Trickcyclists.co.uk - Exam advice for MRCPsych candidates

How might risk be effectively managed in a psychiatric setting?

Introduction

the concept of assessing risk, and specifically the risk of violence towards others and the self, has always been a kind of holy grail for psychiatrists
the ability to assess risk, and perhaps more importantly, manage risk is becoming an increasingly important skill for all psychiatrists, not just those working in a forensic setting
recent inquiries into incidents where violence has been perpetrated with someone with a mental illness have pushed the subject increasingly into the public view, with an implicit understanding that psychiatrists can not only predict risk, but are in a prime position to be able to prevent it
Is this the case?

Essay outline

firstly we will look at whether there is evidence that those with mental disorder are more likely to be involved in violent crime
we will then look at some of the ethical dilemmas which arise as a consequence of risk assessment
finally, we shall look at what tools we have, if they are of use, and the future of accurate risk assessment

Mental disorder and violence

in the UK each year, there are approximately 600 homicides
it is estimated that less than 40-50 are committed by those with mental illness - that makes up less than 10% of the total
most violent acts are committed by those without a psychiatric disorder of any kind, and we are at far more risk from those without a psychiatric history than we are from those with psychiatric contact
there is still much to be discovered in this area, but perhaps what we do know can be summarized as follows:
the vast majority of those with a mental illness or learning disability present no increased danger to others
the best predictors of future violence among mentally disordered offenders are the same as those for the rest of the population - previous offending, criminality in the family, poor parenting, etc.
people with severe mental illness, such as schizophrenia and bipolar affective disorder may present an increased risk to others when they have active symptoms
those with mental illness and active symptoms, and who abuse drugs and alcohol may present a significantly increased risk to others
people with psychopathic disorder (an English legal term), by definition, present an increased risk to others
the above risks are summative
it has been found that the rate of violent offending by the mentally ill is remarkable consistent from country to country, unlike the homicide rate within the population
it may be the case that in a country such as the UK, where the homicide rate is comparatively low, those homicides by mentally disordered offenders stand out more than in other countries

Important studies

Hafner & Boker (1982)
- they looked at all homicides and attempted homicides committed in West Germany between 1955 and 1964
- they found that mental disorder was associated with 5% of these
- however, the prevalence of mental disorder in the community was of the order of 3-5%
- the rate among schizophrenics was 5 per 10,000 and they were 100 times more likely to commit suicide than homicide
- the rate among affective disorders was 6 per 100,000, and they were 1000 times more likely to kill themselves than others
Swanson et al (1990)
- using ECA data
- violence in the last 12 months:
  - schizophrenia = 8%
  - drug misuse = 21%
  - schizophrenia and drug misuse = 30%

Specific disorders and violence risk

it would be helpful to look at one or two conditions to look at the associated risks:

Schizophrenia
- more likely to commit minor offences than serious ones
- violence may be associated with:
  - instructions from hallucinatory voices
  - paranoid delusions
  - active psychotic symptoms, especially attempts to test out the validity of their delusional ideas
Manic-depressive illness
- violence tends to be rare, but more common than in depression
- overspending, stealing, etc.
Depression
- violence again is rare
- homicide may be associated with paranoid delusions
- altruistic homicide may occur, and it is often accompanied by suicide
Learning disability
- few clear associations, but the firmest associations are with:
  - arson
  - sexual offences

Lessons from public inquiries

in England, since 1994, it has been the practice that in cases of homicide, it is deemed necessary to hold and inquiry that is independent of the providers involved (NHS Executive, 1994)
these inquiries receive a lot of public interest
one study (Munro, BJPsych, 2000) looked at 40 of these reports to see what lessons could be learned:
28% felt that violence could have been predicted
72 % felt that there had been insufficient evidence to alert professionals (i.e. the violence was not predictable)
60% had a history of violence
65% were considered to have been preventable - but this was not necessarily due to better risk assessment, but by better psychiatric care in general
- the Norman Dunn case indicates that the considered opinion is not that the violence was predictable, but that the deterioration in his mental state should have alerted professionals earlier on

The problems of assessing risk — false positives vs. false negatives

homicide remains a rare event, and so getting a predictive system with high accuracy becomes difficult
there is no such thing as a perfect test - even those with a high sensitivity and specificity will produce a high number of false positives
the threshold for professional intervention will depend on the acceptability of each type of error
by looking at Taylor-Russell diagrams, we can see that a low threshold for action produces a high rate of false positives (low false negatives), and conversely, a high threshold produced a high rate of false negatives (low false positives)
c.f. the legal system:
'beyond reasonable doubt' produces a lower number of false positives
'on the balance of probability' is a lower threshold and reduces the number of false negatives
how we judge it will depend on how acceptable false positives and false negatives are to us, as a profession, and as a population
high false positives:
increase demands upon an already stretched service
ethical concerns about the erosion of patient's civil liberties since both in the community and in hospital, increasingly coercive methods are used
high false negatives:
failing to detect those that commit violence
arguably, an ethical problem as well, since one is failing to protect the public and also failing to provide a vital service to the mentally ill

Improving services

use of CPA
- includes an assessment of risk along with an assessment of needs
- forces risk assessment to be a continuing process which is constantly under review
good information
- 'nothing predicts behaviour like behaviour'
  - the risk factors for violence tend to be the same for the mentally ill as those for the rest of the population:
    - previous offending
    - use of alcohol or drugs
    - social circumstances (association between homelessness and violence)
    - childhood deprivation, poor parenting, violence in childhood
- interagency working and making sure that this information is available to all, and is taken into consideration
proposals by the Royal College of Psychiatrists that those presenting particular risks should be included on a special register
Continuity of care
- use of CPA
- the Michael Buchanan case (1994) illustrates a withdrawal by the community team after he had broken off contact