As is evident from the major texts, the subject matter of forensic psychiatry lies in the application of psychiatric knowledge to offender populations. In principle, this brief extends to all offender groups, although in practice it is the juxtaposition between mental disorder and criminal behavior that is the traditional concern of forensic psychiatry. With respect to mental disorder and criminal behavior, three populations can be defined: (1) those people with a mental disorder who do not commit crimes; (2) offenders who are not mentally disordered (which includes the majority of offenders); (3) those who display both mental disorder and criminal behavior. It is this third group, mentally disordered offenders, that is of particular concern to forensic psychiatry and so forms the main interest of this article.
The starting point in any discussion of mental disorder and crime lies in the definition of mental disorder. In general psychiatric practice, mental disorder is diagnosed by reference to formal diagnostic systems, such as the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders, the revised fourth edition (DSM-IV-R) of which is in current usage. For the purposes of law, however, a legal definition is required: for example, the Mental Health Act, 1983 distinguished four manifestations of mental disorder: mental illness; arrested or incomplete development of the mind; psychopathic disorder; and any other disorder or disability of mind. There is not an exact correspondence between formal psychiatric diagnostic systems and legal definitions, so that in practice forensic psychiatrists must work with the courts in deciding upon issues of mental disorder as the need arises.
The overlap between mental disorder and criminal behavior has concerned criminologists, lawyers and psychiatrists for many years. There are several reasons for this long-standing concern. It is intrinsic to the dignity and fairness of the legal system that it does not penalize those members of society who are mentally unfit. Any individual who cannot function mentally is seen as potentially unreliable and therefore cannot be assumed to be credible; this failing may weaken and call into question the fairness of the process of justice. Similarly, an accused who is not fully competent cannot reliably be said to be making informed decisions with respect to their case. Finally, the imposition of criminal sanctions, such as imprisonment, is not held to be fair and just if the accused is not fully responsible for their actions.
From the wide range of concerns of forensic psychiatry, three areas have been selected for discussion here, as they encapsulate issues at different legal stages with regard to the mentally disordered offender. These three topics are: assessment of competence prior to committal; risk assessment with respect to placement; and treatment during placement.

Assessment of Competence

It is a fundamental aspect of law that those being tried are seen to be responsible for their actions: first, the accused is responsible for their actions at the time that the crime was committed; second, that the accused is fit to stand trial. Each of these will be considered in turn.
For a crime to be committed in law there must be two elements, mens rea and actus reus. While both terms defy simple definition, the former may be thought of as the intention to commit an act while knowing it was wrong; the latter as the act (or failure to act, as in negligence) itself. Thus, the act may be committed but unless there can be shown to be guilty intent then there is no case to answer. The obvious example of this principle lies in the notion of an age of criminal responsibility: a child may commit a serious act but not be held responsible legally for their actions and therefore cannot be held to have committed a crime. Since the trial of Daniel McNaughton (M’Naghten), the law has evolved to take account of the view that the ‘mad’ or ‘insane’ or mentally disordered cannot be held responsible for their actions. In other words, those not seen to be responsible for their actions can be said to have actus reus but not mens rea.
The practical difficulty in the case of the mentally disordered offender lies in knowing whether the accused was aware of their actions and, if so, whether they knew that their act was wrong at the time that they committed the act. In cases of mental disorder it falls to the defense to mount an ‘insanity defense’; that is, to plead that the accused is, in fact, not guilty by reason of insanity at the time the act was committed. There are variations on this theme across different legal systems; however, the task of the forensic psychiatrist remains the same in such cases. That is, the psychiatrist must look back (‘postdict’) to give a professional opinion of the accused’s state of mind at the time of the crime. Given that there may be months or even years between the act and the need for a ‘psychiatric postdiction’, it is clear that such assessments can never be an exact science.
While the moral and legal issues are similar with respect to competency to stand trial (variously referred to as ‘unfit to plead’ or ‘under disability’), the practical task facing the forensic practitioner is placed in the present rather than the past. It therefore falls to the forensic psychiatrist to inform the court whether the accused is mentally disordered, within the legal meaning of the term, such as to impair understanding of the charges faced. There are several protocols that have been developed to aid assessment of competency, such as Competency Screening Test and the Georgia Court Competency Test. It is unlikely, however, that these instruments will replace the traditional clinical skills of the forensic psychiatrist.

