZ inGeest (NL)
- Corresponding author
- Teismann, Tobias
- Venue
- Frontiers in Psychiatry
- Publisher
- Frontiers Media SA
- Published
- 2026-06-02
- Volume
- 17
- Pages
- 1844322-1844322
- ISSN
- 1664-0640 (ISSN-L)
- PMC
- PMC13270292
- OpenAlex
- W7163211912
- Cited by
- 3
- References
- 62
- Citations by year
- 2026: 3
- FWCI
- 24.03
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Frontiers in Psychiatry (publishedVersion, cc-by)
- PubMed (publishedVersion)
- PubMed Central (publishedVersion, cc-by)
- Pure Amsterdam UMC (acceptedVersion, other-oa)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Funders
- Ruhr-Universität Bochum
- Topics
- Social Sciences › Psychology › Clinical Psychology › Suicide and Self-Harm Studies
- Social Sciences › Psychology › Clinical Psychology › Psychopathy, Forensic Psychiatry, Sexual Offending
- Social Sciences › Psychology › Social Psychology › Mental Health via Writing
- Keywords
- suicide risk assessment, suicidal ideation, suicidal behavior, risk prediction accuracy, composite risk score, clinical judgment, clinical practice implications
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior. https://app.cochat.ai/papers/10.3389/fpsyt.2026.1844322
- The paper
- Teismann, T., Janssen, W. C., & Heering, H. D. (2026). Suicide risk assessment: clinical implications of the unpredictability of suicidal behavior. Frontiers in Psychiatry. https://doi.org/10.3389/fpsyt.2026.1844322
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.
