Improving awareness of the HCR-20 and risk assessment process

1. Background

In 2011, Ms A attacked Ms B at a bus stop, which resulted in Ms B being stabbed in the hand. Following this, Ms B took a knife from a shop and fatally stabbed Ms C. In the hours prior to the attack, Ms A made numerous attempts to seek help. This included contact with the police, two nurses at Accident and Emergency, and staff at an inpatient service. Subsequent investigation highlighted that Ms A’s risk of violence in the community was not appropriately considered by her care team, nor were the factors associated with increments in risk appropriately documented [1].

One of the critical duties of forensic mental health teams concerns assessing service-users’ current and future risks, to implement appropriate risk management strategies and protect the public. The Historical Clinical Risk Management-20 Version 3 (HCR-20) is the most commonly used forensic risk assessment in the UK [2] and evaluates an individual’s risk of violence, the relevance of the risk to their current situation and the appropriate risk management strategies [3]. Table 1 details the items included on the HCR-20.

The HCR-20 has been demonstrated to accurately predict violent behaviour both during treatment [4,5] and subsequent to discharge [6,7] in individuals with psychological difficulties. Ratings on this assessment have also been demonstrated to predict readmission to a psychiatric ward, following discharge from medium-secure facilities [8].

Multidisciplinary involvement in implementing the HCR-20 has been demonstrated to be effective in risk management, highlighting the importance of communicating these assessments to mental health teams and integrating them into practice [2]. Nevertheless, despite evidence that multidisciplinary risk management strategies can reduce the risk of reoffending [9], research aimed at developing guidance on the communication and integration of the HCR-20 into clinical practice is largely absent.

The purpose of this paper can be summarised by the following statement by Haque and Webster “A risk management plan is only as good as the time and effort put into communicating its findings to others” [10]. Whilst the literature supports the validity and reliability of the HCR-20 as a measure of risk, there are serious problems concerning the communication of this information both within and between mental health teams.

3. Results

The interventions resulted in an increased awareness of the existence, purpose and location of the HCR-20. As shown in Fig. 1, the percentage of staff members that had seen a completed HCR-20 significantly increased after training. Indeed, paired samples t tests indicated that the intervention resulted in a statistically significant increase in both the percentage of people that had heard of the HCR-20,p (6) =   −5.523, p = 0.001, and that had seen an HCR-20, p (6) =   −4.950, p = 0.003. Participants noted that HCR-20 training was particularly effective in increasing knowledge and awareness of these risk assessments.

OPEN IN VIEWER

Fig. 1.

Percentage of staff that had seen a completed HCR-20 across all forensic wards.

The interventions produced a sustained reduction in HCR-20 breaches by 73%. As shown in Fig. 2, the HCR-20 breaches across all forensic wards declined subsequent to the induction of the interventions, indicating that the interventions maintained an influence over HCR-20 breaches.

OPEN IN VIEWER

Fig. 2.

HCR-20 breaches across all forensic wards.

Overall, this QI project indicated that including the ‘risk formulation’ and ‘scenarios’ from the HCR-20 in clients’ crisis plans, adding a tab on the summary page of a client’s Rio record, introducing training relating to the HCR-20, monitoring breaches of HCR-20 reviews and including discussions relating to the HCR-20 at the beginning of CPA meetings resulted in improved MDT awareness and knowledge of the location of the HCR-20 on Rio and reduced HCR-20 breaches. A broader understanding and awareness of risk factors enabled the service to move towards a culture of risk being everyone’s business. This approach is likely to lead to more effective management of risk, compared to it being held in isolation by one or two professionals (usually psychology and psychiatry). Adding a tab of the HCR-20 onto the main summary page enables others (such as Community Mental Health Teams or accident and emergency services) to quickly identify someone’s risk history. These results provide a critical insight into the initiatives that can be used to increase knowledge and awareness of HCR-20s, which will support mental health professionals in making risk-informed clinical decisions.

Some of the interventions were more effective than others. Psychologists reported that copying the ‘risk formulation’ and ‘scenarios’ from the HCR-20 into clients’ crisis plans was a time consuming and ineffective task, as other staff would delete the information from Rio. Additionally, as the risk formulation in crisis plans is generally updated more regularly than the HCR-20, copying the formulation from the HCR-20 may result in out-dated information. Therefore, the team worked with the IT department to develop a system that would automatically update the formulation, negating the need to manually update this. This intervention led to the removal of human error, and saved significant time resources on clinicians.

Conflict of interest

There were no known conflicts of interest in the completion of this quality improvement project.