This research is about the involvement in risk assessment and management of mental health service users considered by professionals to pose a potential risk to other people. The research was carried out due to concerns that the voices of these service users are not being heard in research, policy or practice. The research, by Joan Langan and Vivien Lindow from the School for Policy Studies at Bristol University, found that:

• Many service users were aware that they could pose a risk to other people when experiencing psychosis and they wanted help to reduce the chances of this happening.
• Full and frank discussion of risk to others was difficult for many professionals. Some people could not be asked to participate in the study because they did not know they were considered to be a risk to other people.
• Reasons for talking about risk to others included the user's right to know what is written about them, and increased trust between user and professional, with the potential for collaborative risk management.
• Reasons against talking about risk to others included increasing stigma and distress; this prompted disengagement from services, and fears for personal safety.
• Levels of agreement between service users and professionals about risk and how to respond to it ranged from full agreement to very little or none.
• Service user involvement in risk assessment and management was variable and depended upon individual professional initiative. Few professionals were undertaking systematic risk assessment or risk management plans.
• Some service users influenced the support they received but, generally, users' main role was to accept or reject what was offered. Professionals had less room to accommodate user views where the risks were considered too high. Little use was made of advocates, and users were not provided with a full picture of the different supports available.
• Medication and supported housing emerged as key issues for potential tension and disagreement between service users and professionals.
• Serious gaps and inaccuracies were sometimes found in information held about service users that potentially put themselves and others at risk.
• The researchers suggest that:
  
  - a format for assessing and managing risk could be developed to ensure that service users' views about risks are included; and

- a willingness to take appropriate risks, talk about risk and look holistically at all aspects of a person's life may go some way to enable more effective risk management.

Background
Since the early 1990s, mental health service users have become increasingly defined in terms of risk and dangerousness, despite consistent research evidence that their contribution to violence in society is minimal. As mental health policy has moved further in the direction of controlling individuals considered to be a risk to others, assessing and managing risk is now a key requirement for mental health professionals. The continued focus upon risk means that there is a danger that people so defined will be excluded from decision-making about their lives. Defensive practice is also more likely.

"I think the climate has got much more jumpy... I think we feel much less able to take clinical risks. That's a paradox because, if you're looking at care in the community, there are risks. You cannot make it entirely safe. If you're not prepared to take calculated risks you're in trouble." (A professional)

This research focuses upon the experiences of people being discharged from psychiatric in-patient treatment and moving into the community.

Risk to others

"Most of us find it a little bit kind of strong to somehow say because one's had mental health problems, you're immediately at risk to the public... All those lads out on Saturday nights supporting England are probably much more of a risk to the general public than I am." (A service user)

The behaviour of people in the study included assaults, aggression, serious verbal threats and inappropriate sexual behaviour to women. Sometimes staff were unsure whether an assault or threatening behaviour was due to psychosis. For example, one man agreed that he was a risk to other people and himself when experiencing psychosis, but he had been provoked into many fights with other men, sometimes due to racist abuse.

Suicidal ideas, suicide attempts, self harm and substance misuse were also features of many service users' lives, as were harms such as racism, discrimination or trauma arising from being a refugee or having been abused in childhood.

Service users and risk assessment
There were a few examples where service users were fully involved in risk assessment. However, most service users and relatives or friends were not aware that professionals were formally assessing risks.

Most professionals tended to use their own methods or simple checklists to assess risk. Service users were rarely given copies of any risk assessments, although many workers wanted to work towards this.

Most workers said that they discussed risk with the service users in this study, though many found this difficult. Some users were still not clearly aware that they were perceived to pose a risk to other people. Professionals found discussion easier when they knew the user well, including their positive qualities, and had a good relationship with them, built up over time.

Some service users could not be asked if they wanted to take part in the study because they were not aware that staff considered them to pose a risk to other people. This may have implications for civil liberties, and the effectiveness of consequent service responses.