Risk Assessment

What is risk assessment? To conduct a risk assessment is to undertake an exercise in prediction: risk assessment is an estimation of the likelihood that some future event will occur. The estimation of risk is a skill that is required in many areas of life: financiers estimate the risks of their investments; insurers estimate the risks of fire, theft, car accidents; and surgeons estimate the risks of operations.
In forensic psychiatric practice with mentally disordered offenders the assessment of risk is a prime concern. It falls to forensic practitioners to make decisions regarding admittance to secure psychiatric facilities, and similarly to decide on recommendations of discharge from security and return to the community. Thus, it falls to practitioners to estimate the risk of an individual inflicting more harm on the community. In essence, practitioners are faced with a classic decision-outcome matrix as shown in Fig. 1.
Consider the legal decision to send a violent mentally disordered person to a maximum security hospital. It is the role of practitioners to inform the court of their prediction as to whether the level of risk of that person being a continued danger to the public (a key criterion) is such as to justify the decision of detaining that individual for treatment in conditions of security. Suppose the psychiatrist predicts Yes and, indeed, if the individual had stayed in the community there would have been more victims (actual Yes) then a decision to admit to hospital is correct (a Hit) and the public will be protected. However, suppose the prediction is Yes but in fact the individual would not have been a continued danger to the public (actual No). Now the person will wrongly be admitted to hospital (a Miss), with all the attendant issues with respect to human rights and public spending.

Actual Yes Actual No
Predict Yes True positive False Positive
(Hit) (Miss)
Predict No False negative True negative
(Miss) (Hit)

Figure 1 Hits and Misses in prediction.

Alternatively, the practitioner might predict No and the person is not detained but then commits further offences in the community (actual Yes). In this event (a Miss), of which there have been several cases widely reported in the popular media, the scene is set for a public inquiry. Finally, the prediction may be No and the individual remains in the community and does not reoffend (actual No); obviously, this is a good decision (a Hit).
If the same logic is applied to decisions about release from secure hospital conditions back into the community, then it is plain that in its assessment of risk forensic psychiatry is charged with a heavy responsibility. In this light, the topic of risk assessment is of fundamental importance.


A term commonly used in risk assessment is the ‘criterion of risk’. This term refers to an exact description of the event that is being assessed: in other words, a clear definition of the concern of a risk assessment. For example, in physical medicine the criterion of risks might be the occurrence of heart disease or cancer, or even particular forms of illness such as angina or breast cancer.
Once the criterion of risk is clear, then to estimate risk, the associated predictors (or risk factors) must be identified. Thus, for the criterion of heart disease, the predictors might include a family history of heart disease, a history of smoking, a pattern of unhealthy eating and lack of exercise, a high cholesterol level, high blood pressure and a stressful occupation. It is important to emphasize that the presence of these predictors does not mean that the realization of the risk is inevitable. Rather, the presence of a predictor increases the likelihood that the event will occur, recognizing that some predictors carry more weight than others. Further, it is generally the case that the greater the number of extant predictors in a given case, the greater the risk becomes. Thus, an individual with a family history of heart disease, who smokes 20 cigarettes a day and has high blood pressure is more likely to suffer a heart attack than another individual who has a moderately high cholesterol reading. The important word here is likely: for neither individual is a heart attack certain, but the risk would certainly be seen as higher in the former compared with the latter.
The identification of predictors of risk is not a simple task. The issue facing practitioners is how to know exactly what aspects of an individual’s history and current functioning should be of concern in order to estimate a given risk. The research literature traditionally makes the distinction between static predictors of risk and dynamic predictors of risk.
Static risk predictors are historical or background factors, the presence of which is known to elevate the risk of a given criterion. In the example given above, a family history of heart disease would be a static risk predictor. Again, the presence of static predictors, which by definition cannot change, do not mean that the event is predetermined: the presence of static predictors simply raises the probability of the event happening in the future.
Dynamic risk factors are aspects of an individual’s current functioning; thus, smoking, an unhealthy diet and a stressful job are all dynamic risk factors in relation to heart disease. In practice, it is often a combination of static and dynamic predictors that gives the strongest basis by which to predict risk.
In summary, when completing a risk assessment system it is crucial to have clearly defined risk criteria. Once the risk criteria are set, the appropriate risk predictors can then be identified and methods of assessment put into place.


