There has been a growing amount of research in recent years attempting to clarify the connections between problem gambling and suicide. Studies investigating suicidality in gambling-treatment populations establish strong links, and there is evidence that an alcohol problem (typically co-morbid with gambling disorders) significantly increases the risk. It is important to evaluate which clients are most at risk, which necessitates considering problem gambling and suicide statistics.

There is also the question of what constitutes best practice, not just for problem gambling services, but also for the many other agencies that are seeing growing numbers of problem gamblers. The importance of addressing this issue directly is discussed in this chapter, along with a relevant case example.

Finally, consideration is given to the fortunately not common situation in which a client does complete suicide while in contact with a service, and the impact that this can have for counsellors and services.

SUICIDE IN NEW ZEALAND: SETTING THE SCENE

Until recently, suicide trends in New Zealand had shown an upward tendency overall, with an increase of 72% for the male suicide rate from 1978 to 1998 (Ministry of Health, 2003a). However, in 2000, the total volume of suicides was down to 458, which was the lowest rate since 1985, with the female rate registering the least numbers since 1961 (Ministry of Health, 2003a). Despite these statistical improvements, New Zealand has the fourth highest male suicide rate in the world, and ranks second for males 15-24 years of age. Maori deaths by suicide have remained stable over the same time period; this ethnic group is over-represented, numbering approximately 18% of New Zealand suicides. The most common method of suicide for both males and females is hanging, with this accounting for over 40% of the suicides in 1997 (New Zealand Health Information Services, 2001).

The statistics for suicide attempts or intentional self-harm indicate that this is a different group from those who complete suicide. The most recent hospitalisation figures for intentional self-harm are for 2000/2001, in which there were 5,168 presentations at New Zealand hospitals (Ministry of Health, 2003a). The females to male ratio for intentional self-harm in the same time period was 1.8 female hospitalisations to every male one. It is important to note that there are no accurate data on suicide attempts, because records are only kept on those who are admitted to hospital or seen as day patients for longer than three hours. It is also not possible to compare the most recent data with numbers from previous years, since the definition of self-harm has been extended to incorporate cases not previously included.

SUICIDE AND PROBLEM GAMBLING: IMPORTANT CONNECTIONS

There are several types of research relevant for understanding the connections between problem gambling and suicide, such as population studies, social impact assessments, research with clients undergoing treatment, and analysis of individual gambling-related suicides. A selection of these studies is examined here.

The Australian Productivity Commission (1999), in their report on the nation's gambling industries, attempted to estimate Australian gambling-related suicides. They investigated case studies of individual gamblers and surveys of problem gamblers, both in treatment and those identified in general population studies. They felt it was probable that a proportion of problem-gambler suicides reflect wider problems and may have occurred anyway, but equally many suicides may be misdiagnosed as car accidents, drowning, or other forms of death. The Commission used epidemiological evidence to suggest a figure of around 400 deaths per year, but acknowledged that this was probably an over-estimation and that the true total was probably between 40 and 400 deaths per annum. Their conclusion was that there is little doubt that suicides are linked to gambling.

Studies investigating suicidality in treatment populations establish strong links, as would be expected. New Zealand research by Sullivan (1994a) found that over 80% of problem gamblers reported suicidal ideation and saw death as a solution to their problems. These statistics were collected during the first 12 months of operation of a gambling problem hotline. While it might be expected that a new, potentially anonymous service may attract those with the most serious difficulties, the data highlight the way in which those with serious gambling problems identify suicide as a solution. Brown (1994) reports that 59% of clients presenting for gambling treatment had experienced some measure of suicidal thought. He comments that this figure does not include the unknown numbers of potential clients who completed suicide.

While numerous studies have investigated completed suicides, it would appear that gambling generally has not been considered as an issue by these researchers. There are reported cases in the suicide literature for which gambling may well have been the precipitator, or part of the problem, and yet this is not recognised by the investigators. Highlighting this are two case examples offered by Bongar (1996), in which suicide followed within one week of contact with a teaching hospital's emergency mental health service. Both victims were males, aged 50-60 years, with major depressive episodes. Each of them had supportive families, but felt that they were inadequately providing for their households. Neither admitted continued suicidal thoughts but, in both these cases, shame and humiliation appeared to be the catalytic motivation for suicide. Their presentation at the emergency mental health services would have been an ideal opportunity to screen for problem gambling, given that there were several indicators of problem gambling (ie depression, financial problems and shame). These cases typify presentations to problem gambling counselling services and they raise the issue of how often problem gambling may be overlooked in a client presentation to mental health services.

