Re‐focusing risk assessment in forensic mental health nursing

Violence risk assessment has long been held to provide one of the unique characteristics of the specialist mental health nursing role in secure and forensic settings. Nurses have contributed considerably to the development of evidence-based structured tools to guide risk assessment, but the most widely used instruments have been authored by psychologists (Khiroya et al. 2009). Although forensic nurses have also historically recognized the need to incorporate violence risk assessment within an overall therapeutic approach, some authorities believe that it has focused so much on identifying the individual patient's deficits that it may be detrimental to the development of the therapeutic relationships that lie at the heart of mental health nursing practice (Rogers 2000). This editorial examines recent developments that have the potential to facilitate more rounded, holistic and clinically relevant assessments: notably the incorporation of protective factors into risk assessment, and the consideration of a wider range of outcomes in addition to violence. The latter in particular also holds the potential to make risk assessment more relevant beyond the forensic setting where many people with mental disorder are victims of violence, but only a small minority are perpetrators (Desmarais et al. 2014). To date, there are a limited number of tools that have been developed to facilitate the application of a broader and more holistic approach to risk assessment. One used reasonably widely in forensic services internationally is the Short-Term Assessment of Risk and Treatability (START; Webster et al. 2004). Developed in Canada, the START comprises 20 dynamic items and is intended to be repeated every 90 days. Raters are encouraged to consider patients' strengths in equal proportion to their vulnerabilities in relation to each item before making an overall estimate of risk for a range of adverse outcomes: violence, self-harm and suicidality, self-neglect, victimization, substance abuse and unauthorized leave. The START is highly compatible with notions of person-centredness and recovery-oriented practice. The development team for the START included a psychiatric nurse, and the tool is intended to be rated by a multidisciplinary team, an explicit acknowledgment that different professions bring their own specialist fields of knowledge to assessment. The tool is further nuanced by inclusion of instructions to give special consideration to items that are of particular relevance to the individual patient, and it is flexible because it allows the addition of patient-specific risk factors not included in the 20-item scheme. It is not uncommon for nurses to opine that risk management approaches amount to little more than a defensive barrier for clinicians against repercussions in the event of adverse outcomes (Manuel & Crowe, 2014). Others see risk assessment as, at best, a tick-box exercise or, worse, a technocratic attempt to replace clinical expertise and dictate practice. I believe that this would be an unfair criticism directed at the START. The tool is an example of structured professional judgement in that it aims to serve as an aide memoire to clinical assessment and not as a set of hard and fast rules to produce a score on the basis of which decisions about management are made. But what is the evidence? We recently synthesized the available empirical research about the START (O'Shea & Dickens 2014). Our review revealed that START is rated by practitioners as having good utility and ease of use. Additionally, it has good psychometric properties including predictive validity for violence and self-harm. However, evidence is more scant for its ability to predict non-violent outcomes such as victimization. The START has been implemented across all services at St Andrew's, a large UK charity providing secure and forensic care for around 800 people. Using a sample of data from assessments conducted in clinical practice, we have replicated and extended findings from our review. Most importantly, we found that consideration of patient's strengths provides a more accurate indication of the likelihood of them engaging in violence than does focusing on their vulnerabilities; and that there is now more evidence that the tool is significantly predictive of self-harm, victimization, substance abuse and unauthorized leave (O'Shea & Dickens 2015, O'Shea et al. 2015). Furthermore, the overall formulation that rating teams make about patients' risks based on all the START components provides a better indication of likelihood of being aggressive than does consideration of the 20 risk factors alone. Taken together, these findings suggest that the START is a promising tool that can be used to assess a limited range of non-violent outcomes, and that consideration of strengths has demonstrable value as well as being simply the right thing to do. Clearly, the START is not a panacea but a first attempt at broadening risk assessment to include routine consideration of patient's strengths; and to shift the focus purely from violence prediction, it holds considerable promise. Logically, one would expect the patient's experience to be improved if the multidisciplinary team is focused on a holistic and rounded view, and one might expect better outcomes given the focus on a wide range of outcomes. Of course, little in life is that simple, and there remains a need to establish whether this is the case by conducting well-designed research trials. Further, there is a need to better establish the involvement, suitability and advantages of further incorporating the patient's views into the assessment process. Implementation across a large mental health service demonstrates that the START can be incorporated into routine practice and is not an esoteric ‘research-only' instrument. Users of secure and forensic services should, of course, expect that the tools used to inform decisions about their management are supported by evidence. We have shown that large-scale implementation is feasible when there is organizational commitment; further, with investment in infrastructure and training, implementation can also facilitate large-scale research projects to produce this evidence.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.