Suicide is a major public health concern, which claims nearly 800,000 lives worldwide each year (World Health Organization, 2016). In the UK only, it accounts for approximately 6,000 deaths annually (Office for National Statistics, 2019). The assessment of suicide risk thereby constitutes an integral part of clinical practice in psychiatry, but approaches to the management of risk vary considerably between jurisdictions.

One of the main debates in suicide research relates to structured risk assessment tools; in terms of their effectiveness, and as to whether they should be incorporated in risk assessment as adjuncts to clinical judgement. Whilst their use is recommended in the US (Office of the Surgeon General (US) & National Action Alliance for Suicide Prevention (US), 2012), current British (National Institute for Health and Care Excellence, 2011), Australian and New Zealand clinical practice guidelines state otherwise (Carter et al., 2016). They suggest that such tools should not be used for decision-making in regards of treatment and aftercare allocation. Proponents have argued that potential benefits include:

• Providing a framework for baseline risk assessments (Fazel, Wolf, Larsson, Mallett, & Fanshawe, 2019),
• Means of communication amongst healthcare teams, and between clinicians and service users (Graney et al., 2020),
• Short-term risk identification (Cooper et al., 2006),
• Implementation that requires minimal training (Cooper et al., 2006).

However, research has found that most of the currently used tools are limited by poor predictive value (Chan et al., 2016). This does not come as a surprise when considering that the prediction of any low prevalence health-related event (e.g. suicide) will inherently be limited by a small positive predictive value (PPV). Thus, individuals may be wrongly classified as ‘high-risk,’ resulting in unnecessary treatment allocation, and ultimately increase clinician workload (Bolton, Gunnell, & Turecki, 2015).

To investigate the current state of evidence on suicide risk assessment tools in the UK, Graney and colleagues (2020) conducted a national mixed methods study that spanned all mental health organisations. The aims of their study were three-fold:

1. Provide an up-to-date inventory of all suicide-related risk assessment tools
2. Analyse the perspectives of service users, carers and clinicians on these tools
3. Determine the ways in which these tools were used with mental health patients before suicide.

Methods

Three distinct sets of methods were employed to achieve these aims.

1. First, the authors collated information about clinical risk assessment from all mental health organisations in the UK. This included any risk assessment tools or scales used when conducting the assessment. Common themes in respect of the content and structure of these tools, and clinical/sociodemographic information of service users, were subsequently identified through content analysis.
2. Second, an online survey was conducted using convenience sampling amongst clinicians, service users and carers to collect in-depth information on the risk assessment process from various perspectives.
3. Third, semi-structured telephone interviews were carried out to gain additional insights from clinicians who had experienced a patient suicide in the recent years according to the NCISH database.

When applicable, for all three components, descriptive statistics and thematic techniques were employed to summarise quantitative and qualitative data respectively.

Results

Key features of the risk assessment tools

In total, 156 risk assessment tools were identified in all 85 NHS mental health trusts and health boards in the UK. Of these, 85 instruments were selected for the analysis. The main descriptive characteristics of these tools included:

• Type (three most used)
  • Locally devised ones (39%)
  • Rio Risk Screen (20%)
  • FACE (9%)
• Format
  • Checklist style (85%)
  • Formulation-based (15%)
• Scoring system
  • Categorical stratification
    • Low/medium/high (56%)
    • Red/amber/green (7%)
  • Numeric (15%)
• Content included information relating to:
  • Self-harm and suicide (99%)
  • Patient engagement in their care (59%)
  • Suicidal intent (47%)
  • GP consultation (16%)

Views of service users, carers and clinicians about risk assessment tools

In total, 358 individuals completed the online survey, but the responses of 15 patients and 20 carers had to be removed due to missing data. The final sample thus included 323 participants (262 clinicians, 27 patients and 20 carers) that provided their perspective on their encounters with risk assessment tools.

The authors identified three overarching themes within the clinicians’ narratives, which also relate to the perspectives of service users and carers:

1. Personal interaction and communication
2. Training
3. Challenges of using risk assessments.

Interestingly, there was some divergence of opinion regarding these themes. In fact, whilst the vast majority of clinicians (90%) reported using an assessment tool during their consultations, only a third of patients (33%) were aware of the latter being utilised. Moreover, clinicians viewed risk assessment tools as beneficial for establishing a collaborative therapeutic alliance. However, some service users and carers found the assessment to be impersonal, and others emphasised the lack of opportunity to express their personal views and concerns.

Clinician interviews

Two main themes emerged from the phone interviews that were conducted with 22 clinicians: the role tools played in their risk assessments, and their potential benefits for facilitating communication. Clinicians were unanimous in using the tools as a general framework to guide their assessment, rather than relying on them for predictive purposes. They also reported that the tools could simplify information sharing amongst healthcare professionals by providing a brief risk summary, as patients move between various mental healthcare services.