A further way of approaching and understanding the link between gambling and suicide is to conduct a psychological autopsy. Blaszczynski and Farrell (1998) undertook an analysis of completed gambling-related suicides and found that almost a third had previously attempted suicide, and one in four had sought mental health assistance for their gambling problem. Other factors identified in the records included co-morbid depression, large financial debts and relationship difficulties. This descriptive study highlights the overlap between gambling and psychiatric disorders, as well as other life problems and suicide. It may also suggest that there is an important link between problem gambling and serious suicide attempts.

Overall, the literature appears to report a strong association between rates of suicidal thoughts, self-harm attempts and problem gambling. An early study by Moran (1969) of 162 members of Gamblers Anonymous showed that 20% of subjects reported having attempted suicide and 77% disclosed suicidal ideation. While again this highlights the serious end of the problem gambling spectrum, it contributes to a growing body of evidence in respect of the association between gambling and suicide. The Australian Productivity Commission (1999), in tabling evidence from the literature on suicide thoughts and attempts among problem gamblers, reports figures of between 4-31% attempted suicide, and 17-80% with suicidal ideation.

THE CONTRIBUTION OF ALCOHOL

While the exact role of alcohol in suicide is also unclear, a strong association certainly exists. Welte, Abel, and Wieczorek (1988) and Berkelman and colleagues (1985) report that between 18% and 66% of suicide victims have alcohol in their blood at the time of death. A clear connection also exists between alcohol and gambling problems, with a literature review documenting that the rates of lifetime substance abuse disorders among pathological gamblers range from 25-63% (Crockford & el-Guebaly, 1998b). A New Zealand study investigating rates of problem gamblers presenting to alcohol and drug services determined that 11% of clients surveyed were probable pathological gamblers and 18% of them (inclusive of probable pathological gamblers) had some gambling problem (MacKinnon & Paton-Simpson, 1999).

Of significant concern and interest is a study by Ciarrocchi (1987), who noted that clients dually addicted to alcohol and/or drugs and gambling were at a greater risk of suicide. He reported that 100% of chemically-dependant pathological gamblers were diagnosed with major depression and of those, 42% had made a serious suicide attempt. This was five times the frequency of the chemically-dependent group alone. The interaction of both alcohol and/or drugs and problem gambling together with suicide has however not been clearly investigated by other researchers, but these data suggest that problem gambling, alcohol and other drug, and mental health services all need to be aware and take special precautions when clients manifest these combined factors. Blaszczynski and Farrell (1998) comment that given these variables of major depression, alcohol and substance abuse, and also marital dysfunction, which are considered risk factors for suicide in both the general population and among psychiatric patients, it is surprising that only a few studies have investigated risk factors associated with suicide in populations of pathological gamblers.

WHO IS MOST AT RISK?

It is important to consider that many of the client groups that have been noted in problem gambling research as being most at risk are also the groups that are identified as posing concerns in the suicide literature. As stated earlier, those who have psychiatric diagnoses, which include alcohol and other drug problems and pathological gambling behaviour, are more at risk of suicide. Beautrais and her colleagues (1996) determined in their study that the risk of a suicide attempt increased with increasing psychiatric morbidity, and that subjects with two or more disorders had odds of serious suicide attempts that were 89.7 times the odds of those with no psychiatric disorder. Pathological gambling, a diagnosable psychiatric disorder under DSM-IV in itself is often seen with other psychiatric disorders, notably depression and substance misuse. This indicates, therefore, that this is a group of clients who are significantly at risk, and clinicians need to treat them accordingly.

Youth are another at-risk group. Research suggests that gambling problems are much higher amongst youth. The Australian Productivity Commission (1999) identified that those under age 25 were about twice as likely to have a gambling problem as those over age 25, and a meta-analysis undertaken by Shaffer and Hall (1996) confirmed that between 4-8% of youth have serious gambling problems, with another 10-15% remaining at risk of developing a serious gambling problem. Although the most recent New Zealand suicide statistics indicate a drop in the youth suicide rate (Ministry of Health, 2003a), as stated earlier this nation ranks second in the world figures in terms of the male youth suicide rate, and youth have the highest hospitalisation rates for intentional self-harm.

Maori also are over-represented in both problem gambling and suicide statistics. More than a quarter of problem gambling clients attending counselling services are Maori, compared with national representation of only 10.9% in the population 18 years and over (Paton-Simpson et al, 2003). As previously stated, the suicide rate for Maori has not decreased and remains at an over-represented 18% (Ministry of Health, 2003a).

INTERVENTIONS: BEST PRACTICE

The need to intervene in respect of suicidality and problem gambling will be necessary for a range of services. However, because of the high numbers of clients who will be presenting to problem gambling services with co-existing depression and alcohol/drug issues, this is perhaps the ideal service for demonstrating best practice in this area. It is important to remember also that these high correlations do not relate just to those who have family/whanau members or a significant other with a gambling problem.

In the first instance, all clients who present to problem gambling services, including family/whanau, need to be screened for depression and suicidality. As with problem gambling, suicidal thought or intention is not observable, and often clients will perceive their thoughts of suicide as a further inability to cope and will therefore be reluctant to disclose them. Education as part of the assessment is an important process, with information routinely being offered to clients about how often suicide may be perceived by clients as a solution to their difficulties. Once clients have been reassured about the normalcy of their reasoning processes, it is important to ask a direct question regarding their thoughts or plans regarding suicide. This needs to be incorporated, either directly by the worker, or as part of the screening procedure that the agency may undertake.

Alcohol and other drug services are also likely to have many problem gamblers within their services, clearly indicated by the correlations presented above. A challenge in identifying the presence of problem gambling, however, is that research indicates some difficulties when gambling screens and brief interventions are used by services specialising in alcohol land other drug misuse (Sullivan & Penfold, 2000). In a trial examining obstacles to incorporating screening for problem gambling in an alcohol and drug treatment setting, Shepherd (1996) identified several factors including:

• Lack of awareness of the prevalence of gambling addiction among substance abusing populations.
• Some saw it as legitimate that lower socio-economic populations often seen at Alcohol and Drug Clinics should try to solve financial problems through their gambling.
• Preconceived attitudes around their ability to identify problem gambling without screening (ie that it would be obvious).
• Gambling addiction may be seen by clinicians as less life-threatening than substance abuse and de-prioritised.

This trial was repeated in New Zealand in 2000 (Sullivan & Penfold, 2000, August) using the Early Intervention Gambling Health, or Eight Screen (Sullivan, 1999) in two alcohol and other drug services, with the surprising results that the same obstacles were identified. This highlights the importance of education for workers in alcohol and other drug services, as well as additional services likely to see higher numbers of those with gambling problems (eg mental health services). Education needs to concentrate on the risks for dually-affected clients, the prevalence of gambling problems and the difficulty of detecting them.

While it is essential for questions regarding suicidality to be incorporated into an assessment, it is also important that this issue continues to be raised throughout contact with the client. Pathological gambling is described as a persistent and recurrent disorder (American Psychiatric Association, 1994), and as such, while clients can feel pleased with their progress, tolerance for slips can be low, particularly given the amount of financial damage that can occur in a short period of time, and suicidal thoughts can often re-occur at these moments. Blaszczynsky and Farrell (1998) suggest that an awareness of risk factors in general, and those specific to gamblers, is essential if reasonable standards of care in client management are to be achieved.

THE IMPORTANCE OF ASKING: A CASE EXAMPLE

John (a pseudonym) had been attending counselling for approximately ten months, roughly every fortnight. He had been aware of his gambling problem for over ten years. It had contributed to the break-up of his last marriage, and although he was in his late 50s with a steady job, it had left him in a situation where he was living from one pay cheque to the next. John had attended counselling on and off at different services over many years, but always dropped out of contact when he began gambling again, feeling that he had disappointed the service or the counsellor. The crises that had precipitated his latest presentation to counselling services ten months earlier had been the suicide of his older brother, who had also maintained a gambling problem. Although John had gambled on a few occasions over the previous ten-month period, he had broken his previous pattern by agreeing to make it a priority to come back to counselling if he gambled. He had managed this successfully and learnt a lot about himself and his relationships with other people in the process.

At this particular appointment John turned up, but withdrew into his chair in the way that was typical for him when he had gambled. He acknowledged he had gambled since the last session, but talking about this did not appear to provide any relief in the way that it had usually done. In talking through what John had done and how he had coped since last gambling, his avoidance of the subject alerted the therapist to the need to ask again whether John had considered suicide as a way of managing since the last episode of gambling. Asking John appeared to energise him. He stated that even raising the subject was ridiculous, but then went on to talk about the fact that he had been very aware over the last week of the impending anniversary of his brother's suicide, and the rest of the session was spent discussing this further.

At the following session, a week later, John walked into the counselling room, sat down, and said "how did you know?". He was referring to the question from the previous session regarding whether he had considered suicide, and was able to acknowledge that he had in fact been planning to kill himself. The combination of feeling hopeless that he had gambled again, the approaching anniversary of this brother's suicide, and feelings of envy toward his brother who no longer had to cope with feelings about gambling, had made suicide seem a reasonable option to John. Acknowledging this secret had made a difference to him, along with talking about it, which made him feel as though there were other options.

The example of John's situation highlights the importance of asking and continuing to ask about suicidal thoughts, even in longer-term relationships with clients. The nature of gambling problems, and the way they are able to be kept private much more than many other addictions, fits very well with the often secret nature of suicidal thoughts. In addition, the financial implications of a gambling problem can often mean long-term solutions are required in order to balance things in finances, in relationships, and perhaps in work, and suicide can sometimes seem like a more immediate, less painful solution. This is particularly so because of the shame attached to both gambling problems and suicidal thoughts, with both being difficult to voice.

WHEN CLIENTS KILL THEMSELVES: HOW DO HEALTH WORKERS COPE WITH SUICIDE?

It has been found that 57% of psychiatrists and 49% of psychologists reported post-traumatic symptoms similar to those of people who had experienced the death of a parent, when a client or patient had killed themselves while under their care (Chemtob et al, 1988). Despite the high reported association between problem gambling and suicidality, the numbers of suicides completed while clients are in contract with problem-gambling services are fortunately not high. Because it is an unusual event, however, there is often consequently a feeling of shame and a reluctance to talk about the experience. Counsellors may experience similar feelings to their clients, along the lines of: "why me, no one else can understand what this is like, what will other people think, where did I go wrong?". In addition, it can be difficult to acknowledge and express grief when the relationship is not necessarily viewed by other people as a close one warranting a grief reaction. S_derlund (1999), outlined a number of other factors that are likely to increase the emotional impact on the therapist, such as:

• Pre-existing stress in the counsellor due to a high work load.
• The depth of attachment between counsellor and client, rather than the length of acquaintance.
• The counsellor's ignorance about the professional ramifications of the death.
• The availability and use of supports by the counsellor.
• Whether the counsellor was working alone with the client or as part of a team.

There may also be a number of practical considerations to attend to, which places further focus on the suicide and the counsellor whose client has died. If an investigation takes place as a result of the suicide there may be police involvement, and it will be necessary to ensure that all notes and documentation are up to date and available. Thee may be other decisions to make, such as whether to attend the funeral and whether to contract the family. For all these reasons it is important that colleagues, supervisors and managers offer support in the event of a client's suicide, and that the counsellors themselves take responsibility for talking about the experience and their feelings, in the way that they would encourage their clients to do.

CONCLUSION
There is strong research evidence that highlights a close relationship between problem gambling and suicide. For counsellors in the problem gambling field this is a clear indication that questions regarding suicidality need to be incorporated in every assessment. In addition, it is important to keep the issue raised as counselling progresses and circumstances change.

For counsellors in other services, such as mental health and alcohol and other drugs, it is also an important issue. As has been demonstrated, the co-existence of an alcohol problem with problem gambling behaviour raises the risk of suicide significantly, and suggests that problem gambling should be routinely screened for by these services along with an assessment regarding the risk of suicide. It also raises the need for further research in this area, in particular what part gambling may play in the suicide and intentional self-harm statistics.

Finally, although fortunately the experience of clients completing suicide is not common, it is important that this remains a topic of conversation within services. The possibility is always present, and it is essential that supportive processes are developed and instituted when they are required.