Some of the reasons given against full and frank discussion of risk were: fears for personal safety or that such discussion would increase stigma and distress or result in disengagement from services. Reasons for talking about risk to others included the user's right to know what is written about them, and increased trust between user and professional, with the potential for collaborative risk management.

Levels of agreement
Agreement between service user and professional ranged from complete agreement about risks to no agreement at all (in one case). Seven service users were open about discussing risk with professionals and agreed the care plan designed to manage risk.

"What worries me is that when I'm in that state of mind it is logical to take the children with me. I could not leave them with the burden of a mother who killed herself... When my thinking gets like that I need to be away from them because in that state of mind they are just extensions of me." (A female service user)

With five service users there was little or no agreement with professionals. The other service users fell somewhere between the two extremes. In a few cases, attempts to discuss risk were made difficult by users walking out or becoming irate, raising fears for worker safety in one case.

A small number were not allowed to refuse 'offers' of support where mental health professionals used their powers to enforce treatment. Professionals generally saw enforcement as counter-productive where it could risk disengagement with services.

Accuracy of information about risk
Whilst many accounts about risk were consistent, we were also given information by service users that staff did not mention, and vice versa. We found some examples of conflicting descriptions of behaviour. A threatened stabbing (not mentioned by the service user) was described by one professional as a risk factor, while another said:

"He's been accused of going to stab a staff member. I was present on that occasion and it was very half-hearted. He came in waving one of our table knives. There was no way he was going to stick it in him, but that's gone down as an attempted stabbing." (A professional)

Service users have the right to have accurate information about them on record. Over-estimation of risk can lead to the service user developing an unwarranted and life-long reputation for being dangerous and being responded to on that basis by agencies. Under-estimation of risk can lead to service users being under-supported, and professionals and others being put at risk.

Risk management
Each service user's care plan included 'monitoring mental health' as well as 'encouraging compliance with medication' if they were ambivalent or unwilling to take medication. Written risk management or relapse plans were rare although, during the study, the Mental Health Trust developed integrated risk management and care planning.

"He seemed quite concerned. He wanted to help... He wanted to look back in my history and try and find out the best ways of dealing with it." (A service user)

A few service users were considered by professionals as involved and articulate about what they wanted from the care plan.

"I feel quite easy with them. I tell them that I do want to be involved... I want to know exactly what the plans are for myself." (A service user)

Definitions and degrees of involvement varied. Some professionals took involvement to mean the service user expressing their views whilst others meant that they had actually influenced the outcome. Few professionals discussed how to involve people who were not forthcoming. An advocate was involved in only one instance.

Support offered
Although not accepted by all service users, at discharge everyone was offered:

• regular appointments with a psychiatrist;
• attending a group, day-care or a drop-in facility; and
• medication.

With the exception of two people, everyone also saw a community psychiatric nurse or social worker. Other support was provided to certain individuals such as intensive rehabilitation or home treatment services, psychotherapy, help with managing voices or finding employment, and daytime or leisure activities.
Housing was an issue for many. Many service users were homeless at or during admission. Some refused to live in supported housing. Reasons included not wanting to be treated like a child, and threats to personal safety from other residents. However, professionals felt that supported housing was the best option in all but one case.

Medication was another source of disagreement, with some service users remaining under 'section' to enforce it and others planning to or already having stopped taking it by the second interview.

"They put me on a very high dose of medication, which almost killed me. They didn't really think nothing of it." (A service user)

Others found medication helpful:

"All the time I was at the day hospital they always used to say 'Would I like to try depot?' I always felt safer just taking the pills because I could regulate it. But after this last thing, where I just exploded sort of thing and did something very stupid, the depot's the perfect thing for me." (A service user)

All those taking anti-psychotic medication either had current or past experience of side-effects. Whilst many staff spent much effort trying to change medication or reduce dosage to lessen side-effects, others seemed unaware. Staff generally thought that medication was 'the lesser of two evils' but for some service users the effects were so distressing and debilitating that medication became the greater evil.

