Background to the Current Use of the HCR-20

UK Government papers (Department of Health 1996; Department of Health National Risk Management Programme, 2007) have recommended collaboration with the service user and others involved in their care whilst undertaking risk assessments. The latest consultation document (Department of Health, 2012a) which set standards for low secure and psychiatric intensive care units, also emphasises the importance of service-user involvement, outlining that service users should be placed at the centre of their care (‘nothing about me without me’) where this is appropriate. Patient engagement and participation in formulation is actively encouraged and the importance of incorporating risk assessment and management into a service user’s care plan is also documented. Despite these government guidelines, and consultation documents emphasising the importance of service-user involvement, little research has been undertaken to evaluate the frequency or quality of service-user involvement in risk assessment.

Langan and Lindow (Reference Langan and Lindow2004) are among the few who have undertaken research into the level of service-user involvement in violent risk assessment and management. They reviewed 17 service users who were being discharged from an inpatient unit into the community. Although there was evidence of service-user involvement, some of the service users were completely unaware that they were perceived to be a risk to other people. This raises the question of if service users are not aware of their risk, how can they begin to manage it? Langan and Lindow (Reference Langan and Lindow2004) found that some risk information was either absent or inaccurate leading to an under- or overestimation of risk. This distortion can have a detrimental effect on a service user’s care. Dixon (Reference Dixon2012) investigated service users’ awareness and attitudes towards risk assessments. Most participants were unaware that a risk assessment about them had been undertaken, although those who had been involved appeared to engage more with their care staff about risk.

As a result of research into service-user involvement in risk assessment and management on inpatient unit, Langan (Reference Langan2008) reported four key themes that arose:

1. 1. Some professionals lacked confidence or experience in openly discussing risk with service users. This open discussion was not considered to be routine practice by the professionals, and concern was noted that raising the issue of risk with a service user who is acutely unwell may be counter-productive. Despite this, the professionals considered openness to be a good guiding principle.

2. 2. Most staff members did not involve service users in risk assessments, although no staff rejected the principle of service-user involvement.

3. 3. The majority of service users were unaware that professionals were undertaking risk assessments on them. Those that were aware reported being indifferent or viewed it as a positive process.

4. 4. Levels of agreement between professionals and service users regarding risk and its seriousness differed. Professionals considered a high level of concordance between themselves and the service user in assessing risk to be an important aid in risk assessment and management. These themes provide a useful insight into the working practice of involving service users in their risk assessment.

Involving service users in the HCR-20 risk assessment process is likely to be empowering to patients. The Royal College of Psychiatrists (2008) found that patient involvement can substantially support the growth of resilience, confidence and self management in people using services. It promotes collaboration in their care plan and supports them in managing their own risk factors for violence. Henagulph et al. (Reference Henagulph, McIvor and Clarke2012) set up an HCR-20 risk and recovery group in a medium secure unit in London. Meetings are held weekly to facilitate patient discussion about the risk items and the score they warrant. It is the authors’ understanding that the ‘scoring’ gives structure to a discussion by group members about their risk (Clarke, Reference Clarke2012). However, it was acknowledged when this issue was debated at the Division of Clinical Psychology (DCP) Forensic Faculty meeting (April 2012) that scoring and use of a total score is not the normal procedure for the HCR-20. Nonetheless, it has been found that this programme promotes hope amongst patients as they progress from hospitalisation to the community. It provides service users with a better understanding of the relationship between mental illness and risk, provides improved understanding of the reasons for their continued detention, and focuses on goals relating to discharge. Ashcraft and Anthony (Reference Ashcraft and Anthony2008) endorsed patient involvement in a recovery oriented approach that can reduce the use of seclusion and restraint.

The HCR-20 has become a widely used risk assessment tool for assessing violence in a variety of settings and its use is recommended in numerous guidelines (Department of Health, 1996, 2012b; O’Rourke and Bailes, Reference O’Rourke and Bailes2006; Khiroya et al. Reference Khiroya, Weaver and Maden2009). The strength of this instrument lies in proper administration and coding of the HCR-20 which requires considerable professional skill and judgement. The HCR-20 instrument is comprised of the following sections: Historical (past, stable) factors; Clinical (current, dynamic) factors; Risk (future, dynamic) factors; Risk Scenarios (predicted future risky scenarios); and Case Management (summarising risk management plans).

