Research on suicide prediction outside prisons generally presents its findings in terms of risk factors for suicide. It is assumed that the more risk factors a person possesses, the more likely the person is to commit suicide. A number of points should be stressed. First, the assumption that risk factors "add up" in this way has been questioned.(3) The presence of a risk factor is not itself a cause of suicide. Real difficulties are encountered in trying to explain why anyone committed suicide, and even greater difficulties are encountered in trying to predict who is likely to commit suicide tomorrow.

In prisons, this problem of prediction is much worse. Many inmates possess enough risk factors to be automatically defined as high risks by community standards, even before the particular stresses of prison life are included.

Studies on prison suicide tend to follow a stereotyped format. The relatively high risk of suicide in prison systems is stressed (incidence may be said to be rising,(4) but these figures are subject to challenge(5)). Findings include:

• The initial phase of imprisonment is the most vulnerable time (but "some suicides occurred many years after reception into prison"(6)).
• Prisoners on remand are the highest risk group.(7)
• A high proportion of violent and sex offenders and lifers are among those who commit suicide.(8)
• A history of psychiatric problems is common.(9)
• The most common method by far is invariably hanging,(10) followed by slashing and overdose.

The main findings of much recent research are practically identical to those in an 1880 medical inspector's report to the Board of Prison Commissioners in England and Wales (see Table 1).(11)

The frankness of Salive et al. in acknowledging the lack of useful research on prevention programs does not prevent them, like most authors, from making suggestions on how correctional services should prevent suicides.

It is customary to distinguish between secondary prevention efforts aimed at individuals who have already been identified as at risk and primary prevention efforts addressing factors in the environment that might reduce overall suicide rates.

Comprehensive secondary prevention programs might include special facilities to house suicidal inmates and allow special observation, including 24-hour, one-on-one surveillance if necessary; measures to recruit family and friends to help authorities identify those who may be suicidal; and the use of "inmate observation aides."(14) However, little consideration is given to the very real problem of ensuring that such programs are available to the truly suicidal and are not overwhelmed with inmates seeking "relief from the obligations of the imprisoned."(15) Also, there is little evidence that such potentially expensive programs can be effective. Indeed, "Convincing arguments can be made against the effectiveness of prevention strategies aimed at identifying the at-risk individual and attempting to prevent his suicide."(16)

Primary prevention efforts generally require advance planning and implementation. They require the judiciary to change policy on remand and sentencing, and consideration of prison design before prisons are built. They almost always require higher staffing levels, and their implementation can have very little to do with front-line staff who are, nevertheless, generally asked to communicate better with the inmates.

Such statements may read well, but they can appear naïve in the context of rising prison populations and economic constraints.

"Common sense might suggest that people who kill themselves in prison do 50 because they are in prison,"(19) and most administrations are likely to agree with Her Majesty's Chief Inspector of Prisons that "general penal reforms were not justifiable on the basis of a single issue such as suicide."(20)

It is well known that those who attempt suicide are different in significant ways from those who succeed, though obviously the two groups overlap and a history of suicide attempts is recognized as an important risk factor for suicide.(21) In practice, attempted suicide is probably the most common way in which potentially suicidal individuals are identified. The other way is if somebody expresses an intent to commit suicide. It is rare for potentially suicidal individuals to be identified by other behaviour, unless it is part of a mental illness.

Mentally ill inmates consistently represent a considerable proportion of prisoners who commit suicide, and their suicides are probably the easiest to prevent. Treatment of the illness can reduce risk, and it is this group that psychiatric services can most help. Assisting other potentially suicidal inmates is a major problem for professionals.

Every penitentiary in the Ontario region maintains an active system for monitoring those at risk for suicide. This system is illustrated in Table 3. A mental health care team meets on a regular basis to discuss intervention and management of those identified. Possible interventions are limited in prison, and, in crisis, often the only option is to isolate suicidal inmates and deny them access to the usual tools for suicide - a razor blade or a noose. Razor blades are easily concealed in a body cavity, nooses are readily improvised. Therefore, initial intervention usually consists of canvas or paper gowns in a bare ceil, perhaps with 24-hour camera observation. This saves lives, but it can result in a loss of trust, and it cannot be pursued indefinitely.

If little is known about the general problem of suicide, much less is known about clusters of suicide within prisons, except that they occur.(24) Outbreaks of suicide have been observed in communities following media depictions of suicide in fictional characters. One prison suicide cluster has been blamed on a restriction of access to psychiatric services.(25)

In trying to understand epidemic suicide, sociological perspectives are likely to be more useful than medical or psychological concepts. Individuals may have their own reasons for deciding to die, but in an epidemic, these are clearly influenced by wider social factors. The growing public concern with each new suicide can itself feed into, and encourage, the next suicide. Until the fundamental faults are addressed, or until such suicides receive no further publicity, the epidemic may continue. One of the more than 50 recommendations of a Correctional Service of Canada investigation into a cluster of seven suicides in the Atlantic region in 1983 was to "undertake a program which would lead to the media in the area downplaying suicide by an inmate...,"(26) but the attempt to restrict media attention may increase a sense of despair and suspicion, making an epidemic worse.

Suicides in prison come in clusters, some of which may be explained, some may not. Such a cluster may be emerging across the country. The explanation for this probably has to do with wider social factors, perhaps an economic recession or prison overcrowding.

Prediction and prevention of suicides are difficult in this high-risk environment, and front-line staff are easily discouraged when management shows little understanding of their problems. Those involved deserve consideration and respect for their struggle in an area with few clear guidelines. Well-intentioned calls to be more alert to suicide prevention are likely only to erode morale.