Service users' views about care plans and support
Half the service users felt they were getting a reasonable level of support. Others were not necessarily asking for more support from mental health services but wanted 'a better environment' or to be less lonely. A few wanted more support than they were offered, or support that they had valued had been withdrawn.

There was little sense that service users were aware of the full range of services potentially available to them. However, even if users were aware, they had little power to obtain additional services. Promises of high levels of support at discharge failed to materialise for one user, despite his requesting assistance. His needs covered different social work teams with separate eligibility criteria and he was also not considered to have 'a serious mental illness' so was refused additional psychiatric service.

There were some very positive comments about staff.

"I really feel that she understands me and that she's on my side." (A service user)

Good relationships take time, however, and there was dissatisfaction with staff turnover, especially of consultant psychiatrists in outpatient wards.

Delays in support given
Risk management for people living in the community is more likely to be successful when they receive a quick and effective response to any difficulties. Uncertainty about whether this would happen was of great concern to some service users and relatives. One person waited eight months for their first psychiatric outpatient appointment and some talked of being turned away from psychiatric hospitals or accident and emergency departments in the past Delays were particularly worrying for the two service users who had a diagnosis of manic depression; the period between realising that they were becoming unwell and seeking help and becoming so unwell that they rejected help or had become a risk to other people was a matter of days, if not hours.

"It is difficult to get someone in, especially a person like me who's like ill sort of every three years and not in the system all that long." (A service user)

By the second interview, a few service users and relatives were more confident that systems were in place to respond, should there be a crisis.

Support not offered
The following support was either not provided or only provided to some service users when it seemed that others would have benefited:

• supporting service users subject to racial attacks;
• awareness of and respect for cultural differences;
• specialist support for refugees and asylum seekers;
• support for the children of parents with mental health difficulties;
• advocacy;
• therapeutic input;
• specialist assistance for substance misuse;
• anger management; and
• self-help or self-management groups.

In terms of staff, the researchers found that training to counter institutional racism was needed in terms of race equality training, cultural awareness and a review of existing strategies. Some staff said they would value training on risk assessment and management.

Conclusion
The study shows that service user involvement in risk assessment and management was variable and depended upon individual professional initiative. Few professionals were undertaking systematic risk assessment or risk management plans. The researchers do not assume that risks would be more accurately assessed if they were. However, they suggest that a format for assessing and managing risk is developed to ensure that service users' views about risks are included. It seemed that many professionals were interested in working towards user involvement but found it difficult.

A willingness to take appropriate risks, talk about risk and look holistically at all aspects of a person's life (including other harms they may be experiencing, such as racism, trauma, etc) may go some way to enable more effective risk management. Organisational cultures also influence the extent to which professionals feel constrained to practice defensively rather than defensibly.

Service users generally relied on the professionals working with them in terms of what sort of service they were offered. Little use was made of advocates, and users were not provided with a full picture of the different supports available.

The study also revealed a few worrying examples of inaccurate information held about service users, and examples of vague or missing information.

About the project
The research was conducted in one English urban area, and in-depth interviews took place with seventeen mental health service users. Interviews took place as people were being discharged from two psychiatric hospitals and again six months later. The researchers also carried out twenty interviews with ten relatives and six friends, as well as 78 interviews with 56 professionals from health, social services and the voluntary sector across these two time periods. Fieldwork took place between June 1999 and December 2000. The total of 129 interviews was transcribed and analysed using a computer-aided qualitative data analysis package.

The report contains a number of examples and suggestions for improving practice.

How to get further information
The full report, Living with risk: Mental health service user involvement in risk assessment and management by Joan Langan and Vivien Lindow, is published for the Foundation by The Policy Press (ISBN 1 86134 596 8, price £14.95).

Click on the 'order report' icon in the left margin to order online.