Conclusions

The findings of this study underscore the lack of uniformity in risk assessments conducted in mental health organisations across the UK. The authors emphasised:

Risk assessment alone should not be used as a means of allocating treatment. […] Treatment plans must be personalised and collaboratively developed. But, perhaps, we also need services to move away from prediction-based paradigms of providing mental health intervention and focus on safer evidence-based care for all, including improved communication with families and carers.

Strengths and limitations

The study represents the first evidence-based account of risk assessments conducted in all mental organisations across the UK. The authors present a rich body of empirical data, gathered through three distinct methodological approaches. As a result, the authors go beyond previous studies on risk assessments, exemplified by their spotlight on the views of patients and carers; essential voices that tend to be overlooked within contemporary academic debates. In other words, the study puts forward the necessity of a nuanced stance towards prediction-based paradigms, thereby setting the stage for future research that builds on its empirical and conceptual contributions.

Despite these merits, it is paramount to highlight two weaknesses. The authors acknowledge that the survey samples were small and selected. However, they do not mention nor discuss the potential attrition bias that might have arisen as a result of excluding a considerable number of services users and carers on the basis of missing information. Moreover, the study does not engage with comparative information about the predictive performance of the identified risk assessment tools (e.g., area under the curve [AUC]). In the light of this omission, future research should take into account the tools’ discrimination and calibration measures, if those are available, which are likely to inform the extent to which tools can be incorporated into clinical practice. Details about the derivation process and whether the risk prediction tool has been externally validated should also be considered (Fazel & Wolf, 2017).

Implications for practice

• Practically, the study underlines the eminent importance for healthcare professionals to deploy a person-centred approach when conducting clinical risk assessments, which includes explicitly addressing suicidal thoughts of patients.
• Risk assessment tools are likely to gain further technical sophistication due to identification of modifiable risk factors and development of novel approaches, such as machine learning.
• Still, they will not substitute a clinician’s ability to evaluate suicide risk and provide personalised interventions, which form key pillars of evidence-based mental healthcare.

Statement of interests

None.

Links

Primary paper

Graney J, Hunt IM, Quinlivan L, Rodway C, Turnbull P, Gianatsi M, Appleby L, Kapur, N. (2020). Suicide risk assessment in mental health services: a national mixed methods study from the UK. The Lancet Psychiatry, November 12, 2020. https://doi.org/10.1016/S2215-0366(20)30381-3

Anonymous – Commentary (2020) The reality of risk assessment. The Lancet Psychiatry, November 12, 2020. https://doi.org/10.1016/S2215-0366(20)30463-6

Other references

Bolton, J. M., Gunnell, D., & Turecki, G. (2015). Suicide risk assessment and intervention in people with mental illness. British Medical Journal, 351, h4978.

Carter, G., Page, A., Large, M., Hetrick, S., Milner, A. J., Bendit, N., . . . Christensen, H. (2016). Royal Australian and New Zealand College of Psychiatrists clinical practice guideline for the management of deliberate self-harm. Australian and New Zealand Journal of Psychiatry, 50(10), 939-1000.

Chan, M. K. Y., Bhatti, H., Meader, N., Stockton, S., Evans, J., O’Connor, R. C., . . . Kendall, T. (2016). Predicting suicide following self-harm: systematic review of risk factors and risk scales. British Journal of Psychiatry, 209(4), 277-283.

Cooper, J., Kapur, N., Dunning, J., Guthrie, E., Appleby, L., & Mackway-Jones, K. (2006). A Clinical Tool for Assessing Risk After Self-Harm. Annals of Emergency Medicine, 48(4), 459-466.

Fazel, S., & Wolf, A. (2017). Suicide risk assessment tools do not perform worse than clinical judgement. British Journal of Psychiatry, 211(3), 183-183.

Fazel, S., Wolf, A., Larsson, H., Mallett, S., & Fanshawe, T. R. (2019). The prediction of suicide in severe mental illness: development and validation of a clinical prediction rule (OxMIS). Translational Psychiatry, 9(1), 98.

National Institute for Health and Care Excellence. (2011). Self-harm in over 8s: long-term management. Clinical guidance.Retrieved from: https://www.nice.org.uk/guidance/cg133

Office for National Statistics. (2019). Suicide in the United Kingdom: 2018 registrations. Retrieved from: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/suicidesintheunitedkingdom/2018registrations

Office of the Surgeon General (US), & National Action Alliance for Suicide Prevention (US). (2012). 2012 national strategy for suicide prevention: goals and objectives for action. Retrieved from: https://www.ncbi.nlm.nih.gov/books/NBK109917/

World Health Organization. (2016). WHO suicide data. Retrieved from: http://www.who.int/mental_health/prevention/suicide/suicideprevent/en/

Photo credits

• Photo by Scott Graham on Unsplash
• Photo by Jake Hills on Unsplash
• Photo by Ricardo Rocha on Unsplash