The HCR-20 has become the preferred risk assessment instrument, used in conjunction with the RiO Risk Screen and Risk Information, within Somerset Partnership NHS Trust. RiO is an electronic patient record system used by mental health trusts and community service organisations. The RiO Risk Screen form contains ratings regarding service users’ long-term and acute risk to others (rated between zero: no apparent risk; and three: high risk); the Risk Information form contains detailed information regarding clients’ risk and forensic history. To facilitate use of the HCR-20 risk assessment instrument, an extensive two day training programme was rolled out to over 107 mental health staff across the Trust, including all qualified practitioners working in two newly developed inpatient wards: a psychiatric rehabilitation ward (a locked 10-bed, mixed-gender ward) and a low secure ward (12-bed, male ward).

The aim of this audit was two fold:

1. 1. To examine the quality of the HCR-20 assessments completed on the wards;

2. 2. To examine service-user involvement in HCR-20 assessments completed on the wards.

In keeping with the requirements of the audit cycle, the results were to be used for the development of standards for the use of the HCR-20 on these wards.

Method

All current and past service users admitted to the wards were reviewed. At the time of the audit, the wards had been open for just over two years. In this period, 11 service users had been admitted to the low secure ward, and 23 to the psychiatric rehabilitation ward, giving a total sample size of 34. Data was obtained for the most current RiO Risk Screen and Risk Information sections, and the most current HCR-20 report, if one had been completed. The process of completing an HCR-20 risk assessment was implemented when service users were considered to pose a risk of violence to others. Therefore, not all service users in the two wards would have a HCR-20.

The HCR-20 assessments were reviewed according to the following criteria:

• ∙ RiO Risk Screen scoring based on risk to others (acute) and risk to others (long-term).

• ∙ HCR-20 recorded in RiO documents and the date it was completed. If the HCR-20 was completed by an assessor trained in the instrument and whether there is evidence of multi-disciplinary team (MDT) involvement.

• ∙ Sources of information used, such as patient interview data and records.

• ∙ Whether all HCR-20 risk factors are completed.

• ∙ Number of ‘historical’ factors where information cited was consistent with rating awarded. In other words, the information provided by the assessor for each ‘historical’ factor matched the rating given according to the scoring criteria.

• ∙ Appropriate (evidence given does apply to the factor concerned) and sufficient (evidence is sufficient to justify the score given) scoring of the ‘clinical’ and ‘risk factors’.

• ∙ Degree of completion for risk scenarios and case management; and also a summary judgement.

• ∙ Integration of HCR-20 into RiO notes on Care Plan and Risk Information.

• ∙ Patient involvement in the completion of the HCR-20. Was the patient consulted about or active in the HCR-20? Also, reasons for the patient not being consulted or involved.

The HCR-20 assessors (i.e. those clinicians who completed each HCR-20 assessment) were contacted to ascertain patient involvement in the process. They provided information on whether the service user was consulted (informed it was being completed and/or given a copy); involved (actively participated in the process); or, reasons why they were not consulted or involved in the process.

It is noted that ‘appropriate’ and ‘sufficient’ evidence is subjective. Table 1 provides examples which add some clarity to this decision process. The audit assessor (KG) has extensive knowledge of the HCR-20 and provides in-house training in the instrument. Comparisons of evidence were made by cross checking against the definitions set out in the HCR-20 manual and the standards set by examples given during HCR-20 training. These standards are:

• ∙ The HCR-20 manual should be used as a guide throughout;

• ∙ Completing a HCR-20 should be done collaboratively, multi-evidenced, and multi-disciplinary;

• ∙ As much information as appropriate should be cited to add depth to the risk assessment;

• ∙ If there is no information for one of the factors then ‘not known’ should be cited and ways of finding this information should be sought;

• ∙ The presence and relevance of a risk rating should reflect the information given for each factor, so that an accurate formulation can be created;

• ∙ Information in the HCR-20 should be communicated via RiO and placed in the patients care plan;

• ∙ If necessary, there should be liaison with other teams, such as Safeguarding.

