Risk assessment in the field of mental health

Most areas of work these days necessitate the use of audits, targets, protocols, policies and checklists. In fact, many would consider that the administration and oversight of the work they do frustratingly dominates their duties. This is especially so in work that involves the provision of care, building of relationships and empathic interaction with others. With so much documentation required to be completed by healthcare workers and those in “caring” professions it is easy to see how the burden of paperwork can sometimes appear to get in the way of the human side of their roles and the focus of day to day contact with patients and service users in what needs to be a holistic manner.

It is of course vital for healthcare providers to be able to produce evidence of good governance. The ability to present documentation to support the clear “line of sight” from ward to board is an absolute regulatory requirement and the subject of significant audits and reporting. The CQC regulatory functions are a good example. The CQC monitors services and grades them with colour coded ratings of Outstanding, Good, Requires Improvement and Inadequate. We have recently seen in the media calls to scrap the same rating process for Ofsted as it is thought to be too binary. But the tick boxes that must be used to make these evaluations have evolved over time as have the data-gathering processes used to ensure that adherence to fundamental principles in caring professions are adhered to and that there is a generic approach to data collection.

Those in the healthcare professions have long ago absorbed the mantra “if it is not documented it didn’t happen” and when giving evidence regarding complex processes in a court scenario (for example) it is extremely useful to be able to produce checklists, rag rated action plans and tick boxed questionnaires to show how the distillation of complex information is simplified and disseminated. It is also a vital source of data-gathering and it is comforting for professionals to be able to turn to policy and guidance for a generic approach to work. Nevertheless, it is important to remember that this documentation is only part of the picture when providing person-centred care. Risk assessments are a crucial part of ensuring that vulnerable people are kept as safe as possible whilst affording as much autonomy within those boundaries as is feasible.

This is particularly important in mental healthcare settings where capacity and risk often fluctuates. The balance between keeping a person safe whilst at the same time helping them to function within the community or a care setting is one of the biggest challenges mental health practitioners face. It is simply not possible to allow risk-taking as part of someone’s road to recovery whilst at the same time keeping that person safe at all times.

There has long been a desire to move away from slavish risk assessments as, of course, we know that they are only evidence of what is documented at any particular point in time and risk can change. For years the phrase “dynamic” has been attached to the assessment of risk to illustrate the difficulties experienced in changing circumstances. If it is difficult to assess the risk of an elderly person admitted to hospital who needs a level of autonomy, but might fall due to medication changes, how much more difficult to assess the risk of a person with mental health problems, looked after in the community by mental healthcare professionals who cannot be there at all times and have to rely on what is being said to assist them in their risk assessment judgment-making. There will inevitably be the application of years of experience and instinct as well as reference to generic guidance.

Recently a Healthcare Safety Investigations Branch Report (HSIB) warned of a need to change suicide risk assessments carried out by community mental health teams after the very sad death of Frances Wellburn who died by suicide after being discharged from hospital on two occasions into community teams. In their review of her care, HSIB raised concerns about the use of tick box exercises in determining risk and noted that while the categorisation of risk as high, medium or low was in line with the trust policy at the time, it did not sit with the NICE guidelines.

At the inquest into Mrs Wellburn’s death, the coroner noted that there were gaps in her care which reflected the Office for National Statistic’s own findings where concerns relating to assessment and clinical judgement were raised in more than a third of reports examined. Despite those identified gaps, the coroner said that it was “not possible to say whether they caused or contributed to her death”.

When considering risk prior to her death, a consultant had assessed Ms Wellburn as being in the category of “low risk”. She was engaging with services, concordant with medication and following advice given to her. We do not know how the determination of low risk was reached in this case, but these factors clearly influenced the consultant’s opinion. There was no specific comment in the coroner’s findings about the use of “tick-box” exercises. However, the national investigator for HSIB, Amber Sargent, concluded that it is absolutely crucial that patients at risk of suicide and self-harm are given a holistic assessment that moves away from tick boxes.

It would appear that in this case there was too much reliance upon the categorisation of risk within the high, medium or low risk categories. We do not know what the algorithms were that were used to determine that level of risk. It is interesting to consider the desire to move away from tick boxes when some services rely so heavily upon them. Perhaps it is the sophistication of the algorithm that should be the focus? Consider, for example, the triaging of emergencies by the ambulance service to determine risk in a multitude of different presentations. There must be a flow chart for call handlers to follow and a proper record of how decision-making has been reached. That approach clearly does not sit well within the context of risk assessment for mental health services.

Compare the HSIB criticisms in the very tragic case of Mrs Wellburn with a recent Report for the Prevention of Future Deaths issued in relation to a mental health trust working within a prison setting, where there appeared to be a lack of a formulaic (and it was assumed, therefore, coherent) approach to risk contained within the documentation. In that case risk was evidenced as being assessed within case notes but there was no determination of level (high, medium or low). It was considered that this meant the rationale for determination of risk was difficult to identify when reviewed by other clinicians as it did not directly correlate to a specific level of risk.

Naturally, there must be clear documentation and record-keeping around the determination for risk assessment. Busy healthcare professionals must strike a balance between taking a holistic approach, recording their findings and giving a clear rationale for the level of risk determined. Tick boxes have a place to act as an aide memoire only.

The NICE Guidance “Self-harm: assessment, management and preventing recurrence” [NG225] contains very useful information regarding the assessment of risk. It made recommendations stating that “the committee agreed that risk assessment tools and scales cannot accurately predict risk of self-harm and suicide, and that determining access to treatment or hospital admission based on inaccurate risk assessment tools could lead to repeat self-harm, distress and lower patient satisfaction it goes on to advise: The potential harms of risk stratification, including the implication that risk is static instead of dynamic, outweigh any benefits it has as a clinical communication tool or an adjunct to clinical assessment…… and safety should be a part of every assessment and that risk should not be used to determine care management in isolation of other factors…”

It is clear, then, that individual clinical judgement is vital, tools to assist in risk assessment are useful but there should be a move away from categorisation. It will also always be necessary to be able to record information in a manner that can be audited and used for data analysis. The need for clear documentation including rationale for decision-making, however, remains an absolute priority.