The classic distinction with respect to method of risk assessment is to be found in clinical and actuarial prediction of risk.
Clinical prediction The clinical method of risk prediction is based on professional judgment and decision-making. This judgment of risk may be made by an individual practitioner, or by a clinical team, or by a case conference.
Actuarial prediction The actuarial method of risk prediction relies on the use of statistical methods to identify the risk factors for a given behavior, say violent conduct. These statistically identified risk factors may then be combined into an algorithm to give a standardized risk assessment.
In developing systems of actuarial risk assessment, the first task is to identify the appropriate risk predictors for the given criterion. Initially, through the use of clinical records, case material and the research literature, potential risk factors are identified for the criterion of concern. An appropriate cohort is identified and measures are taken of all the potential risk factors. The cohort is then followed over a lengthy period (typically years) to establish which potential risk factors actually best predict the occurrence of the criterion of risk.
A good example of this approach to risk assessment is seen in research, conducted in Canada, concerned with the development of an actuarial prediction scale for violent recidivism in mentally disordered offenders. The cohort in this study consisted of more than 600 males treated in a maximum security psychiatric hospital. After a mean time at risk of almost 7 years, the cohort could be divided into those who had shown violent recidivism and those who had not shown this behavior. These two groups could then be compared across a range of demographic, clinical and offense-related measures, with an aim of determining statistically what particular factors differentiated the two groups and hence predicted violent recidivism.
As shown in Table 1, a total of 12 predictors of violent recidivism were identified using actuarial, statistical methods. These factors are efficient in the sense that they maximize the likelihood of making a Hit and minimize the likelihood of a Miss.
In order to make this research available for clinical practice, the Violence Prediction Scheme, for predicting violence in high-risk men, was produced. The Violence Prediction Scheme presents practitioners with full details on assessing the various predictors. Simply, in making an assessment each risk factor is assigned a score, the scores are totaled and the higher the score the greater the risk of violent recidivism.
Of course, the same procedure can be followed for other criteria. For example, sex offending in mentally disordered populations is an obvious concern: actuarial studies have identified several predictors of sexual recidivism. Another Canadian study reported that rapists presented a greater risk of recidivism than sex offenders against children. As shown in Table 2,a range of predictors of sexual offending was identified; however, while predicting sexual reoffending, the predictors did not discriminate between rape and offending against children.

Table 1 Actuarial predictors of violent recidivism

1. Level of victim injury in index offence
2. Never married
3. Female victim – index offence8
4. Failure on prior conditional release
5. Property offence history
6. Age at index offence
7. Alcohol abuse history
8. High score on psychopathy checklist
9. DSM-III personality disorder
10. Separation from parents under age 16
11. Victim injury in index offence8
12. DSM-III schizophrenia
There is now a great deal of information available to inform actuarial risk assessment with mentally disordered offenders but there remains much to be understood; for example, the research informing the development of the Violence Prediction Scheme reported that a DSM-III diagnosis of schizophrenia had a negative relationship with violent reoffending. It is, however, the case that there is a minority of people with schizophrenia who are at a greater risk of reoffending than others with the same diagnosis. The nature of the relationship between the finer aspects of the diagnostic symptoms of schizophrenia and offending within the population of people diagnosed as schizophrenic remains uncertain.
There is a long-running debate regarding the relative efficacy of clinical and actuarial prediction. Proponents of actuarial methods proclaim the superior accuracy and scientific basis of statistical prediction; proponents of the clinical method argue for the skill of the clinician in making individually based risk assessments and against extrapolating from broad-based research findings to the individual patient. Rather than one method being compared with the other, it would be advantageous to see controlled studies of a combination of methods.