Table 1

Examples of appropriate and sufficient evidence to give clarity to the decision process.

Two of the HCR-20 assessments were cross checked by a colleague, who also provides training on using the HCR-20, for the purpose of checking rater reliability given the relative subjectivity of the ratings. Overall there was a near match between the HCR-20 assessor (KG) and the second reviewer of the results. The only discrepancies were apparent on the issue of ‘sufficient’ evidence being given. This seemed to represent a tension behind quality evidence being sought against practical constraints (e.g. the second reviewer was slightly more lenient in terms of the amount of evidence presented in recognition of time constraints on staff completing the HCR-20).

Recommendations

The results of this audit enabled standards to be proposed for HCR-20 risk assessments on the low secure and psychiatric rehabilitation wards that were reviewed:

• ∙ All inpatients on the low secure ward should have an up-to-date HCR-20 risk assessment;

• ∙ All inpatients on the psychiatric rehabilitation ward with a RiO Risk Screen score of two or more (significant or high risk) for either acute or long-term risk of violence/harm to others should have an up-to-date HCR-20 risk assessment;

• ∙ All HCR-20 risk assessments should have been completed in the last six months which is in keeping with routine clinical practice for Care Programme Approach (CPA) reviews;

• ∙ HCR-20 risk assessments should be fully completed with evidence and a rating for all 20 items, and entries should be made for the risk scenarios, case management and prioritisation sections;

• ∙ HCR-20s should be cross-referenced in the care plan and RiO risk information;

• ∙ All service users should be informed that a HCR-20 risk assessment has been undertaken and, if clinically appropriate, they should be actively involved in the process; assessors must document reasons if this is inappropriate;

• ∙ Training should be provided in HCR-20V3 and the Trust Forensic Liaison Team was available to provide supervision and support to colleagues completing HCR-20 assessments.

Strengths and Limitations

This audit was useful as it obtained data relating to the quality of assessments, which in turn has identified staff training needs and can be used to determine the standards set relating to this process. In addition, it allowed data to be accessed regarding service-user involvement. Most of the assessors were trained in HCR-20 assessments to the standard expected of the auditor. Therefore, any shortcomings would be either due to a lack of effort or understanding on the assessor’s part. Knowing that most of the assessors were trained to standard made it easier to highlight faults that needed to be rectified. One limitation of this audit is that it is based solely on data from two wards within Somerset Partnership NHS Foundation Trust. Other trusts may have different guidelines for assessors using the HCR-20. It should be noted that the audit was largely conducted by KG. Rater bias is a possibility in that KG is a trainer of the HCR-20 within the Trust. However, KG is not based on either of the two wards audited.

This audit addressed the issue of service-user involvement in completing HCR-20 assessments, and determined this by directly asking staff. Given the noted discrepancies in research comparing clinician and service-user perspectives, a limitation may be the restrictive method of determining service-user involvement in the process. Although this audit looked at an area which has not been investigated much before, and highlights some key points relating to use of the HCR-20, future research should focus on amalgamating this data with data from other trusts to enable more widespread generalisation of the recommendations.

This audit had a small sample and therefore generalisability would be difficult. The sample reflects the number of patients admitted to the low secure unit and psychiatric rehabilitation ward in a year. Potentially the audit could have extended to the low risk psychiatric intensive care unit (PICU) to increase sample size. This audit shows that even when trained to a high standard, the quality of the HCR-20 assessments can slip, hence the usefulness of this type of audit. A further limitation is that the HCR-20V3 has recently been released which has improved the vague item descriptions for the clinical and risk factor items. This may aid assessors in matching evidence with ratings given for these items. It would be interesting to replicate this audit with this new version to determine whether all of the recommendations are still appropriate and this will be conducted as a follow-up.