Treatment and the Mentally Disordered Offender

Alongside court work, the delivery of treatment to the mentally disordered is a mainstay of forensic psychiatry services. Such services can be delivered in the community, in hospitals (both open and secure) and in prisons. The task of those offering treatment to the mentally disordered offender is complex. Alongside an awareness of risk to the public, and in some cases the need to work within conditions of security, forensic practitioners are faced with two tasks: the treatment of the disorder and lowering the risk offending. It is not always the case that the two are synonymous.

Table 2 Actuarial predictors of sex offence recidivism

1. Never married
2. Violent convictions
3. Admissions to corrections
4. Previous sexual convictions
5. Previous female victims
6. Previous male child victims
7. Deviant sexual interest
8. High score on psychopathy checklist
Take, for example, the case of a person with schizophrenia who exhibits angry outbursts and commits violent assault. The forensic practitioner is faced with the need to treat the schizophrenia, to treat the anger and to lower the risk of violent behaviour. It cannot be assumed that any or all of these three dimensions of the patient’s behavior, i.e. the schizophrenia, anger and violent conduct, are related. The key clinical skill lies in conducting an assessment that will distinguish between treatment targets related to clinical outcome (i.e. improved mental health) and targets related to lowering the risk of harm to the public. How might the identification of risk factors work in practice? The example below highlights issues in the assessment and management of the risk of violence.

Risk assessment and management

As discussed above, research points to a range of actuarial predictors of violent behavior. In a comprehensive risk assessment system, a patient’s files and case material would be systematically trawled for the presence of these predictors. Clinical interview and observation would then look for the presence of these and other dynamic predictors of violence (e.g. high levels of anger arousal). Thus, a combination of actuarial and clinical risk assessment of the violent conduct will alert clinical teams to the factors that are relevant for an individual patient. This assessment will, in turn, help the clinical teams in setting initial treatment plans to bring about change on the factors related to violence. The process of risk assessment will continue throughout a patient’s contact with clinical services. It is likely that new risk factors will arise as clinical teams develop a fuller understanding of their patient. Finally, practitioners must be confident that their assessment and treatment has had the necessary effects, so that the individual can be released to the community.
When stated in this way the process might seem quite straightforward; in fact, it is far from being an exact science. The complexities are perhaps best illustrated by what is known of the effectiveness of treatment for those sent for treatment as psychopaths.

Treatment of psychopathy

Psychopathy is a legal term which probably best translates into clinical terminology as a personality disorder, or, more exactly, antisocial personality disorder. The characteristics of such individuals, who have generally committed a string of violent and sexually violent offences, is generally assessed in terms of their lack of remorse for their actions, lack of empathy, and deceit and manipulation of other people. The problems of definition and diagnosis aside, there is a (probably small) group of psychopathic/personality disordered people who are dangerous to society at large. Such individuals will be received into forensic services: what is the optimum way to treat them?
Following an extensive review of studies of the treatment of offenders with personality disorder, it was noted that a variety of therapeutic approaches have been tried: this includes psychopharmacological treatment, psychodynamic psychotherapy, and cognitive-behavior therapy. In concluding the review, the author comments that no particular approach has consistently been shown to be beneficial, but procedures that structure the therapeutic environment, such as the therapeutic community, can claim some positive effects.
However, to add to the confusion, those treatment studies that have focused on assessment of psychopathy (using a measure called the Psychopathy Checklist), rather than clinical diagnosis of personality disorder, suggest that such individuals show lower levels of motivation to participate in treatment. Further, they show significantly lower levels of clinical improvement, and are more likely to drop out of treatment before it is completed. Indeed, one outcome study went so far as to suggest that psychopathic offenders showed higher levels of recidivism after time in a therapeutic community.


While relatively few in number, mentally disordered offenders are of great public concern and set many challenges for forensic psychiatry. The complexities of the moral and legal issues with respect to responsibility, the intricacies of assessment and the problems of treatment are all live issues. In looking at the problems it is perhaps easy to forget that forensic psychiatry, as with other forensic specialities, has a short history. If the current pace of growth in knowledge can be maintained, answers to many of the questions set here will surely be found.

Next post:

Previous post: