Effectiveness of problem gambling brief telephone interventions: an uncontrolled outcome study

Effectiveness of problem gambling brief telephone
interventions: An uncontrolled outcome study
Provider Number: 467589
Agreement Number: 326673/00 and 326673/01
FINAL REPORT
Re-issued 18 July 2013
Prepared for:
Ministry of Health
PO Box 5013
Wellington

Authors:
Professor Max Abbott
Dr Maria Bellringer
Nick Garrett
Associate Professor Alain Vandal
Professor David Hodgins
Katie Palmer Du Preez
Dr Jason Landon
Dr Sean Sullivan
This report has been prepared by the Gambling and Addictions Research Centre, National Institute for Public Health and Mental Health Research, School of Public Health and Psychosocial Studies, Faculty of Health and Environmental Sciences, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand. The authors are highly appreciative of, and would like to thank, Gambling Helpline Ltd (now part of Lifeline Auckland) which was the partner organisation for this research. Gambling Helpline staff embraced this project with enthusiasm and recruited all participants to the study, providing the interventions with fidelity and integrity. Without Gambling Helpline's commitment to the study, this research would not have been possible. Thanks are also due to Professor Philip Schluter and Professor Valery Feigin for advice in the developmental stages of the project and Dr Justin Pulford for project management of the pilot phase. Grateful acknowledgement is made of all the gamblers who participated in this study, the majority of whom continued to participate throughout the follow-up interviews. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
CONTENTS

Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
LIST OF TABLES
LIST OF FIGURES
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
EXECUTIVE SUMMARY

Background

Problem gambling and other gambling-related harms have substantial health, social and
financial impacts. They present a significant challenge to governments, communities,
families and individuals. They follow and deepen existing lines of social disadvantage and
health disparity. Various policies and services have been developed with the object of
preventing and reducing problem gambling and associated harms. In New Zealand, as in a
number of other jurisdictions, this includes a national gambling helpline and face-to-face
counselling services. Little is known about how well these services perform, in New Zealand
and elsewhere. Given that services are now increasingly available in many parts of the world,
and were first established in New Zealand 20 years ago, it is surprising that the evidence base
is weak. Internationally there have been relatively few studies of even moderate quality
where people recruited from the community or seeking help from counselling services are
randomly allocated to receive defined interventions to assess their relative efficacy or
effectiveness. It is more surprising that there are not substantially more outcome studies
where people who seek help from, and engage in, a particular service are systematically
tracked afterwards to assess the extent and durability of change over time and obtain more
information about factors that precipitate relapse and sustain recovery. The former type of
investigation, randomised controlled trials (RCTs), is expensive to do well and is difficult to
conduct in real-life clinical settings. While not without challenges, outcome studies are
generally less expensive and somewhat easier to undertake.
The Gambling and Addictions Research Centre at Auckland University of Technology was
commissioned by the Ministry of Health to conduct a RCT of four different, defined,
interventions to be delivered by the national Gambling Helpline. These interventions
included: (1) Helpline standard care (TAU, Treatment as usual), (2) single motivational
interview (MI), (3) single motivational interview plus cognitive-behavioural self-help
workbook (MI+W), and (4) single motivational interview plus workbook plus four follow-up
motivational telephone interviews (MI+W+B). Four hundred and sixty-two Helpline callers
who met the eligibility criteria were randomly assigned to the four groups. Inclusion criteria
were minimum age of 18 years; perception of having a gambling problem; and willingness to
read a short workbook (to ensure reading ability), have calls recorded, provide follow-up data
on gambling, and provide the name of collaterals. Past or present involvement in treatment or
mutual help groups for gambling or other mental health problems was documented and did
not preclude participation. Callers were excluded if they were considered to be actively
psychotic or required immediate crisis or police intervention because they posed a serious risk
to themselves or others. It was found that participants in all four groups improved
significantly, both statistically and clinically, and that these improvements were sustained at
12 months. A 36-month follow-up has recently commenced. While there were no overall
primary outcome differences between the four treatments, some client subgroups did better in
one or more treatments than in others. These findings have implications for potential
matching of clients to interventions to further enhance treatment outcome.
The outcome study reported here is not part of the RCT. That is reported separately. It does,
however, include one of the four intervention groups of the trial participants who received
Helpline standard care. This group was supplemented by additional clients recruited after
trial recruitment had ceased. Recruitment was extended to increase participant numbers and
provide the basis for a stand-alone outcome study. The main purpose of this study was to
ascertain whether there are differences in outcome between those who only access telephone
care and those who also access professional counselling or therapy services additional to the
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
initial telephone intervention, and also to identify client characteristics associated with
treatment outcome.
Methodology

The study was designed as an uncontrolled outcome study. This is a type of prospective
cohort study. All callers initially received brief non-directive counselling to identify
presenting concerns and establish rapport. If the caller met eligibility criteria he or she was
invited to participate. The inclusion criteria were the same as those used in the RCT and
outlined above. In addition to the 116 participants from the RCT, 34 participants were
recruited making a total of 150 callers recruited and followed for 12 months. After consent
was given, participants received a baseline assessment and then received a manualised version
of the Helpline's standard care. This included brief screening, reflective listening to clients'
concerns and, in the case of first-time callers or regular callers who were experiencing
persistent difficulties, referral to face-to-face problem gambling counselling services and/or
suggestions for self-care. Prior to the commencement of the study Helpline counsellors
received extensive training in delivering manualised standard care and their performance
before and during the study was assessed for compliance and consistency. Both were high.
In contrast, there is considerable variability in the approach that face-to-face counsellors take
within and across agencies in New Zealand. This variability is not assessed in this report.
Study participants could choose their own treatment goal (quit some or all forms of gambling,
or control their gambling). Outcome measures were self-reports of days gambled, money lost
gambling and treatment goal success. Other outcome measures included problem gambling
severity, control over gambling, gambling impacts, psychiatric comorbidity, general
psychological distress and quality of life. Although some baseline information was obtained
by counsellors pre-intervention, additional information was obtained by a research team
member within seven days post-intervention.
Results

Of the 150 callers who received Helpline treatment, 86% were assessed at three months,
79% at six months and 66% at 12 months. There were slightly more female participants
(57%) than males. Fourteen percent were aged 18 to 24 years, 25% were aged 25 to 34 years,
27% were aged 34 to 44 years, 20% were aged 45 to 54 years and 14% were aged 55 years
and older. The majority of participants identified their primary ethnicity as either Maori
(43%) or European (42%) with 10% primarily Pacific and five percent Asian or other.
Approximately half (49%) were partnered. Twenty-seven percent had no educational
qualification, 32% a secondary school qualification, 18% a trade or technical qualification
and 23% a tertiary or professional qualification. Forty-four percent were in full-time
employment and 13% had part-time employment. There is no evidence of major differential
attrition with respect to socio-demographic variables. Most participants reported electronic
gaming machines as their primary mode of problem gambling (89%) with 76% citing
machines in pubs, nine percent machines in clubs and three percent citing machines in
casinos.
With respect to gambling outcome measures there were substantial improvements from
baseline to three months, with these improvements being maintained at six and 12 months.
This included problem gambling severity as measured by the PGSI-3 (past three-month time
frame) and PGSI-12 (past 12 month time frame). The median score reduced from 17 to 9. At
baseline 96.6% met the PGSI criteria for problem gambling and 3.5% for moderate risk
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
gambling. At 12 months 57.5% were assessed as problem gamblers, 25.5% as moderate risk and 17% as non-problem or low risk. The percentage of problem gamblers at the baseline and 12-month assessments was similar to that noted for this group (without the additional 34 participants) in the RCT (97.2% and 60.8% respectively). Treatment goals changed somewhat during the 12-month follow-up period, with proportionately more over time seeking to maintain abstinence and fewer seeking to quit all gambling. There was a slight increase in people wanting to gamble in a controlled manner. Participants reported substantial reductions in the adverse impacts of gambling on work, social life, family/home and physical health. Again most improvement occurred between baseline and three months with little change subsequently. A similar pattern was found with respect to psychological distress as measured by the Kessler-10. At baseline, 56% were in the clinically significant high distress category. This reduced to 12.5% at three months and remained slightly below this percentage at six and 12 months. At baseline only 2.8% were in the low distress category. This increased to 62.6% at 12 months. Notable changes were also apparent with respect to mood disorders. At baseline, 57.7% of participants, as assessed by the PRIME-MD, were deemed likely to have a major depressive disorder and 12.4% a minor depressive disorder. At 12 months the corresponding percentages were 18.2% and four percent. Less substantial changes were apparent for dysthymia. There were slight reductions in smoking across the four assessment points (60% at baseline; 50% at 12 months) and alcohol abuse/dependence from baseline to six months (62.4%; 50%); however, at 12 months the percentage was 59.6%. While most participants did not report other drug misuse problems, those indicating moderate or serious problems reduced from 12.3% at baseline to two percent at 12 months. During the first three months of the follow-up period, 31% of participants reported receiving assistance for their gambling problem from a professional treatment service and somewhat more (39%) reported obtaining informal support. Both professional and informal assistance reduced at each subsequent assessment point. At 12 months, the corresponding figures were 19% and 25%. There did not appear to be a clear preference for treatment provider and some sought help from more than one source. From multivariate analyses it was found that males were significantly more likely than females to take up formal treatment during the first three months post-intervention (41.5% versus 22.1%) with an odds ratio of 2.85. In the case of males, this applied irrespective of whether they had, or had not, previously received treatment for a mental health problem (50% versus 40%). However, it was only females who had previously received treatment who were more likely to seek professional gambling treatment (43.8%) with an odds ratio of 4.41 compared with females who had not previously sought treatment (15.4%). None of the other socio-demographic variables including area of residence predicted receiving formal treatment. Participants who had received treatment for a gambling problem and those who had received treatment for another mental health problem during the past 12 months were also more likely to engage in counselling or therapy during this period. A large number of factors were found not to be associated with receiving formal treatment including primary problem gambling mode (EGMs versus other); various indicators of problem gambling severity, impact and duration; motivation to overcome gambling problem; treatment goal; level of belief in treatment success; and perceived difficulty in overcoming problems. Similarly, the various measures of psychological distress, mental health disorder, substance use/misuse, suicidal ideation, quality of life, and prior treatment for an alcohol or drug problem in the past 12 months did not predict treatment engagement. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
With respect to improvements on the PGSI-12 from baseline to 12 months (i.e. reductions in
problem gambling severity) there was no significant difference between those who did and
did not access formal gambling treatment services during the first three months post-
intervention. Some groups of people outside the paid workforce (disabled, had an illness, on
sick leave) improved less on this measure. This was also the case for people residing in more
deprived areas. When confounding factors were accounted for, participants who gambled on
pub EGMs showed substantially less improvement than participants who gambled on casino
EGMs and casino table games (p=0.01). Those who were classified as having minor
depressive disorder improved more (p=0.003). The wide range of other measures referred to
with respect to the analyses of formal gambling treatment involvement were not found to be
associated with reduced problem gambling severity.
Participants who accessed formal gambling treatment services were somewhat more likely to
report a reduction in time-averaged mean number of days gambled per month over the follow-
up assessment period. Widowed participants evidenced less change than those in other
relationship categories. None of the other large number of variables examined was associated
with change on this outcome measure.
In contrast to number of days gambled, no association was found between the uptake of
formal gambling treatment services and time-averaged mean change in money lost gambling
per month. In the multivariate analyses, baseline problem gambling severity (PGSI-12) was
the only statistically significant predictor of change on this outcome measure (p=0.02).
Participants with high scores (more serious gambling problems) improved less.
Participants who had higher belief in treatment success at baseline were more likely to have
somewhat larger improvement in time-averaged self-assessment of control over gambling
(p=0.022). On the other hand, those who perceived a high level of difficulty in overcoming
their gambling problem had somewhat less improvement on this outcome measure (p=0.032).
Participants who scored as having a low quality of life at baseline also improved less, as did
those who had received treatment for a mental health problem in the past year.
None of the various factors examined in relation to time-averaged, self-assessed treatment
success (gambling-quit or improved) reached statistical significance in univariate and
multivariate analyses.
Discussion

Perhaps the most notable study finding is that clients improved substantially, both in
statistical and clinical terms, with respect to problem gambling and some associated mental
health problems. In many instances improvement occurred in the first three months and was
sustained at the 12-month assessment. The durability of these changes will be further
assessed at 36 months. They were achieved even though most callers received only one
Helpline call and did not access other, more intensive, gambling counselling or therapy. Less
change was evident for tobacco use and alcohol misuse. Given that substantial numbers of
clients smoked and experienced alcohol problems, these findings raise questions about how
they could be addressed, either by extending Helpline services or further encouraging referral
to specialist services. There were some client characteristics that were associated with better
or worse outcomes on some measures. There are a number of potential implications for
future research and professional practice. However, the most striking finding was that, for the
most part, clients showed fairly similar overall gains, irrespective of socio-demographic and
other background differences.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
The other major finding is that engagement in additional gambling treatment was not associated with better treatment outcome. This does not mean that these services are not of value. It might be that most clients who do not perceive a need for additional professional assistance do not require it, and that those who do perceive a need, obtain it and benefit. This could explain why there are generally no differences between those who do and do not receive additional therapy. However, if this is the case, it is surprising that those who obtained additional treatment did not have more serious gambling problems, psychopathology and less confidence in achieving treatment goals. Further research is required to assist in the matching of clients to services. Little is known about the nature of face-to-face gambling services in New Zealand and their outcomes. It is not possible to conclude from an outcome study that client improvements are attributable to the intervention provided. However, some of the interventions that were included in the RCT from which most of the outcome study participants were sourced had been evaluated previously relative to wait-list control groups in previous trials. Given that Helpline standard care and these interventions achieved comparable outcomes, it is highly likely Helpline standard care per se contributes to the outcomes observed. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
BACKGROUND
In December 2008, the Gambling and Addictions Research Centre at Auckland University of Technology was commissioned by the Ministry of Health to conduct the research project National problem gambling intervention effectiveness which is reported separately1 and which should be read in conjunction with the current report. The above-mentioned project was a single-site randomised controlled trial of brief telephone interventions for problem gambling involving four groups with repeated measures (pre-treatment, three months, six months and 12 months) enabling investigation of independent, and some interaction, effects of the different interventions. Participants were recruited from gambler callers to the national gambling helpline. The four groups were:  Group 1: Helpline standard care (control group)  Group 2: Single brief motivational interview  Group 3: Single brief motivational interview plus self-help workbook  Group 4: Single brief motivational interview plus self-help workbook plus four follow-up motivational booster sessions.
Current study
The current study is an uncontrolled outcome study. An outcome study involves the
prospective tracking and assessment of a cohort of participants. Participants comprised the
Group 1 participants from the aforementioned randomised controlled trial together with
additional participants recruited to Group 1. All participants received the same treatment.
The main aims of the current study are to:
 Ascertain if there are any differences in participant outcomes between those who only access telephone assistance for gambling problems in comparison with those who also access professional counselling services additional to the initial telephone intervention.  Identify client characteristics associated with treatment outcome. 1 Abbott, M., Bellringer, M., Vandal, A., Hodgins, D., Palmer Du Preez, K., Landon, J., Sullivan, S., & Feigin, V. (2012). Effectiveness of problem gambling brief telephone interventions: A randomised controlled trial. Auckland: Gambling and Addictions Research Centre, Auckland University of Technology. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
LITERATURE REVIEW

Background

Gambling is a popular recreational activity in New Zealand, with between 80% and 90% of
the population reporting participating in the past six- or 12-months (Abbott & Volberg, 1996,
2000; Amey, 2001; Christoffel, 1992; Department of Internal Affairs, 2008; Health
Sponsorship Council, 2007; Reid & Searle, 1996; Wither, 1987). The Department of Internal
Affairs' five-yearly national surveys on people's participation in, and attitudes to, gambling
from 1985 to 2005 showed that rates of participation in gambling have dropped slightly over
time despite the number of gambling opportunities increasing (Department of Internal Affairs,
2008). In 2000, 10% of the population reported not participating in any forms of gambling in
the past year. This doubled to 20% when the survey was repeated in 2005 using the same
methodology. Lower reported participation rates of 63% to 65% have been reported in other
studies (SHORE & Whariki, 2008; Ministry of Health, 2006; 2009) but these differences in
gambling participation from previous estimates may be due to methodological differences in
the conduct of the surveys (Williams & Volberg, 2010).
A wide variety of modes of gambling are approved by the Department of Internal Affairs
under the Gambling Act 2003. These include:
 Lotteries Commission products, e.g. Lotto, Bullseye, Big Wednesday  Scratch tickets  Electronic gaming machines in pubs, clubs and casinos  Casino table games  Track betting with the TAB  Sports betting with the TAB  Card games, e.g. poker  Casino/gaming evenings
These various legalised modes of gambling attract large amounts of spending each year. Over
$1,900 million was spent on gambling in 2011 (Department of Internal Affairs website,
2012), with non-casino electronic gaming machines accounting for $856 million of that
expenditure. Table A shows the breakdown of gambling expenditure over the most dominant
modes in 2011.
Table A: Gambling expenditure in 2011
Gambling sector
Expenditure in 2011
TAB racing and sports betting NZ Lotteries products Non-casino electronic gaming machines Casino gambling (electronic gaming machines and table games) Total $1967 million (Department of Internal Affairs website, 2012) Although most people are able to gamble without ever experiencing any harm, it is estimated that one percent to two percent of the adult population meet the criteria for problem gambling and experience significant harm as a result of their gambling. A national survey of Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
4,053 adults over the age of 18 years, estimated that the prevalence rate for current problem
and probable pathological gambling was 2.1% (±0.4) and 1.2% (±0.3) respectively (Abbott &
Volberg, 1996; 2000). Another face-to-face national survey of a random sample of
12,488 New Zealanders over 15 years of age was conducted in 2006/07 (the New Zealand
Health Survey). That survey found that 0.6% of past-year gamblers could be classified as
problem gamblers, with an additional two percent being classified as moderate-risk gamblers
(Ministry of Health, 2009).
In addition to the harm experienced by problem gamblers, affected others (i.e. people close to
a problem gambler, such as nuclear family, whanau, and work colleagues) also experience
harm due to someone else's gambling. Analysis of data from the 2006/07 New Zealand
Health Survey indicated that 2.8% of New Zealanders over the age of 15 years had
experienced problems due to someone else's gambling in the past 12 months (Ministry of
Health, 2009).
Treatment in New Zealand

There is a wide range of problem gambling treatment providers located throughout New
Zealand both in urban and rural localities, all of which provide services free of charge. The
gambling helpline is the primary telephone counselling provider and is staffed 24 hours a day,
365 days per year by trained counsellors. In 2011, the gambling helpline received calls from
1,242 new problem gamblers and 600 new affected others (Gambling Helpline, 2012). Face-
to-face counselling services assisted 12,090 people seeking help for gambling-related issues
over a similar period (July 2010 to June 2011) (Ministry of Health, 2012).
The Problem Gambling Foundation (PGF) is the largest nationwide provider of gambling
intervention services. PGF offers counselling in a range of languages including, but not
limited to, Maori, Pasifika, Mandarin and Cantonese. As well as face-to-face counselling,
PGF also has gambling ‘hotlines' for telephone counselling for the general public, and for
Asian and Pasifika help seekers (PGF website). PGF face-to-face counselling varies in
counselling technique based on the preferred modality of individual counsellors. Thus,
different clients may receive different types of intervention (e.g. motivational interviewing or
cognitive behavioural therapy). This is most likely the case for many of the face-to-face
counselling services as there is no requirement for standardised care at present. This presents
a difficulty when making recommendations based on treatment outcomes reported by service
providers.
The National Oasis Centres, a branch of the Salvation Army, also offer nationwide face-to-
face counselling for problem gamblers and affected others. Oasis also makes referrals to self-
help groups such as Gamblers Anonymous (Oasis website).
Gamblers Anonymous (GA) is a member-run self-help organisation that is founded on the 12-
step programme instituted by Alcoholics Anonymous. GA promotes abstinence from
gambling through peer support and has groups throughout New Zealand (Gamblers
Anonymous website).
Hapai Te Hauora Tapui is a Maori public health provider that uses a Maori framework as a
base for providing information, development and guidance for problem gambling services and
the general public at a local (Auckland) and national level (Hapai Te Hauora Tapui website).
There is also a range of regional Maori, Pacific and Asian treatment providers that work with
local communities. For example, Te Piringa Tupono and South Seas Healthcare Trust both
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
provide problem gambling treatment services for the people of Manukau in Auckland, the
former assisting Maori and the later, Pasifika.
Problem gamblers accessing treatment

In terms of the number of people assisted by face-to-face problem gambling treatment
services each year, Table B shows the total clients (excluding brief interventions) over the last
seven years (Ministry of Health website). It should be noted that the way data were collected
changed over time so the numbers are not directly comparable. However, it can probably be
concluded that the number of people seeking help for gambling has risen overall between
2004 and 2011. There were large increases occurring in the years 2006/07 and 2008/09 and a
small decrease in help-seekers in 2010/11 but it is yet to be seen whether this decrease will
continue into 2012.
Table B: Total clients recorded by face-to-face problem gambling treatment services
(excluding brief interventions)
clients ↑ from prev. (Ministry of Health website)
The gambling helpline has a reduction in numbers of clients seeking help over the past seven
years especially in the case of existing clients (Table C) (Gambling Helpline, 2012). This
reduction in existing clients may be due to the gambling helpline referring clients to face-to-
face services, such as PGF. This may have resulted in a skew of the total number of existing
clients due to the fact that the gambling helpline mainly offers brief interventions over the
telephone, rather than more structured long-term counselling.

Table C: Total clients recorded by the gambling helpline
Jan - Dec Jan - Dec Jan - Dec Jan - Dec
clients ↑ from prev. (Gambling Helpline, 2012)
Treatment modalities

The following sections of this literature review focus on the major different modalities of
treatment available to people who are experiencing harm from gambling and for which
outcome study data are available. We acknowledge that other treatment modalities exist. The
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
fact that the various treatment modalities are generally not standardised make effectiveness
difficult to assess. This review examines the evidence presented by various outcome studies
and does not attempt to differentiate between the techniques used, aside from those
differences mentioned in the methodologies of these studies. Whilst randomised controlled
studies provide more robust data than outcome studies regarding the efficacy and
effectiveness of the treatment modalities, they have been discussed in the literature review
within the main report for the effectiveness study of brief interventions. The current report is
a companion report focusing on outcomes of helpline standard care participants from the main
study who did, or did not, access professional counselling services additional to the initial
telephone intervention.

Gamblers Anonymous

Gamblers Anonymous, founded in 1957, is based on the Alcoholics Anonymous 12-step
model of self-help treatment. Members attend meetings with other recovering problem
gamblers and share stories about their experiences with gambling (Gambler's Anonymous
website). This treatment programme promotes abstinence as the only acceptable form of
recovery.
Multiple studies have reported GA as being ineffective at treating problem gambling (Petry &
Armentano, 1999). Retrospective studies show that less than 10% of those who attend GA
become active members (and are thus abstinent) (Stewart & Brown, 1988) and of those who
attend meetings, only a small percentage achieve a year of continued abstinence from
gambling (Brown, 1985). In a study of 232 GA attendees conducted by Brown (1985), it was
found that 7.5% were abstinent after one year of attending GA meetings and 7.3% were
abstinent after two years. Brown argued that this shows that GA has some benefit in
achieving long-term behaviour change, but only to a minority of attendees.
When GA is combined with other treatment types, it may be more effective than when it is
used as the sole treatment. However, due to outcome studies not having control groups and
the treatment not being strictly monitored in many studies, especially post-treatment, the role
that GA plays in any success cannot be concluded (Russo, Taber, McCormick, & Ramirez,
1984; Petry, 2003; Taber et al., 1987).
Behavioural therapy

Behavioural therapy is based on the theory that behaviour is learned through negative and
positive reinforcement. There are different forms of behavioural therapy including aversion
therapy, exposure therapy (urge reduction or extinction), and techniques to assist in avoidance
of behavioural triggers. In the case of gambling, it is argued that arousal experienced by
gamblers when in the gambling environment and winning (and associated random
reinforcement schedule2) act to reinforce pathological behaviour (Anderson & Brown, 1984).
McConaghy, Armstrong, Blaszczynski and Allcock (1988) argue that the behaviour becomes
so habituated that if one does not perform the behaviour when stimulated, the experience is so
negative that one is compelled to participate in the behaviour in order to relieve these
feelings.
2 A random reinforcement schedule refers to a ‘reward' of some kind being offered in response to a specific act, in this case, at random intervals. In the context of gambling, wagering money is the act and the reward is winning money or, in the case of electronic gaming machines, winning ‘free spins'. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Early behavioural treatment for problem gambling included aversion treatment, the most common of which involved electric shocks and inducing vomiting through medications (Lester, 1980). These techniques resulted in short-term abstinence from gambling but did not produce long-term behaviour change. Walker (1993) illustrated this in a review of seven studies that used aversion treatment with a combined total of 53 participants. When participants were followed up between six months and three years after termination of treatment, only 12 participants were abstinent and nine considered themselves improved. Two case studies conducted in the late 1970s added the use of other behavioural techniques to aversion therapy: contingency management (strategies the individual can draw on to reduce the chances of relapsing) and controlled gambling (Dickerson & Weeks, 1979; Rankin, 1982). These case studies found that the addition of these other behavioural aspects to therapy resulted in both patients achieving their stated goals of reduced gambling over the course of treatment. Many behavioural treatments for problem gambling are based on the treatment developed for obsessive compulsive disorder (OCD), in which a participant is repeatedly exposed to stimuli, including imaginary scenarios (imaginal desensitisation), that cause them to experience problematic urges until they no longer feel the urge (Marks, 1987). In the problem gambling area, these stimuli are often those relating to the mode of gambling in which the participant is experiencing problems. In a study comparing two groups of ten problem gamblers, McConaghy, Armstrong, Blaszczynski, and Allcock (1983) found imaginal desensitisation more effective in decreasing gambling urges, gambling frequency and anxiety than aversion therapy. McConaghy, Blaszczynski and Frankover (1991) conducted a study comparing the outcomes of 120 problem gamblers who underwent a five-day inpatient treatment programme. Participants were consecutively allocated to one of four different behavioural therapy programmes and followed-up two to nine years later. Sixty participants were allocated to receive imaginal desensitisation (ID), whereby the participant visualised scenarios in detail in which they experienced the urge to gamble but did not act on that urge. Twenty participants were allocated to aversion therapy, where they were exposed to mild electric shocks to their fingers when they read aloud phrases that described their past gambling behaviour. In each session, the last phrase read described them avoiding a gambling situation and that did not result in a shock. Twenty participants were allocated to imaginal relaxation, where they were taught to relax by visualising scenes that they found soothing. A typical scene reported by the authors was of the participant lying in a hammock with a drink. Once the participant had achieved a state of relaxation, they signalled to the therapist and then moved on to imagining the next relaxing scenario. Lastly, ten participants were allocated to brief in-vivo exposure and ten were allocated to prolonged in-vivo exposure. In-vivo exposure consisted of participants being exposed to gambling stimuli but not acting on their urges to gamble. In this case, participants accompanied the therapist to a gambling venue, with those in the brief exposure condition staying at the venue for 20 minutes each day and those in the prolonged exposure condition staying at the venue for 60 minutes each day. Sixty-three participants completed follow-up measures (52% of the sample). At follow-up, 18 participants reported that they had ceased gambling, 25 reported having continued gambling in a controlled manner, and 20 reported continuing gambling in an uncontrolled manner. Of the 33 participants in the ID group who completed follow-up measures, 26 (79%) had ceased gambling or gambled in a controlled manner. This is significantly different from those followed-up who were allocated to the other behavioural procedures, 16 (53%) of whom had ceased gambling or gambled in a controlled manner. Differences in gambling behaviour at follow-up between participants allocated to the three Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
alternative behavioural therapies were not significant. Due to the lack of control group, small sample size and the low retention of participants for follow-up, conclusions from this study need to be interpreted with caution. Another study examining the effectiveness of a behavioural intervention, in this case exposure therapy, was conducted by Symes and Nicki (1997). Using progressively more intense stimuli, two participants were exposed to gambling situations without the financial reward being present. Both participants experienced a decrease in their urges to gamble and were abstinent at the end of treatment (80 days consisting of 11 days of baseline measures being taken and 69 days of therapy). In a case study reported by Oakes, Battersby, Pols and Cromarty (2008), a 31 year old woman who fit the criteria for pathological gambling was treated with an exposure therapy and relapse prevention programme. The exposure hierarchy used for this therapy commenced with several stepped stages progressing gradually from imaginal exposure of gambling cues, through to pictures of favourite gaming machines, with the final stage being to sit at a machine and put money into the machine without playing, then cashing out the money when the urge to gamble had dissipated. Following treatment completion, the participant completed measures, including the South Oaks Gambling Screen (SOGS), the Beck Depression and Anxiety Index, and goal evaluations every three months for one year. Following that, measures were completed yearly up to four years post-treatment. Immediately post-treatment, the participant was no longer classified as a pathological gambler. These treatment gains remained stable at the four year follow-up. Gains were also made on the Work and Social Adjustment Scale and the Beck Depression and Anxiety Index, moving from the clinical range to normal post-treatment, and remaining stable at four years. A larger study examining exposure therapy was undertaken by Smith and colleagues (2010), using 127 clients of the Statewide Gambling Therapy Service in South Australia. Participants were treated with exposure therapy and followed-up one, three, six and twelve months post-treatment. Results showed that exposure therapy was effective at improving measures of problem gambling severity, psychological distress and impaired functioning at short- and mid-term follow-up. However, these gains were not held long-term. Additionally, retention was relatively low with 91 (71.7%) participants completing measures at three months and 80 (63%) at six months. Other studies of exposure therapy have included cognitive therapy as an adjunct or supplement. Battersby, Oakes, Tolchard, Forbes and Pols (2008) assessed outcomes following treatment at the Flinders Therapy Service for 123 problem gamblers who completed treatment. Flinders Therapy Service is a treatment facility in South Australia for problem gamblers referred by state-funded agencies. Treatment was administered via a variety of modes, including face-to-face, telephone, videoconferencing and group sessions. Of the 123 participants who completed treatment, 54 (44%) completed follow-up measures at six months. Of these 54, only three participants (2%) reported having lapsed into problem gambling. The majority of participants also showed improvements in their depression and anxiety levels. The authors acknowledged that a weakness of the study was the lack of resources that could be dedicated to re-contacting participants. Participants were encouraged to return to the Service to complete follow-up measures but were not actively pursued by staff. No differential analysis was performed on the non-completers (Battersby et al., 2008). The Flinders Therapy Service has also published a case study of a woman treated for problem gambling using the exposure therapy with cognitive therapy model (Battersby et al., 2008). Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
"Melissa" attended eight weekly sessions and was followed up over two years. At the
beginning of her treatment, Melissa scored 14 on the SOGS, indicating a severe gambling
problem. Mid-treatment, immediately post-treatment, and at six-, 12-, and 24-months,
Melissa scored zero on the SOGS and had also significantly reduced her depression and
anxiety to nonclinical levels.
In another study using the Flinders Therapy Service's model of exposure therapy plus
cognitive therapy, the effectiveness of group therapy in a rural setting was examined (Oakes,
Gardiner, McLaughlin & Battersby, 2012). Seven problem gamblers completed pre-screening
for gambling problem severity and comorbid problems and all were recommended to attend
an inpatient programme at Flinders Medical Centre. However, all chose to participate in an
intensive outpatient programme in the rural community where they lived. This programme
consisted of two five-day blocks over a three-week period, as opposed to the usual outpatient
programme of 12 once-weekly sessions.
Of the initial seven problem gamblers recruited for the rural Flinders study, six completed the
treatment (including homework tasks) and the follow-up measures. At the initial assessment,
the median score on the SOGS was greater than five indicating probable pathological
gambling. At all five post-treatment points (immediately post-treatment, three-, five-, six-
and 12-months following treatment), the median SOGS score was below three, indicating that
the participants no longer fit the criteria for probable pathological gambler. Participant
statements at 12-months also showed that all six no longer considered themselves problem
gamblers. Five of the six participants reported being able to enter a gambling venue without
feeling any urge to gamble (one of the six had relapsed). The authors reported that the above
gains from a three-week intensive outpatient programme show the effectiveness of this
treatment, even over short time periods of treatment (Oakes et al., 2012).
Cognitive therapy

Cognitive treatment is based on the theory that problem gambling stems from gamblers'
erroneous beliefs in the randomness of gambling. Gamblers often report cognitions of having
some control over the outcome, of having strategies that influence the game, or of being able
to predict when a win will occur (often thinking that losing is necessary in order to have a big
win later, or that winning predicts more wins to follow) (Ladouceur & Walker, 1996,
Gaboury & Ladouceur, 1989). Cognitive treatments aim to challenge these erroneous beliefs,
leading to a change in behaviour in the problem gambler.
In a case study conducted by Toneatto and Sobell (1990) it was found that cognitive treatment
resulted in a reduction of gambling frequency from seven times per month to once a month
immediately post-treatment. The frequency of gambling dropped further to once every two
months, six months following treatment.
In a small study conducted by Ladouceur, Sylvain, Letarte, Giroux and Jacques (1998),
cognitive therapy was used to treat five participants diagnosed as pathological gamblers by a
clinician using the DSM-IV. A psychologist delivered the cognitive therapy to participants
once or twice a week for 60 to 90 minutes, with a maximum of 20 hours of treatment per
person. At the end of treatment, four participants no longer met the criteria for pathological
gambling. These treatment gains were maintained at six months post-treatment.
Similar to many behavioural treatment studies, cognitive treatment studies suffer from small
sample size and lack of control group. There is also a lack of studies for techniques that focus
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
on the use of cognitive or behavioural treatments only, rather than in combination (Petry &
Armentano, 1999; Emshoff et al., 2007).
Cognitive behavioural therapy

Cognitive Behavioural Therapy (CBT) is a broad term encompassing behavioural and
cognitive interventions. It was described in 1996 as involving "a highly diverse set of terms
and procedures" (Brewin, 1996, p31) and more recently as the "generic term referring to
therapies that incorporate both behavioural interventions (direct attempts to reduce
dysfunctional emotions and behaviour by altering behaviour) and cognitive interventions
(attempts to reduce dysfunctional emotions and behaviour by altering individual appraisals
and thinking patterns)" (Problem Gambling Research and Treatment Centre (PGRTC), 2011,
p66).
Studies have indicated that CBT is associated with positive outcomes though these have not
necessarily used CBT alone and the CBT is not necessarily more efficacious than other
treatments for problem gambling. For example, one small study assessed the outcome of
problem gamblers treated with a combination of motivational enhancement therapy (MET)
(one to three sessions), CBT (ten to fifteen sessions) and, finally, relapse prevention (two
sessions), all administered individually and in a face-to-face setting. The nine problem
gamblers assessed all showed a decrease in SOGS scores following the completion of
treatment (Freidenberg, Blanchard, Wulfert, and Malta, 2002).
The authors acknowledged that their study was limited by the small number of participants
and the lack of control group. In addition, follow-up was limited to immediately post-
treatment which does not allow any investigation into how long the treatment gains lasted.
Furthermore, there was no attempt to separate the three treatments in any way in order to
assess if any one treatment was more effective than the other, or whether they were most
effective in combination.
Another study with a larger sample size was conducted with 99 problem gamblers randomly
assigned to one of four treatment groups:
 Cognitive therapy (six individual treatment sessions over 8 to 10 weeks)  Behavioural therapy (as above)  Motivational enhancement therapy (as above)  Minimal intervention (one 90 minute feedback session consisting of practical advice for stopping gambling, including a booklet of practical strategies plus the feeding back of findings from baseline assessment). (Toneatto & Gunaratne, 2009) Over 90% (n=92) of the participants were followed-up immediately post-treatment and 74% (n=73) at 12 months. Toneatto and Gunaratne (2009) found that there were no significant differences in the effectiveness of all four treatments. All participants were found to have a lower number of symptoms of the DSM-IV pathological gambling criteria and gambled less frequently, though the changes were mild to moderate. There were no differences in the socio-demographics between groups. In a more recent outcome study of 502 problem electronic gaming machine gamblers, participants were assigned (not randomly) to CBT-only (n=313) and CBT plus exposure and response prevention (ERT) treatment groups (n=189) (Jimenez-Murcia, et al., 2012). ERT is a form of in-vivo exposure therapy. Participants were exposed to stimuli that elicited the urge to gamble, for example, a gaming machine. Participants remained with the stimuli until the Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
urge to gamble dissipated. The intensity of the stimuli was gradually increased to reflect real-
life scenarios, for example, being in a gambling venue with accessible cash. The treatment
for each of the two groups consisted of 16 weekly group therapy sessions. It was found that
CBT was an effective treatment for problem gambling, with positive results evident
immediately post-treatment and at the six-month follow-up. However, due to the high
attrition rates seen in the CBT + ERT group (53.4%) compared with the CBT group (29.7%),
this treatment was concluded not to be effective (Jimenez-Murcia et al., 2012).
The authors commented that the high attrition rates may be due to the extra demands made on
those in the latter group, for example, the CBT + ERT group had significantly more
homework than the CBT-only group. They suggested that making the treatment less complex
may result in lower attrition, as seen in the CBT-only group. Conversely, the authors also
stated that this assumed that participant drop-out was a negative result, indicating relapse.
This may not be the case as drop-out may also indicate that the participant felt that they no
longer needed treatment as their problem had been resolved. Due to the lack of
randomisation, lack of control group and limitations in following-up participants, the authors
recommended that more research was required to support these results.
In a study comparing treatment outcomes of 126 problem gamblers 12-months following
either a CBT or modified Gamblers Anonymous (GA) intervention, Toneatto & Dragonetti
(2008) found there to be no difference between either group. Both treatments resulted in
significantly reduced gambling frequency and expenditure 12 months post-treatment. Of
those who completed follow-up at 12 months, approximately half reported being abstinent
from gambling. At the 12-month follow-up, of those who continued to gamble,
approximately one quarter still met the criteria for pathological gambling, according to the
DSM-IV criteria. Less than half of the initial sample completed 12-month follow-up
measures (n=81), presenting a limitation in this study.
Motivational interviewing

Motivational interviewing is a form of brief treatment that typically includes limited
interaction time between the therapist and the participant and the use of self-help tools such as
workbooks. Motivational interviewing focuses on building the desire of the participant to
change by using five therapeutic guidelines: expressing empathy, highlighting inconsistencies
between a participant's behaviour and their goals, avoiding arguing or confronting the
participant, reinforcing accurate and correcting inaccurate perceptions whilst going along with
any resistance, and endorsing self-efficacy (Miller & Rollnick, 2002; Miller, Zweben,
DiClemente & Rychtarik, 1992). It is used for treating problem gambling due to the fact that
the majority of problem gamblers never seek treatment and that those who do often drop out
of treatment early (Slutske, 2006). These features of motivational interviewing also make this
kind of intervention more suitable for telephone helplines and for assisting people in rural
communities where formal treatment is lacking (Hodgins, Currie, Currie & Fick, 2009).
In relation to outcome studies of motivational interviewing interventions, one case study
examined the combination of motivational interviewing with pharmacotherapy (fluoxetine)
and found that this combination of treatment was successful in reducing gambling symptoms
immediately post-treatment and at three-month follow-up (Kuentzel et al., 2003).
Many of the studies that have been conducted assessing the effectiveness of motivational
interviewing have been randomised controlled trials, which are beyond the scope of this
literature review. They are, however, discussed in detail in the previously mentioned
literature review within the main report for the effectiveness study of brief interventions.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Multimodal treatment

Many general addiction treatment facilities offer multimodal treatment for problem gambling.
Multimodal treatment is where multiple modalities of therapy are used to treat a patient's
issues. These can include individual face-to-face therapy, group therapy, family group
therapy, education, financial counselling, relationship counselling, and GA (twelve-step)
therapy (Stinchfield & Winters, 2001). Commonly, not only is a patient's urge to gamble
addressed but also other areas of life that may have been affected by their gambling.
Multimodal treatment is premised on the complex nature of problem gambling and the impact
that it can have on multiple areas of one's life including physical health, mental health, family
relationships, and one's financial situation.
Korn and Shaffer (2004) support the view that problem gambling requires a multimodal
approach, arguing that pathological gambling is best viewed as a syndrome, as it often
presents with multiple symptoms including depression, anxiety, and alcohol and substance
abuse. However, treatment programmes which involve interventions over multiple problem
areas are typically not run in specialised gambling centres and vary in terms of quality and
implementation of best practice principles (Griffiths & MacDonald, 1999).
Outcome studies on multimodal treatment have reported positive findings. Blackman,
Simone and Thoms (1989) studied 155 participants admitted to multimodal treatment at an
American multimodal treatment centre between 1982 and 1985. Treatment length was on
average 5.5 months. Pre- and immediately post-treatment measures were available for
88 participants. When presenting for treatment, 79% of the 88 participants followed up were
gambling daily and 67% rated their gambling problem as severe or very severe. At the
completion of treatment, six percent were gambling on a daily basis and 29% rated their
gambling problem as severe or very severe. In addition, 43% of participants described
themselves as pre-occupied with gambling when beginning treatment, compared to 25% at
treatment termination.
Russo, Taber, McCormick and Ramirez (1984) also reported positive findings from a follow-
up of 60 former inpatients of a multimodal treatment facility. They found that 55% of
participants reported being abstinent from gambling one year post-treatment. However, the
60 participants who responded to requests for follow-up represented only 48% of the sample
that Russo and colleagues were investigating. This low response rate increases the possibility
of the sample being skewed.
A large-scale outcome study assessing the effectiveness of Minnesota's (USA) gambling
treatment programmes was published in 2001 and, due to the large sample size of 568 patients
and the inclusion of standardised pre-, post-, and follow-up measures, represents a high
quality source of information on gambling treatment outcomes (Stinchfield & Winters, 2001).
The Minnesota study included four state-supported out-patient gambling programmes offering
multimodal treatment aimed at cultivating gambling abstinence. Despite the programmes
having a GA grounding, they all offered individual counselling, group therapy, lectures,
discussion groups, financial counselling, and various other forms of therapy. Data were
collected between January 1992 and January 1995, with the number of patients from each
facility being as follows: 171, 187, 138 and 96. The sample comprised 348 males and 220
females, the majority of whom were white and employed, with an average age of 39 years but
ranging from 18 to 74 years. The treatment goals were gambling abstinence and to learn how
to live without gambling.
One of the strengths of the Minnesota study is that assessment batteries were administered
pre-treatment, post-treatment, six-months following and 12-months following treatment.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Questions included gambling frequency, legal problems, recovery attitude, and substance use
issues, amongst others. In addition, the patient's nominated ‘significant other' completed a
questionnaire about the patient's behaviours at the same time points. The SOGS was included
in all assessment batteries in order to measure problem gambling severity. The majority of
patients reported experiencing problems with electronic gaming machines and casino card
games and reported gambling either daily or weekly. All patients met the diagnostic criteria
for pathological gambling at admission to the programme (Stinchfield & Winters, 2001).
Daily gambling rates dropped markedly from 36% of patients to between one and three
percent following treatment. Abstinence rates increased markedly from pre-treatment to
immediately post-treatment but then declined at follow-up. SOGS scores also showed a large
shift from the majority (87%) of patients scoring over five on the measure, to less than one-
quarter at the six- and 12-month follow-up points. Multivariate analysis indicated there was
significant improvement in gambling frequency and SOGS score post-treatment and over the
follow-up periods. In addition, improvements were also seen post-treatment and over follow-
up on other quality of life and psychological measures. However, due to a large amount of
missing cases (over 50% of the sample by the 12-month follow-up) it is difficult to directly
compare percentages. The authors also noted that it is difficult to judge whether to include
missing data as treatment failures or to exclude it from the analyses (Stinchfield & Winters,
2001).
General counselling
General non-specific counselling or clinical social work is commonly used by treatment
providers across the problem gambling sector (MacDonald, Sheldon & Gillespie, 1992;
Smith, Thomas & Jackson, 2004; Jackson, Thomas, Thomason, Borrell, Crisp, Ho, Holt &
Smith, 2000). Those who use these non-specific techniques are often trained counsellors,
trained social workers, or individuals who hold some other form of qualification which
permits them to work in the industry. They use counselling techniques drawn from a range of
strategies and sources and can also be expected to liaise with courts and provide budgeting or
financial advice (Smith et al., 2004; Delfabbro, 2011).
Jackson and colleagues (1997) examined the type of therapy that counsellors believed had the
more beneficial effect on their clients' gambling problems. Counsellors who had the best
rates of problem resolution with their clients were more likely to report using a combination
of the following:
 Client-centred humanistic psychology  Solution-focused counselling  Assessment of the client's readiness to change  Goal setting Outcome studies on non-specific gambling counselling are few. Jackson and colleagues (2000) conducted an evaluation of the Victoria, Australia ‘BreakEven' counselling service and found that 43% of clients reported experiencing a full or satisfactory resolution of their general problems following treatment and 41% reported a partial resolution of their general problems. In terms of their specific problems with gambling, 70% of clients reported that their situation had improved, with over half of these reporting that it had improved significantly. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Natural recovery

Natural recovery refers to moving from having a gambling problem to no longer having a
problem without accessing any formal treatment services. Obviously, this is not a form of
structured intervention but is discussed in this review as it is a path followed by many
problem gamblers. This is evidenced by the lack of problem gamblers that seek formal
treatment, with international research showing that between six percent and 17% of problem
gamblers ever access help from formal treatment providers (Productivity Commission, 2010;
Slutske, 2006; Suurvali, Hodgins, Toneatto, & Cunningham, 2008). In addition, the steady or
declining rates of problem gambling prevalence and incidence across many populations
indicate that problem gambling is not a lifelong chronic affliction for many people (Ministry
of Health, 2009; Abbott & Volberg, 1996). Natural recovery could be used to inform more
formal interventions.
There is a lack of literature on natural recovery from problem gambling, most likely due to
the difficulty in studying this phenomenon. In a national longitudinal study of 143 problem
and non-problem gamblers, and smaller in-depth study seven years later, Abbott and
colleagues reported that of those considered probable pathological gamblers at the first data
collection point, less than one quarter remained in this category seven years later. The
greatest change was noted for people with less serious problems. None of those probable
pathological gamblers reported seeking formal treatment of any kind in those intervening
years. The authors argue that this indicates a high rate of natural recovery amongst those with
serious gambling problems (Abbott, Williams and Volberg, 1999; 2004).
Shaffer and Hall (2001) assert that problem gambling is a dynamic disorder, with those
affected moving in and out of risk categories over time. When looking at casino employees (a
group with a high risk of developing gambling problems) they found that those with severe
gambling problems were very likely to have moved to having less serious problems one year
later, rather than remain in that high risk category, despite not undergoing any formal
treatment (Shaffer & Hall, 2001).
In a study by Hodgins and Peden (2005), 63 pathological gamblers were followed for a period
of three and a half years in order to explore the process of natural recovery. Participants
completed baseline measures of gambling severity (SOGS) and gambling frequency. Of the
63 participants at baseline, 40 completed measures at follow-up, with the mean follow-up
period being 40.2 months. At follow-up, twelve participants considered themselves to no
longer have a gambling problem. Of these twelve, seven were abstinent and five were
gambling in a controlled manner. Two members of the abstinent group and two members of
the controlled gambling group had received formal treatment for their gambling problem
since baseline measures were taken.
Those who reported no longer having a gambling problem but still participated in controlled
gambling generally had the lowest SOGS scores of the sample at baseline and those who
reported being abstinent from gambling had the highest SOGS scores. Hodgins and Peden
(2005) argue that this shows that different treatment goals may be necessary for different
severities of problem gambling, with those experiencing more negative effects requiring a
more absolute goal such as abstinence. Of the 28 participants who remained pathological
gamblers, eight had received formal treatment, four had received minimal treatment and
sixteen had received no treatment since baseline measures were taken.
Another study investigating students aged 18 to 29 years showed that over an 11-year period
the prevalence of problem gambling stayed the same but individual gambling problems were
characterised as being episodic and dynamic (Slutske, Jackson & Sher, 2003).
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Natural recovery was also explored in two national studies conducted in the United States.
Data from the Gambling Impact and Behavior Study (GIBS) (N=2,417) and the National
Epidemiologic Survey on Alcohol and Related Conditions (NESARC) (N=43,093) were
examined by Slutske (2006). Of the 21 participants in the GIBS study who reported ever
having had a gambling problem, 18 no longer met the DSM-IV criteria for pathological
gambling and nine reported having no symptoms in the past 12-months. Only two
participants who reported ever having a gambling problem also reported seeking treatment.
Of the 185 participants in the NESARC who reported a lifetime history of gambling
problems, 111 (63%) no longer met the DSM-IV criteria for pathological gambling and
70 (36%) had no symptoms in the past 12-months. Of this group of 185 lifetime pathological
gamblers, only 22 (9.9%) had ever sought treatment. However, in this group there was a
strong relationship between the likelihood of seeking treatment and the severity of the
problem. Of those who had 5, 6, 7, 8, 9 and 10 criteria for pathological gambling, 6%, 4%,
6%, 17%, 31% and 76% respectively sought treatment. When examining the data as a whole,
the author concluded that 33% to 36% of those who experience pathological gambling at
some time in their lives will recover naturally (Slutske, 2006).
Therapist effect

The ‘therapist effect' refers to a person experiencing a positive outcome as a result of
treatment, due to the relationship developed with the therapist, not the actual treatment that is
administered. This theory may be supported by the fact that it has been concluded that most
psychological interventions are successful (e.g. Lambert & Bergin, 1994; Smith, Glass &
Miller, 1980; Stiles, Shapiro & Elliot, 1986). In a meta-analysis of 69 studies that examined
therapeutic alliance and patient outcome Martin, Garske and Davis (2000), found that it is the
alliance between therapist and patient that has a greater effect on the patient's outcome than
the intervention chosen. They concluded that it is, thus, more important to focus on the
therapeutic alliance than which intervention to use. However, research in this area is lacking.
Pharmacotherapy

Studies examining the effectiveness of pharmacotherapy on problem gambling are, for the
vast majority, randomised controlled trials. Hence, these studies are not included in this
literature review. This section will examine case studies and trials of pharmacotherapy for
treating problem gambling that do not contain either control groups or randomisation.
In a study by Hollander and colleagues (1998), 16 problem gamblers recruited through
professional referrals and the media underwent a 16-week drug treatment trial. Measures
completed at baseline included the Clinical Global Impression (CGI) pathological gambling
scale, measuring pathological gambling severity, and a modified Yale-Brown Obsessive
Compulsive Scale for pathological gambling, measuring urges and behaviour. The first eight
weeks of the single-blind study consisted of participants being administered a placebo. This
was an attempt to minimise any short-term placebo effects and ensuing relapse. The second
stage of the trial was eight weeks of fluvoxamine, a selective serotonin re-uptake inhibitor
often used in the treatment of obsessive compulsive disorder.
Of the 16 participants in the fluvoxamine study, six dropped out during the placebo stage.
The remaining 10 participants all completed the full course of treatment. The authors
reported that seven of the 10 treatment completers responded positively to the treatment. The
seven that responded positively all were rated ‘very much improved' or ‘much improved' on
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
their CGI score and had a mean improvement of 86% on their Yale-Brown gambling
behaviour score. In addition, all seven were abstinent at the end of treatment. However,
Hollander et al. (1998) warn that these results must be interpreted with caution due to the lack
of double-blinding, randomisation and control groups.
In another pharmacotherapy study, Black (2004) conducted an open-label trial of bupropion, a
norepinephrine-dopamine re-uptake inhibitor that is commonly used to treat depression and
assist with smoking cessation. Ten problem gamblers participated in the eight week trial and
at the conclusion, seven of the ten were rated as being much or very much improved in their
CGI score. However, Black (2004) includes caveats, such as the lack of blinding and placebo
control.
Conclusion

The current literature review examined outcome studies and case studies that investigated the
outcomes of various treatment types in treating problem gambling. Due to the fact that these
studies do not adhere to the rigours inherent in randomised controlled trials, there are
limitations. These limitations include small sample sizes, a lack of control groups, a lack of
blinding and the use of non-standardised measures. Due to these limitations, results must be
interpreted with caution.
As can be seen from the variety of treatments available and lack of strong evidence pointing
to any one being significantly more effective than another, more research is needed in this
area. The current study aims to evaluate the outcome of a group of participants who called a
national gambling helpline, received standard helpline care and were then referred to face-to-
face gambling help services for additional treatment. The array of treatment services
available across New Zealand means that different participants will have encountered
different intervention modalities, similar to those discussed in the current literature review.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
RESEARCH METHODOLOGY
Ethics approval
All participant materials (i.e. survey questionnaires, information sheets and consent forms) and other relevant documents were submitted to the Multi-Region Health and Disability Ethics Committee (a Health Research Council accredited human ethics committee) which considers the ethical implications of proposals for research with humans where participants are asked questions in relation to their health. The ethics approval for the study was granted on 3 June 2009 (Appendix 1). The Ethics Committee was kept apprised of any changes as the study progressed. During the research the following measures were taken to protect the identity of the participants:  All participants were allocated a code by the research team to protect their identities  No personal identifying information has been reported.  Participants were informed that participation in the research was voluntary and that they could withdraw at any time, prior to data reporting. Study design

3.2.1
Study aims
The main aims of the study are to:  Ascertain if there are any differences in participant outcomes between those who only access telephone assistance for gambling problems in comparison with those who also access professional counselling services additional to the initial telephone intervention  Identify client characteristics associated with treatment outcome.
3.2.2

This study was a (prospective) cohort study of problem gamblers calling the Gambling
Helpline and receiving the standard treatment. Recruitment occurred in parallel with, and
using the same entry criteria, as a randomised controlled trial of four different helpline
interventions, the Control (or "Treatment as Usual") arm of which was recruited into the
cohort.
3.2.3
Study population
As described throughout this report, the data discussed here represent an extension of the ‘standard care' control group from a large four group randomised controlled trial recruited from callers to the Gambling Helpline. The extended recruitment for the standard care group Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
was designed so that the group could function as a stand-alone outcome study, describing and
documenting the outcomes for those accessing help services for their gambling problems in
New Zealand. In the randomised controlled trial, 116 participants were recruited to the
standard care control group. An additional 34 participants were recruited to the group to
increase the group size to 150 for the purpose of this outcome study. Recruitment was
terminated at this point due to funding and time constraints.
All callers to the helpline initially received brief non-directive counselling to identify
presenting concern/s and establish rapport. If the caller met eligibility criteria they were
asked if they would like to participate in the study. Immediately after consenting to take part
in the study, participants underwent an initial baseline assessment (detailed in section 3.2.6)
and then received their treatment which was delivered by telephone by a trained Gambling
Helpline counsellor.
The participants received a manualised version of the helpline's standard care. This standard
care included brief screening, reflective listening to clients' concerns and, in the instance of
first time callers or regular callers who were experiencing persistent difficulties, referral to
face-to-face problem gambling counselling services and/or suggestions for self-care.
Treatment comprised a single session. The protocol was developed with staff to ensure it was
functional and similar to their normal practice; it did not contain any elements of motivational
interviewing techniques. Additionally, participants were offered an information pack
(relevant information pamphlets, for example detailing venue self-exclusion processes, or
budgeting advice).
While the protocol for the Gambling Helpline counsellors was manualised and assessed for
compliance and consistency, as part of the protocol included referrals to face-to-face problem
gambling counselling services, there is variability in the approach to face-to-face delivery
across service providers that is not described or accounted for in the results presented in the
current report. Nonetheless, these results provide a reasonable account of the outcomes for
people seeking help for their gambling problems in New Zealand.
3.2.4
Eligibility criteria
The inclusion criteria were:  Minimum age of 18 years  Perception of having a gambling problem  Willingness to: o Read a short workbook (to ensure reading ability) o Have calls recorded o Provide follow-up data on gambling o Provide the name of collateral/s. Present or past involvement in treatment or mutual help groups for gambling or other mental health problems was documented and did not preclude participation. Callers were excluded from the study if:  They were considered by the counsellor to be actively psychotic  They required immediate crisis or police intervention because they posed a serious risk to themselves or others. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Setting and location

The study took place at the Gambling Helpline, Auckland, New Zealand. Treatment was
delivered by trained Gambling Helpline counsellors. As the treatment was delivered by
telephone, participants were based throughout New Zealand. Recruitment and delivery of
treatment occurred from August 2009 to May 2011.
Follow-up assessment calls were made by telephone by trained university research assistants
from Auckland University of Technology (AUT), Auckland, New Zealand.
3.2.6
Schedule of assessments

Initial assessment

The initial assessment was conducted with the participant by a helpline counsellor prior to the
participant receiving the treatment. Due to the length of the initial assessment, some of the
baseline initial assessment was conducted by an AUT researcher within seven days of a
participant receiving the intervention. This is detailed at the end of this section.
Gambling/problem gambling history, impacts and past help-seeking
A brief gambling history was obtained including length of gambling problem; type/s of
gambling causing problems; number, nature and outcomes of past attempts to quit or reduce
gambling; and past treatment and mutual help involvement. The impacts of gambling on
financial status, employment, family and other relationships, criminal offending and general
health (adapted from Abbott & Volberg, 1992; Abbott, 2001) were also assessed.
Problem gambling
The nine-item Problem Gambling Severity Index (PGSI) (Ferris & Wynne, 2001) was used to
measure severity of gambling problems. It was administered in both a past 12-month and a
past three-month time frame (reported as PGSI-12 and PGSI-3, respectively). The PGSI has
been validated against clinician-derived DSM-IV pathological gambling diagnoses and other
problem gambling measures including the widely used South Oaks Gambling Screen/South
Oaks Gambling Screen-Revised (SOGS/SOGS-R) (Abbott & Volberg, 2006).
Comorbidity and substance use
A brief version (10-item scale) of the Drug Abuse Screening Test (DAST; Skinner, 1982) was
administered to assess drug abuse. The DAST has very good internal reliability in samples of
substance abusers and psychiatric patients and correlates strongly with a number of drug use
measures (Cocco & Carey, 1998).
To identify hazardous alcohol consumption or active alcohol use disorders (including alcohol
abuse or dependence) a brief version (AUDIT-C, three-item scale) of the Alcohol Use
Disorders Identification Test (AUDIT; Saunders et al., 1993) was administered. In a review
of research using the AUDIT and shortened versions, the AUDIT-C was reported as showing
promise in being time-efficient and accurate when compared with full AUDIT results (Reinert
& Allen, 2002).
Participants were also asked about lifetime and current tobacco use and any previous success
at quitting a problematic behaviour (i.e. smoking, alcohol, other drugs and other behaviour).
The Kessler-10 (K-10) questionnaire was included to provide a continuous measure of
general psychological distress that is responsive to change over time. The K-10 has been well
validated internationally. Its brevity and simple response format are attractive features. It
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
also produces a summary measure indicating probability of currently experiencing an anxiety
or depressive disorder (Kessler & Mroczek, 1994).
Quality of life
Quality of life was assessed by the WHOQoL-8, an eight item version of a widely used
measure. This short form has been used in a number of countries, is robust psychometrically,
and overall performance is strongly correlated with scores from the original WHOQoL
instrument (Schmidt, Muhlan & Power, 2005).
Treatment goal
Participants were asked whether their goal was to stop all forms of gambling, stop only
problematic forms of gambling, or to gamble in a non-problematic manner.
Self-efficacy
A simple rating was employed to assess belief in likelihood of a participant achieving their
treatment goal (0 "not at all confident" to 10 "extremely confident") in the next six months.
Motivation and perceived control over gambling
Treatment goal motivation was measured on the same 0 to 10-point scale ("not at all" to
"extremely"). Participant-rated sense of control over gambling was assessed using a similar
0 to 10-point scale ("no control" to "total control").
Socio-demographics
Age, gender, ethnicity, marital status, highest educational level, employment status and area
of residence data were collected.
Follow-up baseline questions
Within seven days of the initial assessment and treatment delivery, an AUT researcher asked
some follow-up baseline questions of each participant. These included:
Gambling/problem gambling history
A detailed timeline of types of gambling, frequency and money spent gambling over the past
two months was administered (based on Sobell & Sobell, 1992). Participants were provided
with memory cues such as recent holidays and news events to facilitate retrieval of this
information.
Comorbidity and substance use
The mood module of the Primary Care Evaluation of Mental Disorders (PRIME-MD, Spitzer
et al., 1994) was administered to provide diagnoses of major depressive disorder, dysthymia,
minor depressive disorder, and alcohol abuse/dependence. This is a structured interview
designed for primary care clinicians and researchers to diagnose these and other current
DSM-IV mental health disorders. It has been validated against the Structured Clinical
Interview for the DSM-IV (Spitzer et al., 1992) and has been administered by telephone and
shown to yield valid diagnoses (Spitzer et al., 1994; Kobak et al., 1997). The use of
psychotropic medication and history of manic episodes was assessed using questions
modified from the Gambling Impact Study (Gerstein et al., 1999).
Socio-demographics
The eight-item New Zealand Index of Socio-economic Deprivation for Individuals (Salmond,
2005) was administered.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Follow-up assessments

Participants were contacted by an AUT researcher to complete a follow-up assessment at
three, six and 12 months post-treatment. At each follow-up assessment, a timeline follow-
back interview was conducted to capture the number of days gambling during the follow-up
period and the amount of money lost on each occasion. Participants were asked whether they
had met their goal (‘not at all', ‘partially', ‘mostly', ‘completely') and what their present goal
and personal sense of control over their gambling were (0 ‘no control' to 10 ‘total control').
At each assessment, participants were asked what other treatment or help, if any, they
received for their problem gambling during the follow-up period. These forms of treatment/
help were listed and, for each, they were asked how often the treatment or help was obtained
(number of occasions) and how helpful it was in reaching their goal (‘not at all', ‘partially',
‘mostly', ‘completely'). At the three- six- and 12-month assessments, in addition to the
previously mentioned assessments, the current tobacco use, gambling impacts, AUDIT-C, K-
10 and WHOQol-8 were re-administered. At the 12-month assessment point, participants
were also re-administered the DAST and PRIME-MD mood module as well as being asked to
reflect on their overall experience during the past 12 months in seeking and receiving help for
gambling and making changes in their lives. They were also asked about the cultural
appropriateness of the help they received (linked to source/type of help) and, if inappropriate,
what could be changed to make it more appropriate. They were invited to comment on other
changes that could make help more accessible, appropriate and effective.
3.2.7

Training for gambling helpline counsellors included practice in introducing the project,
recruitment of participants, and the initial assessment questions. The training also
incorporated how to use the protocol. All training included pilot interviews with volunteers
that were digitally recorded and assessed for compliance and consistency by Dr Sean Sullivan
and Professor David Hodgins (two of this report's authors). The training included additional
ad hoc sessions, particularly at the beginning of the study, to address any issues.
Dr Sullivan also trained the AUT researchers who conducted the follow-up assessments. The
training included identification of risk level of participants and how to safely intervene when
participants expressed symptoms of risk or suicidal ideation.
3.2.8
Treatment integrity and fidelity
To assess how well the counsellors delivered the intervention (i.e. to assess treatment integrity and fidelity), approximately 20% of telephone calls (including intervention delivery) with participants were randomly digitally recorded. The recordings were subsequently (usually within one month) listened to and assessed by Dr Sean Sullivan who provided personal feedback and, where required, additional training to counsellors. Approximately one-third of the recordings were also listened to by Professor David Hodgins in order to assess reliability of the first assessment of the recordings. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Data analyses
Participant profiles Baseline data utilised for the participant profiles are primarily descriptive. These include frequencies of demographics, gambling, mental health and behavioural measures, as well as means and standard deviations or medians and ranges, where appropriate. Baseline data for profiles include:  Demographics: o Age group, ethnicity, gender, marital status, employment status, highest educational level, area (town/city) of residence, total household income  Gambling-related, mental health and other behavioural measures including: o Number of days gambled, amount of money spent o Problem Gambling Severity Index (PGSI), control over gambling, gambling o Psychological distress (Kessler-10), Mood Module of Primary Care Evaluation of Mental Disorders (Prime-MD) o Quality of life (WHOQoL-8), individual deprivation o Alcohol abuse or dependence (AUDIT-C), Drug Abuse Screening Test (DAST), tobacco use Trends Follow-up data from three-, six- and twelve-months have been presented descriptively, examining measures at each time point and change since treatment, providing frequencies, means and standard deviations, or medians and ranges, where appropriate. Data for profiles by time point (baseline, three-, six- and 12-months) include:  Follow-up rates of participants  Gambling measures o Number of days gambled, amount of money spent o Problem Gambling Severity Index (PGSI), control over gambling, gambling  Additional assistance from problem gambling intervention services, other services, family or friends
Predictors of further utilisation of formal gambling treatment services
The participants have been classified into those who utilised, and those who did not utilise,
any additional formal gambling treatment services (Problem Gambling Foundation, Salvation
Army Oasis Centres, Gamblers Anonymous, any other problem gambling support services,
and additional gambling helpline contact) within the three months to the first follow-up
assessment.
Baseline characteristics were examined for any differences in formal service utilisation in the
first three months (Yes/No) using logistic regression to identify predictors of uptake of formal
services. Baseline characteristics examined included those detailed previously in Participant
profiles.
Multivariate models were then examined utilising all baseline characteristics with p values
≤ 0.20 to identify the best combination of variables to predict further utilisation of formal
gambling treatment services and, therefore, to adjust for any confounding effects.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of successful problem gambling outcomes through helpline standard care Successful problem gambling outcome measures were derived from the following factors:  Number of days gambled, amount of money spent  Control over gambling  Problem Gambling Severity Index (PGSI)  Goals met of treatment success (Gambling-Quit or improved) Data for the past 12-month PGSI measure was only collected at the baseline and 12-month assessments, therefore, the statistical analysis for this measure was an unweighted linear mixed effects model of the 12-month assessment measures, adjusting for baseline measures. The number of days gambled, amount of money spent and control over gambling, which were measured at baseline, three months, six months and 12 months, were analysed using weighted linear mixed effects models for the three follow-up time points, adjusting for baseline measures. The treatment success measures which were dichotomous measures (i.e. true or false), and measured at all four time points, were analysed using a weighted logistic mixed effects model. Baseline characteristics detailed previously in Participant profiles were examined to identify any significant associations with the positive problem gambling outcomes. Additionally, the following factors were examined:  Treatment assistance o Utilisation of referral to formal services (as defined above) o Utilisation of any problem gambling intervention services o Utilisation of any other services o Utilisation of family or friends assistance Multivariate models were then examined utilising all baseline characteristics with p values ≤ 0.20 to identify the best combination of variables to predict successful problem gambling outcomes through helpline standard care and, therefore, to adjust for any confounding effects. Whilst power calculations were not specifically computed for this outcome study (they were computed for the randomised controlled trial), post-hoc power calculations for PGSI as an outcome measure indicated there was enough power for a difference of +/-3.5 PGSI units. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
This chapter details the results of data analyses obtained from the 150 participants who received the helpline's standard care intervention. Section 4.1 contains the descriptive statistics detailing socio-demographic characteristics of the participants, their gambling participation data, and outcome data showing time trends over the assessment points. Section 4.2 details the predictors for utilisation of formal (professional) treatment services for problem gambling within three months of receiving the study intervention. Sections 4.3 to 4.7 detail predictors of successful problem gambling outcomes focusing on PGSI, number of days gambling per month, money lost gambling per day, control over gambling, and treatment success. Descriptive statistics

4.1.1
Socio-demographic characteristics at each assessment point
One hundred and fifty participants received the helpline's standard care intervention. Ninety-nine participants received a follow-up assessment at 12-months representing a 66% retention rat Table 1: Number and percentage of participants at each assessment point Assessment point
Baseline
12 months
Percentage of baseline
Socio-demographic characteristics of the participants are detailed in Appendix 2, Table 2.1.
Visual examination of percentages of socio-demographic characteristics over time revealed
no major differences. Thus, differential attrition is considered not to be an issue.
4.1.2
Baseline socio-demographic characteristics by baseline and follow-up data
availability and whether formal assistance accessed in first three months
Socio-demographic characteristics of the participants are detailed in Appendix 2, Table 2.2. The characteristics are detailed as total number and percentage of participants at the initial baseline assessment; the number and percentage with any follow-up assessment data; and the number and percentage, at the three-month assessment, who responded (‘yes' or ‘no') to the question asking whether they had sought assistance from a professional problem gambling treatment provider (other than the initial gambling helpline intervention) during the last three months. Of the 150 participants, follow-up assessment data were available for 130 (87%), and 121 (81%) responded to the question regarding accessing professional assistance for their problem gambling during the previous three months. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
There were slightly more females (57%) than males (43%) at the baseline assessment. This
profile differed slightly from the overall gambler caller profile to the Gambling Helpline from
2009 to 2011 (the period of study recruitment) where the gender split across the years was
47% to 49% female and 52% to 53% male (Gambling Helpline, 2012).
Just less than half (49%) of the participants were partnered, either married (23%) or in a de
facto relationship (26%). Participants represented all adult age groups with the larger
percentages being in the 25 to 34 year (25%) and 35 to 44 year (27%) age groups.
A majority of participants identified primary ethnicity as either Maori (43%) or European
(42%) with 10% identifying primarily as Pacific and five percent as Asian/Other. The
ethnicity profiles differed slightly from the overall new gambler caller profile to the gambling
helpline from 2009 to 2011 where: 28% to 35% (in each individual year) were European,
18% to 23% were Maori, seven percent to nine percent were Pacific, six percent to nine
percent were Asian, and 25% to 27% were Other/multiple ethnicity (Gambling Helpline,
2012). Based on this, the study recruited more participants of European, Maori or Pacific
ethnicity and less of Asian/Other ethnicity than the general Gambling Helpline gambler caller
profile at that time.
One-quarter (27%) of the participants had no educational qualification, one-third (32%)
reported being educated to secondary school level, one-fifth (18%) reported having a trade or
technical certificate, and the remainder had a tertiary or professional qualification.
Two-fifths (44%) of the participants were in full-time employment and a further 13% had
part-time employment. Just less than half (46%) of the participants reported gross family
income in the past 12-months as either < $20,000 (28%) or between $20,001 and $30,000
(18%). Almost one-quarter (23%) of participants reported gross annual family income of
$30,001 to $50,000, two-fifths (21%) between $50,001 and $100,000, and 11% reported more
than $100,001.
Participants were recruited from around the country with a larger proportion residing in the
three major cities of Auckland, Christchurch and Wellington (Appendix 2, Table 2.3).
4.1.3
Primary mode of problem gambling

Eighty-nine percent of the participants reported electronic gaming machines to be their
primary mode of problem gambling at the baseline assessment with a majority (76%) citing
machines in pubs, nine percent citing machines in clubs and three percent citing machines in
casinos. The remaining participants cited track betting, casino table games, card gambling,
keno, sports betting or other forms of gambling as their primary problem mode (Appendix 2,
Table 2.4).
4.1.4
Time trends
Tabulated data of changes over time are presented in Appendix 2, Table 2.5. For each variable examined, the greatest change was in the first three months from baseline assessment to the three-month assessment3. Thereafter, changes were generally maintained. 3 These data are not adjusted for attrition. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Days gambling per month, expenditure per day and control over gambling
presents median values for self-reported number of days gambling per month and
gambling expenditure per day, and self-reported control over gambling (10-point scale from
0 ‘no control' to 10 ‘total control) across the four assessment points. Number of days
gambling and expenditure were both substantially lower at the three-month assessment
compared to the baseline assessment, with the improvement maintained at the subsequent
assessments. Self-reported control over gambling median value increased from 3 at the
baseline assessment to 7 at the three-month assessment. A further improvement (median
score 8) was noted at the six- and 12-month assessments.

Figure 1: Days gambling, dollars lost and control over gambling by assessment point
Gambling goal At each assessment point, participants were asked to state their current goal. Goals were to quit all modes of gambling, quit some modes of gambling (problematic modes), gamble in a controlled manner, maintain gambling abstinence, or some other goal. At the baseline assessment, a majority (61%) of participants reported their goal to be to quit all forms of gambling. This decreased to 38% at the three-month assessment with a gradual decrease continuing at the six- and 12-month assessments. Mirroring this finding, only five percent of participants reported their goal to be to maintain gambling abstinence at the baseline assessment, with an increase to 21.5% at the three-month assessment and continued gradual increases at the six- month assessment, which then stabilised. Participants whose goal was to gamble in a controlled manner showed a slight increase across time, from 13% of participants at the baseline assessment to 19% at the 12-month assessment. Overall, across time, the percentage of participants reporting their goal to be to quit some modes of gambling decreased slightly from 20% to 16%. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Figure 2: Gambling goal by assessment point Problem gambling severity index Participants' PGSI scores improved over time. When the PGSI-12 (past 12-month time frame) was administered to participants, the median score at the 12-month assessment was almost half the score at the baseline assessment (9 vs. 17). The PGSI-3 (past three-month time frame) scores show that the largest decrease in score was noted at the three-month assessment from baseline and that this was maintained at the six- and 12-month assessments . Figure 3: PGSI by assessment point At the baseline assessment, 97% of participants were categorised as problem gamblers using the PGSI-12, with 3.5% in the moderate risk group and no participants categorised as low risk or non-problem gamblers. At the 12-month assessment, 58% of participants remained categorised as problem gamblers with one-quarter (26%) in the moderate-risk group, 12% in the low-risk group, and five percent categorised as non-problem gamblers . Thus, although there was an overall decrease in PGSI-12 score over time (as detailed above) and improvement made in gambling outcome, four-fifths (83%) of the participants remained problem gamblers or at moderate risk for developing problem gambling at the 12-month assessment. However, almost one-fifth (17%) had improved significantly to a low risk or non-problem gambling level. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Figure 4: PGSI-12 categorisation by assessment point Psychological distress Over time, the percentage of participants with a high level of psychological distress (measured via Kessler-10) decreased from the baseline assessment (56%) to the 12-month assessment (10%) although the greatest decrease was noted at the three-month assessment which stabilised thereafter. Conversely, a minority of participants reported psychological distress at a low level at the baseline assessment (3%), which increased to approximately 50% at the three- and six-month assessments, with a further increase noted at the 12-month assessment (63%). Similar percentages of participants reporting moderate psychological distress were noted at the first three assessment points (approximately 40%) with a decrease to 27% at the 12-month assessment Figure 5: Psychological distress by assessment point On the PRIME-MD scale, a substantially lower percentage of participants reported major or minor depressive disorder or dysthymia at the 12-month assessment in comparison with the baseline assessment. For major depressive disorder, over half (58%) of participants reported this mental health issue at the baseline assessment with less than one-fifth (18%) reporting it at the 12-month assessment. This was a decrease of about one-third. A similar finding was noted for minor depression (12% at baseline, 4% at 12-months). A smaller percentage Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
decrease occurred for dysthymia over time (42% to 32%). However, a decrease was not noted in percentage of participants reporting bipolar disorder whereby three percent reported the disorder at the baseline assessment and four percent reported it at the 12-month assessment; due to the small number of participants with bipolar disorder (n=4 at both time points) this finding could be due to attrition and should be treated with caution Figure 6: Major and minor depressive disorder, dysthymia, bipolar disorder by assessment point Substance abuse/dependence There appeared to be a slight decreasing trend over time in regard to participants who smoked tobacco, with 60% reporting smoking at the baseline assessment, and half reporting smoking by the 12-month assessment. A similar trend was not noted for participants reporting alcohol abuse or dependence (measured via AUDIT-C). Sixty-two percent of participants reported alcohol abuse/dependence at the baseline assessment and whilst the percentage gradually decreased over the next two assessment points (55% and 50% respectively), by the 12-month assessment, the percentage had returned to baseline levels (60%) . Figure 7: Alcohol abuse/dependence and smoking by assessment point A majority of participants (77%) showed no drug use problems (measured via DAST) at the baseline assessment, with the percentage increasing to almost all participants (92%) at the 12- Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
month assessment. The percentage of participants showing low, moderate or substantial drug problems at the baseline assessment decreased at the 12-month assessment. Figure 8: Drug abuse by assessment point Life aspects affected by gambling The extent to which gambling had affected aspects of participants' lives (10-point scale from 0 ‘not at all' to 10 ‘very severely) again showed marked improvement from the baseline assessment to the three-month assessment. Slight improvement continued to be noted at the six- and 12-month assessments for the self-rated mean scores for gambling affecting past-month work, social life and family life/home responsibilities. The continued slight improvement was not noted for gambling affecting physical health in the past month, which stabilised from the three-month assessment Figure 9: Aspects of life affected by gambling by assessment point Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Additional assistance At the baseline assessment, 38 participants reported that they had previously received formal (professional treatment services) and/or informal (e.g. family, friend and work colleagues) assistance for their problem gambling, and 23 participants reported that they were currently receiving formal or informal assistance. Eight of the participants were both currently receiving, and had previously received, assistance . Table 2: Previously and/or currently receiving assistance at baseline assessment Currently receiving assistance
Previously received assistance
At the three-month assessment, 60% of respondents reported receiving assistance for their problem gambling from formal and/or informal sources in the past three months (additional to the study intervention). A greater percentage reported receiving assistance from informal sources (39%) than formal services (31%). The percentage seeking assistance decreased at the subsequent assessments to 19% and 25% for formal and informal assistance at the 12-month assessment, respectively. Figure 10: Assistance received for gambling problem by assessment point Ten percent or less of respondents at the three-month assessment reported accessing any individual formal treatment service for their problem gambling with this percentage, although fluctuating, remaining similar at the six- and 12-month assessment points. Some respondents accessed formal treatment services across the assessment points. There was no apparent preference for treatment provider although only one respondent accessed an online/internet provider for their problem gambling, reported only at the three-month assessment . The median number of sessions, for individual treatment services accessed by respondents, was six or less at the three-month assessment, 12 or less at the six-month assessment and 24 or less at the 12-month assessment. This latter assessment covered the previous six-month period whilst the three- and six-month assessments covered a previous three-month period. The maximum number of sessions attended by any respondent was 12 at the three- and six- Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
month assessments and 24 at the 12-month assessment and represents approximately weekly attendance at a service . One-fifth (22%) or less of respondents at the three-month assessment reported receiving informal assistance from a partner, family member or friend for their problem gambling. Generally, the percentage of respondents reporting informal assistance from any single source decreased at subsequent assessments. Some respondents sought informal assistance across the assessment points. Respondents appeared slightly more likely to seek assistance from partners or family members rather than from friends or other people . Some respondents sought assistance from more than one source. For example, at the three-month assessment, 31% of participants reported receiving assistance from formal sources compared with a maximum of 10% seeking assistance from any one organisation. This finding continued to be apparent at the six- and 12-month assessments and was also apparent for informal assistance received . Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 3: Individual additional assistance access Assessment point
At any time
12 months
Formal treatment service
Gambling Helpline
Problem Gambling Foundation
Salvation Army Oasis Centres
Gamblers Anonymous
Other problem gambling service
Online/internet service
Any of above formal services
Informal assistance
Family member
Any of above informal assistance
Other person
Any formal/informal
Total N
N Missing
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 4: Formal service access by assessment point - number of respondents, median, minimum and maximum number of sessions Assessment point
12 months
No. Median
Gambling Helpline
Problem Gambling
Foundation
Salvation Army Oasis
Centres
Gamblers Anonymous
Other problem gambling
support service
Online/internet service
# One respondent did not provide number of sessions Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of utilisation of formal treatment services

This section presents data pertaining to associations between use of formal treatment4 services
in the first three-months (i.e. whether the participants accessed formal gambling treatment
services in the first three months after receiving their helpline treatment) and socio-
demographic characteristics, as well as gambling and related behaviours and other baseline
covariates.
4.2.1
Uptake of formal services by socio-demographic characteristics
details the odds ratios of uptake of formal gambling treatment services in the first three-months and associations with socio-demographic characteristics which attained a level of statistical significance. Males had statistically significant (p=0.02) greater odds for uptake of formal services in comparison with females. This significant association was retained in the multivariate analyses controlling for confounding factors, with greater odds of 2.85 times. Table 5: Odds ratios for uptake of services by gender Variable
Category
Uptake of
Univariate odds ratio
Adjusted odds ratio†
services
OR (95% CI)
† Adjusted by previous assistance received for gambling problem and baseline treatment received for
mental health in past year.
Variables which did not achieve a level of statistical significance are presented in Appendix 3,
Table 3.1. These related to baseline marital status, age, ethnicity, employment status, highest
educational qualification, gross annual family income, and area of residence.
4.2.2
Uptake of formal services by baseline gambling and related behaviours
details the odds ratios of uptake of formal gambling treatment services in the first three-months and associations with baseline gambling and related behaviours which attained a level of statistical significance. Participants whose had previously received assistance for a gambling problem had statistically significant (p=0.03) greater odds for uptake of formal services in comparison with participants who had not previously received assistance for a gambling problem. This significant association was retained in the multivariate analyses controlling for confounding factors, with greater odds of 2.95 times. A statistically significant association between participants self-rated control5 over gambling at the baseline assessment and uptake of formal services was noted in the univariate analyses. 4 This includes face-to-face counselling services, Gamblers Anonymous, online/internet treatment services, and additional gambling helpline contact. 5 Control over gambling was rated on a scale of 0 to 10 where 0 = ‘no control' and 10 = ‘total control'. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
However, as a level of statistical significance was not retained in the multivariate analyses, this finding is likely to be due to confounding from one or more other variables. Table 6: Odds ratios for uptake of services by baseline gambling and related behaviours Variable
Category
Uptake of
Univariate odds ratio
Adjusted odds ratio†
services
Prev. assist. for
gambling prob.
Control over
gambling
† Adjusted by gender and baseline treatment received for mental health in past year.
Variables which did not achieve a level of statistical significance are presented in Appendix 3,
Table 3.2. These related to baseline: primary problem gambling mode (dichotomised to EGM
vs. other); self-reported number of days gambling per month and gambling expenditure per
day; length of problem duration; number of days since the last gamble; level of motivation to
overcome gambling problem; current goal (i.e. to quit some or all modes of gambling or to
control gambling); belief in treatment success, and perceived level of difficulty in overcoming
gambling problems; and if assistance was currently being received for a gambling problem.
4.2.3
Uptake of formal services by other baseline covariates
details the odds ratios of uptake of formal gambling treatment services in the first three-months and associations with other baseline covariates which attained a level of statistical significance. Participants who had, at baseline, received treatment for a mental health issue in the past 12 months had statistically significant (p=0.03) greater odds for uptake of formal services in comparison with participants who had not received treatment. This significant association was retained in the multivariate analyses controlling for confounding factors, with greater odds of 2.77 times. A statistically significant association between participants who, at baseline, had been prescribed medication for an emotional, nervous or mental health issue in the past 12 months and uptake of formal services was noted in the univariate analyses. However, as a level of statistical significance was not retained in the multivariate analyses, this finding is likely to be due to confounding from one or more other variables. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 7: Odds ratios for uptake of services by other baseline covariates Variable
Category
Uptake of
Univariate odds ratio
Adjusted odds ratios†
services
Treatment,
mental health
last year
Prescription,
mental health
last year
† Adjusted by gender and previous assistance received for gambling problem.
Variables which did not achieve a level of statistical significance are presented in Appendix 3,
Table 3.3. These related to baseline: problem gambling status (PGSI, past 12-month and past
three-month time frame); psychological distress (Kessler-10); alcohol abuse/dependence
(Audit-C, dichotomised to low risk and high risk); drug abuse/dependence (DAST); suicide
ideation (dichotomised to no ideation and some ideation); major depressive disorder,
dysthymia and minor depressive disorder (PRIME-MD); current tobacco use; quality of life
(WHOQoL); deprivation level (NZDI); treatment for an alcohol or drug problem in the past
12 months; how work, social life, family and home life, and health were affected in the past
month; and legal problems in the past 12 months.
4.2.4
Multivariate model examining interactions

When examining for interactions between variables which retained a level of statistical
significance in the multivariate analyses, a statistically significant interaction (p=0.02) was
noted between gender and receiving treatment for a mental health issue in the past year
(baseline assessment). Females who had not previously received mental health treatment
were less likely to access formal services for their problem gambling. Females who had
received mental health treatment and males who either had or had not received mental health
treatment had greater odds for accessing formal services for problem gambling (4.41 to 6.55
times greater) than females who had not received mental health treatment. Previously
receiving assistance for a gambling problem did not show an interaction with the other
variables

Table 8: Multivariate model of interactions between variables
Variable
Category
Uptake of services
Multivariate odds ratio
OR (95% CI) p-value
Female, no prev. treatment mental health
Female, prev. treatment treatment last Male, no prev. treatment
Male, prev. treatment Prev. assist.
for gambling
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of successful problem gambling outcomes - PGSI

This section presents data pertaining to associations between change in PGSI-12 (past 12-
month time frame) score at the 12-month assessment from the baseline assessment score and
uptake of formal gambling treatment services in the first three-months (section 4.3.1). It also
details data pertaining to associations between change in PGSI-12 score and socio-
demographic characteristics (section 4.3.2), baseline gambling and related behaviours (section
4.3.3) and other baseline covariates (section 4.3.4). As the difference in PGSI scores between
the time points gives an indication of change in problem gambling severity/risk level, these
analyses indicate which variables are associated with likelihood of better (i.e. improved)
problem gambling severity outcomes 12-months after initial helpline contact and treatment.
4.3.1
PGSI-12 change at 12-months by uptake of formal services in first three months
details mean PGSI-12 score change at the 12-month assessment from the baseline score and associations with uptake of formal gambling treatment services in the first three months. There was no significant difference in change in mean PGSI-12 score between participants who sought additional formal treatment in the first three months compared with participants who did not seek formal treatment6. Table 9: PGSI change by uptake of formal services in first 3 months Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Uptake of formal No
services


4.3.2
PGSI-12 change at 12-months by socio-demographic characteristics

details mean PGSI-12 score change at the 12-month assessment from the baseline
score and associations with baseline gambling and related behaviours which attained a level
of statistical significance.
Marital status was statistically significantly associated with mean PGSI-12 score change
(unadjusted values). However, as a level of statistical significance was not retained in the
multivariate analyses, this finding is likely to be due to confounding from one or more other
variables.
In the multivariate analyses, when confounding factors were accounted for, a statistically
significant (p=0.03) association between mean PGSI-12 score change and baseline
employment status was noted. On average, participants who were disabled, had an illness or
were on sick leave, showed a smaller improvement in mean PGSI-12 score (estimated mean
PGSI change -8.06) than other participants (estimated mean PGSI change -11.56 or more).
6 The estimated least squares mean difference between no uptake of formal services, and uptake, was -1.4 which is not a clinically significant difference. Post-hoc power calculations indicated that there was sufficient power for a difference of +/-3.5 to be identified, which would be a clinically significant difference. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 10: PGSI-12 change by socio-demographic characteristics Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Employment
Disabled/illness † Adjusted by baseline NZDI, primary problem gambling mode, and PRIME-MD minor depressive
disorder.
Variables which did not achieve a level of statistical significance are presented in Appendix 4,
Table 4.1. These related to gender, age, ethnicity, highest educational qualification, gross
annual family income, and area of residence.
4.3.3
PGSI-12 change at 12-months by baseline gambling and related behaviours
details mean PGSI-12 score change at the 12-month assessment from the baseline score and associations with baseline gambling and related behaviours which attained a level of statistical significance. A statistically significantly association for number of days since the last gamble and mean PGSI-12 score change was noted (unadjusted values). However, as a level of statistical significance was not retained in the multivariate analyses, this finding is likely to be due to confounding from one or more other variables. In the multivariate analyses, when confounding factors were accounted for, a statistically significant (p=0.01) association between mean PGSI-12 score change and baseline primary mode of problem gambling was noted. On average, participants who gambled on pub electronic gaming machines showed a smaller improvement in mean PGSI-12 score (estimated mean PGSI change -8.22) than other participants, particularly participants who gambled on casino electronic gaming machines or casino table games (estimated mean PGSI change -18.15 and -17.17 respectively)7. 7 Note that the statistical significance is across the whole table of primary modes of problem gambling and is not a pairwise comparison. Some of the pairwise comparisons (e.g. between pub and club EGMs) are not statistically significant but the pairwise comparison between pub EGMs and casino EGMs is. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 11: PGSI-12 change by baseline gambling and related behaviours Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Number of
days since
last gamble
gambling
† Adjusted by baseline employment status, NZDI and PRIME-MD minor depressive disorder.
Variables which did not achieve a level of statistical significance are presented in Appendix 4,
Table 4.2. These related to baseline: self-reported number of days gambling per month and
gambling expenditure per day; self-rated control over gambling; length of problem duration;
level of motivation to overcome gambling problem; current goal (i.e. to quit some or all
modes of gambling or to control gambling); belief in treatment success, and perceived level of
difficulty in overcoming gambling problems; if assistance was currently being received for a
gambling problem; and if assistance had previously been received for a gambling problem.
4.3.4
PGSI-12 change at 12-month assessment by other baseline covariates
details mean PGSI-12 score change at the 12-month assessment from the baseline score and associations with other baseline covariates which attained a level of statistical significance. Statistically significantly associations between baseline major and minor depressive disorder (measured by PRIME-MD) and mean PGSI-12 score change were noted (unadjusted values). However, the association was different dependent on whether participants had major or minor depressive disorder at the baseline assessment. In the multivariate analyses, when confounding factors were accounted for, the association between mean PGSI-12 score change and major depressive disorder was no longer evident suggesting that it is likely to be due to confounding from one or more other variables. However, the statistically significant association was retained for minor depressive disorder and mean PGSI-12 score change (p=0.003) when confounding factors were accounted for. On average, participants who had minor depressive disorder at the baseline assessment were more likely to have a greater mean PGSI-12 score change at the 12 month assessment (estimated mean PGSI change -15.41) whilst participants who did not have minor depression at the baseline assessment showed a smaller improvement in mean PGSI-12 score change (estimated mean PGSI change -10.32). Level of deprivation (NZDI) was also statistically significantly associated with mean PGSI-12 score change. This finding was retained (p=0.014) in the multivariate analyses when confounding factors were accounted for. On average, participants in the lowest two deprivation quartiles were more likely to have a larger mean PGSI-12 score change (estimated mean PGSI change -14.75 or more) than participants in the higher two deprivation quartiles (estimated mean PGSI change -11.62 or lower). Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Statistically significantly associations were also noted for mean PGSI-12 score change and participants who had received treatment or prescriptions for mental health issues in the past 12 months when assessed at baseline (unadjusted values). However, as a level of statistical significance was not retained in the multivariate analyses, these findings are likely to be due to confounding from one or more other variables. Table 12: PGSI change by other baseline covariates Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Prime MD - Major
depressive disorder
Prime MD - Minor
depressive disorder
Treatment, mental
health last year
Prescription, mental
health last year
NZDI (quartiles)
† Adjusted by baseline employment status and primary problem gambling mode. Variables which did not achieve a level of statistical significance are presented in Appendix 4, Table 4.3. These related to baseline psychological distress (Kessler-10); alcohol abuse/ dependence (Audit-C, dichotomised to low risk and high risk); drug abuse/dependence (DAST); suicide ideation (dichotomised to no ideation and some ideation); dysthymia (PRIME-MD); current tobacco use; quality of life (WHOQoL); treatment for an alcohol or drug problem in the past 12 months; how work, social life, family and home life, and health were affected in the past month; and legal problems in the past 12 months. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of successful problem gambling outcomes - days gambled

This section presents data pertaining to mean change in time-averaged number of days
gambled per month from the baseline values. Section 4.4.1 details change in days gambled
from baseline values by assessment point. Associations between change in time-averaged
number of days gambled per month from the baseline values and uptake of formal gambling
treatment services is presented in section 4.4.2. Also presented are data pertaining to
associations between change in time-averaged number of days gambled per month from the
baseline values and socio-demographic characteristics (section 4.4.3), baseline gambling and
related behaviours and other baseline covariates (section 4.4.4). As the difference in number
of days gambled per month gives an indication of change in amount of gambling, these
analyses indicate which variables are associated with likelihood of better (i.e. improved)
outcomes after initial helpline contact and intervention delivery.
4.4.1
Days gambled by assessment point
No statistically significant difference was noted between assessment points for mean change in number of days gambled per month from baseline value Table 13: Change in days gambled by assessment point Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Assessment point 3 months

4.4.2
Days gambled by uptake of formal services
Uptake of formal gambling treatment services in the first three months was statistically significantly associated with time-averaged mean change in number of days gambled per month . This finding was retained (p=0.04) in the multivariate analyses when confounding factors were accounted for. On average, participants who accessed formal treatment services were more likely to report a slightly greater number of days per month when they did not gamble (estimated mean days gambled change -6.33) than participants who did not access formal treatment services (estimated mean days gambled change -4.98). A level of statistical significance was not attained when uptake of formal services by any of the follow-up assessments (three, six or 12 months) was examined . Table 14: Change in days gambled by uptake of formal services Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Uptake of formal
services by 3 months Yes
Uptake of formal
services by 3, 6 or 12 Yes
† Adjusted by marital status. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Days gambled by socio-demographic characteristics
details mean change in time-averaged number of days gambled per month from the baseline score and associations with socio-demographic characteristics which attained a level of statistical significance. In the multivariate analyses, when confounding factors were accounted for, a statistically significantly association (p=0.03) between mean change in time-averaged number of days gambled per month and marital status was noted. On average, widowed participants were least likely to have reduced their number of days gambling per month (estimated mean days gambled change -1.61) than other participants (estimated mean days gambled change -5.81 or more). Table 15: Change in days gambled by socio-demographic characteristics Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Marital status
† Adjusted by uptake of formal services in first three months.
Variables which did not achieve a level of statistical significance are presented in Appendix 5,
Table 5.1. These related to gender, age, ethnicity, employment status, highest educational
qualification, gross annual family income, and area of residence.
4.4.4
Days gambled by baseline gambling and related behaviours and other baseline
No statistically significant differences were noted for the change in time-averaged mean number of days gambled per month from baseline values and baseline gambling (dichotomised to EGM vs. other) and related behaviours, or any other baseline covariates examined (Appendix 5, Tables 5.2 to 5.3). Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of successful problem gambling outcomes - money lost

This section presents data pertaining to change in time-averaged money lost (dollars)
gambling per month from the baseline values. Section 4.5.1 details change in money lost per
month from baseline values by assessment point. Associations between change in time-
averaged money lost gambling per month from the baseline values and uptake of formal
gambling treatment services is presented in section 4.5.2. Also presented are data pertaining
to associations between change in time-averaged money lost gambling per month from the
baseline values and socio-demographic characteristics (section 4.5.3), baseline gambling and
related behaviours (section 4.5.4) and other baseline covariates (section 4.5.5). As the
difference in money lost gambling per month gives an indication of change in amount of
gambling, these analyses indicate which variables are associated with likelihood of better (i.e.
improved) outcomes after initial helpline contact and treatment.
4.5.1
Money lost by assessment point
No statistically significant difference was noted between assessment points for the mean change in money lost gambling per month from baseline values Table 16: Change in money lost by assessment point Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Assessment point 3 months

4.5.2
Money lost by uptake of formal services
A level of statistical significance for time-averaged mean change in money lost gambling per month (unadjusted values) was not attained when uptake of formal services in the first three months, or by any of the follow-up assessments (three, six or 12 months) was examined Table 17: Change in money lost by uptake of formal services Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Uptake of formal services
by 3 months
Uptake of formal services
by 3, 6 or 12 months

4.5.3
Money lost by socio-demographic characteristics
No statistically significant differences were noted for time-averaged mean change in money lost gambling per month and any socio-demographic characteristics (Appendix 6, Table 6.1). Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Money lost by baseline gambling and related behaviours
details change in time-averaged mean change in money lost gambling per month from the baseline score and associations with baseline gambling and related behaviours which attained a level of statistical significance. A statistically significantly association between time-averaged mean change in money lost gambling per month and PGSI-12 (past 12-month time frame) at the baseline assessment was noted. This finding was retained (p=0.02) in the multivariate analyses. On average, participants who scored 18 or more on the PGSI-12 were more likely to have a lower reduction in money lost gambling per month (estimated mean money lost change -$34.76 or less) than participants who scored 17 or less on the PGSI-12 (estimated mean money lost change -$38.82 or more). Table 18: Change in money lost by baseline PGSI-12 Unadjusted values
Multivariate values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
PGSI-12 (12 month
time frame)
(quartiles)
† PGSI-12 was the only factor appearing in the multivariate analyses.
Variables which did not achieve a level of statistical significance are presented in Appendix 6,
Table 6.2. These related to baseline: primary mode of problem gambling (dichotomised to
EGM vs. other); self-reported number of days gambling per month; self-rated control over
gambling; length of problem duration; level of motivation to overcome gambling problem;
current goal (i.e. to quit some or all modes of gambling or to control gambling); belief in
treatment success, and perceived level of difficulty in overcoming gambling problems; if
assistance was currently being received for a gambling problem; and if assistance had
previously been received for a gambling problem.
4.5.5
Money lost by other baseline covariates
details mean change in time-averaged money lost gambling per month from the baseline score and associations with other baseline covariates which attained a level of statistical significance. On average, participants who had major depressive disorder at the baseline assessment showed a statistically significant lower mean change in time-averaged money lost gambling per month than participants who did not have major depression (unadjusted values). However, as a level of statistical significance was not retained in the multivariate analyses, this finding is likely to be due to confounding from one or more other variables. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 19: Change in money lost by baseline major depressive disorder Unadjusted values
Multivariate values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
PRIME-MD -
Major depressive
disorder
† PGSI-12 was the only factor appearing in the multivariate analyses. Variables which did not achieve a level of statistical significance are presented in Appendix 6, Table 6.3. These related to baseline psychological distress (Kessler-10); alcohol abuse/ dependence (Audit-C, dichotomised to low risk and high risk); drug abuse/dependence (DAST); suicide ideation (dichotomised to no ideation and some ideation); minor depressive disorder and dysthymia (PRIME-MD); current tobacco use; quality of life (WHOQoL); level of deprivation (NZDI); treatment or prescriptions for mental health issues in the past 12 months; treatment for an alcohol or drug problem in the past 12 months; how work, social life, family and home life, and health were affected in the past month; and legal problems in the past 12 months. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of successful problem gambling outcomes - control over gambling

This section presents data pertaining to change in time-averaged control over gambling from
the baseline values. Section 4.6.1 details change in control over gambling by assessment
point. Associations between change in time-averaged control over gambling from the
baseline values and uptake of formal gambling treatment services is presented in section
4.6.2. Also presented are data pertaining to associations between change in time-averaged
control over gambling from the baseline values and socio-demographic characteristics
(section 4.6.3), baseline gambling and related behaviours (section 4.6.4) and other baseline
covariates (section 4.6.5). These analyses indicate which variables are associated with
likelihood of better (i.e. improved) outcomes after initial helpline contact and treatment.
4.6.1
Control over gambling by assessment point
No statistically significant difference was noted between assessment points for mean change in control over gambling from baseline values Table 20: Change in control over gambling by assessment point Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Assessment point 3 months

4.6.2
Control over gambling by uptake of formal services
A level of statistical significance for mean change in time-averaged control over gambling was not attained when uptake of formal services in the first three months, or by any of the follow-up assessments (three, six or 12 months) was exami Table 21: Change in control over gambling by uptake of formal services Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Uptake of formal services
by 3 months
Uptake of formal services
by 3, 6 or 12 months

4.6.3
Control over gambling by socio-demographic characteristics

details mean change in time-averaged control over gambling from the baseline score
and associations with socio-demographic characteristics which attained a level of statistical
significance.
Marital status was statistically significantly associated with change in time-averaged mean
control over gambling (unadjusted values). However, as a level of statistical significance was
not retained in the multivariate analyses, this finding is likely to be due to confounding from
one or more other variables.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 22: Change in control over gambling by marital status Unadjusted values
Variable
Category
Estimated least
Standard
squares mean diff.
Marital status

Variables which did not achieve a level of statistical significance are presented in Appendix 7,
Table 7.1. These related to gender, age, ethnicity, employment status, highest educational
qualification, gross annual family income, and area of residence.
4.6.4
Control over gambling by baseline gambling and related behaviours
details mean change in time-averaged control over gambling from the baseline value and associations with baseline gambling and related behaviours which attained a level of statistical significance. A statistically significantly association between baseline level of belief in treatment success and mean change in time-averaged control over gambling was noted (unadjusted values). This finding was retained in the multivariate analyses controlling for confounding factors (p=0.022). On average, participants who had higher than median belief in treatment success at the baseline assessment were more likely to have a slightly larger mean improvement in control over gambling (estimated mean change in control 3.90) than participants who had a lower than median belief in treatment success (estimated mean change in control 2.91). A statistically significantly association between baseline perceived level of difficulty in overcoming the gambling problem and mean change in time-averaged control over gambling was noted (unadjusted values). This finding was retained in the multivariate analyses controlling for confounding factors (p=0.032). On average, participants who perceived a high level of difficulty at the baseline assessment (score 8+) were more likely to have a slightly lower mean improvement in control over gambling (estimated mean change in control 2.88 or less) than participants who had a lower perception of difficulty (estimated mean change in control 3.53 or more). Table 23: Change in control over gambling by baseline gambling and related behaviours Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Belief in treatment
success,
Perceived level of
difficulty in
overcoming
problem (scale 1-
10) (quartiles)
† Adjusted by baseline WHOQoL and received treatment for mental health issues in past 12 months. Variables which did not achieve a level of statistical significance are presented in Appendix 7, Table 7.2. These related to baseline: primary problem gambling mode (dichotomised to EGM vs. other); PGSI-12 (12-month time frame); self-reported number of days gambling per month and gambling expenditure per day; length of problem duration; level of motivation to Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
overcome gambling problem; current goal (i.e. to quit some or all modes of gambling or to
control gambling); if assistance was currently being received for a gambling problem; and if
assistance had previously been received for a gambling problem.
4.6.5
Control over gambling by other baseline covariates
details mean change in time-averaged control over gambling from the baseline value and associations with other baseline covariates which attained a level of statistical significance. A statistically significantly association between baseline quality of life (WHOQoL) and mean change in time-averaged control over gambling was noted (unadjusted values). This finding was retained in the multivariate analyses controlling for confounding factors (p=0.002). On average, participants in the lowest quartile for quality of life were less likely to improve in mean control over gambling (estimated mean change in control 1.90) than participants who scored in the other three quartiles for quality of life (estimated mean change in control 3.59 or more). A statistically significant association was also noted for mean change in control over gambling and participants who had received treatment for mental health issues in the past 12 months when assessed at baseline (unadjusted values). This finding was retained in the multivariate analyses controlling for confounding factors (p=0.004). On average, participants who had not received treatment for mental health issues were more likely to have a greater improvement in mean control over gambling (estimated mean change in control 4.13) than participants who had received treatment (estimated mean change in control 2.67). Although the unadjusted analyses also indicated that there was an association between change in control over gambling and: baseline psychological distress (Kessler-10), major depressive disorder and dysthymia (PRIME-MD), level of deprivation (NZDI), and effect of gambling on physical health, a statistically significant association was not retained in the multivariate analyses controlling for confounding factors. Thus, these findings are likely to be due to confounding from one or more other variables. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 24: Change in control over gambling by other baseline covariates Unadjusted values
Adjusted values†
Variable
Category
Est. least
Standard
Est. least
Standard
mean diff.
mean diff.
Kessler-10
(quartiles)
Prime MD - Major No
depressive
disorder
Prime MD -
Dysthymia
(quartiles)
NZDI (quartiles)
Treatment, mental
health last year
How was health
affected in past
month? (10 point
scale) (quartiles)
† Adjusted by baseline belief in treatment success and perceived level of difficulty in overcoming problem. Variables which did not achieve a level of statistical significance are presented in Appendix 7, Table 7.3. These related to alcohol abuse/dependence (Audit-C, dichotomised to low risk and high risk); drug abuse/dependence (DAST); suicide ideation (dichotomised to no ideation and some ideation); minor depressive disorder (PRIME-MD); current tobacco use; treatment for an alcohol or drug problem in the past 12 months; prescription for a mental health issue in the past 12 months, how work, social life, and family and home life were affected in the past month; and legal problems in the past 12 months. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Predictors of successful problem gambling outcomes - treatment success

This section presents data pertaining to associations between difference in time-averaged
treatment success (Gambling-Quit or improved). Section 4.7.1 details treatment success by
assessment point. Association between time-averaged treatment success and uptake of formal
gambling treatment services is presented in section 4.7.2. Also presented are data pertaining
to associations between time-averaged treatment success and socio-demographic
characteristics, baseline gambling and related behaviours and other baseline covariates
(section 4.7.3).
4.7.1
Treatment success by assessment point
No statistically significant difference was noted between assessment points for treatment success Table 25: Treatment success by assessment point Variable
Category
Univariate odds ratio
OR (95% CI) p-value
Assessment point 3 months

4.7.2
Treatment success by uptake of formal services
There was no difference in time-averaged treatment success when examined by uptake of formal services in the first three months, or by any of the follow-up assessments (three, six or 12 months) Table 26: Treatment success by uptake of formal services Variable
Category
Univariate odds ratio
OR (95% CI)
Uptake of formal services
by 3 months
Uptake of formal services
by 3, 6 or 12 months

4.7.3
Treatment success by socio-demographic characteristics, gambling and related
behaviours and other baseline covariates
No statistically significant differences were noted for time-averaged treatment success and socio-demographic characteristics, gambling and related behaviours, or any other baseline covariates examined (Appendix 8, Tables 8.1 to 8.3). Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
DISCUSSION
The major purpose of the study was to ascertain whether there are differences in treatment outcomes between those who only access the Helpline and those who subsequently obtain additional professional counselling or treatment for problem gambling. A further purpose was to identify client characteristics associated with treatment outcome. The present study had sufficient participant numbers to examine relationships between many variables relevant to the study purposes. However, a larger sample would have been required to explore more complex interactions between variables. Participants appeared to be broadly representative of Helpline callers during the period of recruitment and the inclusion/exclusion criteria were such that the great majority of callers were eligible. This increases confidence that the results of the present study can be generalised to Helpline clients generally. A number of standardised measures were included that facilitate comparison with findings from other research. While necessary to assess outcomes of theoretical and practical interest, it is possible that the nature of the questions involved and researcher contact with clients during the course of the study may have had a therapeutic effect additional to that associated with Helpline treatment per se and subsequent clinical interventions. Participant retention over the follow-up assessment period was satisfactory given the nature of the population and the mode of contact (telephone). Furthermore, attrition was not differential, at least with regard to a wide range of socio-demographic variables and primary problem gambling mode. This increases confidence in the findings. However, the methodology has some limitations. This is an uncontrolled outcome study. While changes following intervention can be assessed, it is not possible to determine directly whether or not these changes are a consequence of the intervention per se. That requires the inclusion of defined interventions in a randomised controlled trial that incorporates a no treatment or placebo control group and/or one or more interventions of known effectiveness. Measurement relies on self-report and there is always the possibility that the accuracy of participant responses will be compromised to an unknown degree. In the present study, treatment was clearly defined and was delivered by trained Helpline counsellors with a high degree of integrity and consistency across time and counsellors. However, in the case of face-to-face and other treatment subsequently accessed by some participants, it is not known what interventions were involved. For the most part these services leave it to individual counsellors to decide what modality or modalities to use. This compromises understanding of what therapeutic features contribute to client outcomes. As part of intervention delivery, participants were referred to, or informed about, other professional gambling treatment services. During the first three months of the follow-up period just under a third (31%) of participants reported receiving assistance for their gambling problem from a professional treatment service and a larger number (39%) indicated that they had received informal support; Both forms of assistance for gambling problems reduced in frequency at subsequent assessment points. At 12 months the corresponding percentages were 19% and 25%. Counselling was received from the range of major providers mentioned previously in the report and there did not appear to be a clear preference. Some participants sought professional help from more than one provider. Men were more likely to access additional treatment than women and this was the case irrespective of whether or not they had previously, during the past 12 months, received help for a mental health problem. However, for women, this was just the case for those who had previously received treatment for a mental health problem. Only 15.4% of women who had not previously received treatment accessed additional problem gambling treatment during the first three months following Helpline intervention. People who had received treatment for a gambling problem during the Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
past 12 months prior to contacting the Helpline were also more likely to access further gambling treatment during this period. It is perhaps not surprising that many people who had in the past 12 months obtained professional help for a gambling or mental health problem sought additional, more intensive, specialist assistance with their gambling problems. This may be because they have more serious gambling and mental health problems and/or because they are more familiar and perhaps comfortable with accessing this form of assistance than callers who have not previously. These possibilities require further consideration. It is unclear why almost as many men (40%) who had not received treatment for a mental health problem obtained additional gambling treatment as those who had (50%). Again further examination of the study data and additional research is required to understand this. Callers contacted the Helpline from all parts of the country. It is of interest that area of residence did not predict the accessing of additional gambling treatment, perhaps reflecting the wide availability of formal problem gambling services in New Zealand and suggesting that geographical location is not a significant access barrier for most callers in regard to follow-up treatment from gambling treatment services. It should be noted that a large number of factors examined did not predict additional treatment involvement. Among others this included primary problem gambling mode (EGM versus other); various indicators of problem gambling severity, impact and duration; motivation to overcome gambling problem; treatment goal; level of belief in treatment success; or perceived difficulty in overcoming problems. This was also the case for the various measures of psychological distress, mental health disorder, substance use/misuse, suicidal ideation, quality of life, and treatment for an alcohol or drug problem in the past 12 months. It is perhaps surprising that people with more serious gambling and comorbid mental health disorders, or with greater perceived difficulty in overcoming gambling problems, were not more likely to seek additional, more intensive assistance. This suggests that the possible association mentioned above concerning past involvement in gambling or mental health treatment and receipt of post-Helpline intervention is more linked to previous experience of help-seeking than to problem severity and need for more intensive treatment. This requires further investigation and has relevance to the development of stepped-care service delivery. Perhaps the most notable study finding is that participants evidenced substantial, statistically and clinically significant improvement with respect to problem gambling and some associated mental health problems. In many instances these improvements occurred during the first three months and were sustained throughout subsequent assessments. With regard to problem gambling this included days gambled, money lost gambling and control over gambling, and problem gambling severity and impacts on work, social life, family/home and physical health. There were notable reductions in psychological distress and the prevalence of major and minor depression and drug abuse. Less change was evident for tobacco use and while there was some reduction in alcohol misuse post-intervention, by 12 months it had increased to just below the baseline rate. These outcomes are impressive, particularly given the severity of gambling problems and associated comorbidity. The changes are, for the most part, of large magnitude and durable throughout the 12 month follow-up period. They were achieved even through most clients had received only one Helpline counselling session and did not subsequently access other, more intensive, gambling counselling. The large reductions in levels of psychological distress and the prevalence of common mood disorders is also of note and suggests that the resolution of gambling problems may play a major role. This requires further research including closer examination of associated changes in life situation and use of medical interventions. It is likely that many of these clients would have been prescribed an antidepressant, especially the 57.7% who met the clinical criteria for major depression at the time they contacted the Helpline, if they had presented to a general medical practitioner and their depressive symptoms had been noted. This underlines the Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
importance of general practitioners and other primary health care practitioners inquiring about problem gambling and other psychological and social factors when patients present with anxiety and mood disorders. The finding that many participants continued to smoke and experience alcohol problems requires consideration. What is the role of problem gambling services in addressing these serious health issues, either by focusing on them concurrently with gambling treatment or by referral? Would this broadening of focus compromise gambling treatment outcomes? These are questions for further discussion and study. It is sometimes asserted that ‘symptom substitution' occurs following treatment for a particular disorder, for example that successful treatment for gambling leads to an alcohol problem or increased tobacco use. Analysis of the data in the present study focuses on the overall group, rather than examines individual trajectories over time. These and other matters concerning change in individuals can be subsequently examined further, particularly when the outcome data from the 36-month assessment become available. Given that there was no overall increase in smoking or alcohol misuse and that the prevalence of drug abuse, psychological distress and major and minor depression reduced markedly, this type of substitution is unlikely except, perhaps, in a small number of cases. The finding that 57.5% of participants, based on the PGSI-12, remained problem gamblers at the 12-month assessment requires discussion. First, it needs to be noted that the mean PGSI-12 score reduced substantially, from 17 to 9. This indicates a clinically meaningful reduction in the severity of gambling problems for most participants, even though the majority continued to meet the criteria for problem gambling. Additionally, a number of the PGSI items refer to experiences and consequences that could be expected to continue for some time, even when people stop gambling completely. For example the PGSI-12 asks if, during the past 12 months, respondents have felt they might have a problem with gambling, experienced health problems caused by gambling, had financial problems caused by gambling and felt guilty about the way they gamble or what happens when they gamble. Further analysis is required to assess changes in responses to individual questions over time. It is likely that responses to some questions will change markedly within the first three months and that others will change later. People who accessed formal gambling treatment services during the first three months of follow-up were more likely than those who did not access services to report time-averaged reductions in mean days gambled per month throughout the follow-up assessment period. However, while statistically significant, the difference was not large and accessing additional gambling services at later periods during the 12-month follow-up period was not associated with better outcome on this measure and may be due, in part, to the fact that active referral or information about other services was only provided to participants when they received their Helpline intervention (i.e. at initial contact). Involvement in additional gambling treatment was also found not to be associated with PGSI-12 problem reduction, time-averaged mean change in money lost gambling per month or time-averaged self-assessed treatment success (gambling- quit or improved). On the face of it these findings suggest that involvement in gambling treatment, additional to the Helpline intervention, contributed little if at all to treatment outcome. However, it may be that many or most participants who chose not to seek problem gambling assistance additional to that received from the Helpline (including the self-help manual) were of the view that they did not require additional professional assistance, and that this assessment was often correct. Many will have received informal support. Similarly, perhaps those who are of the view that they require more professional support and subsequently received it may have had a better outcome than would otherwise be the case. This could also explain the finding that there is little or no difference in gambling treatment outcomes between those who accessed additional treatment and those who did not. If this is Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
the case, however, it is odd that it was not reflected in more serious gambling problems, psychopathology and some other relevant measures at baseline. While additional treatment engagement appeared to have little impact on treatment outcomes, some participant attributes did. The following were associated with worse outcomes on one of the problem gambling outcome measures: having pub EGMs as primary problem gambling mode; PGSI-12 problem gambling severity at initial contact; living in high deprivation areas; being disabled, having an illness or being on sick leave; being a widow; having a low quality of life; having received treatment for a mental health problem in the past year; and perceiving a high level of difficulty in overcoming their gambling problem. People with a higher belief in treatment success and having a minor (but not major) depressive disorder, improved more. These findings are interesting and potentially important. However, the association in all cases was with only one outcome measure, not across the board. The finding of large differences in treatment outcome (measured by changes in PGSI-12 scores) depending on where people predominantly access EGMs was unexpected. However, on reflection, there might be a link between poorer outcomes for problem gamblers who play machines in pubs, rather than in casinos, and living in high deprivation areas. Residence in high deprivation areas was also associated with lower levels of improvement on the PGSI-12 measure. Pub EGMs are more heavily concentrated in high deprivation neighbourhoods and previous New Zealand studies have found high levels of participation and gambling problems among people living in closer proximity to EGMs as well as in more deprived areas. To date there has been a tendency to categorise EGM participants together irrespective of where they are located. Location could, however, be important, both in terms of access and contextual features that may influence participation, as well with regard to the types of people who frequent the different settings. Some of the groups that did not change as much as others may also be more likely to reside in high deprivation areas, as well as experience lower levels of social capital including social support and a lack of meaningful occupation. Problem severity was found to be a strong predictor of future problem gambling in a previous New Zealand study (Abbott, Williams & Volberg, 2004). This was a longitudinal study of pathological, problem and frequent non-problem gamblers recruited from a national prevalence study, not a treatment sample. Self-efficacy has been shown to predict better treatment outcomes in previous studies involving problem gamblers and other clinical groups (Project MATCH Research Group, 1997; Hodgins et al., 2009). Perceiving higher levels of difficulty in overcoming a gambling problem and believing in treatment success are measures of this construct. Further research on the role of self-efficacy in treatment and ways to enhance it early in treatment would be helpful. The association of minor (but not major) depression at baseline and better treatment outcome might suggest that moderate but not high levels of depression in some way help to facilitate behaviour change. However, again this association was found with regard to just one outcome measure and probably not too much should be made of it. The factors discussed above have relevance to understanding client variation in treatment-seeking and treatment response and ways to better match client subgroups with interventions to enhance outcomes. However, a major conclusion, and an important one, is that many of the factors examined had little or no association with treatment outcomes. This includes age, gender and ethnicity, characteristics often found to influence access and response to health and social services. While some caution is required owing to low numbers of people in some groups such as Pacific people and Asians, perhaps the most notable finding is that clients generally, irrespective of their socio-demographic characteristics, experienced clinically significant, sustained improvement for their gambling and some related problems following Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
access to the Helpline. This requires further examination in larger studies with longer-term follow-up assessments. As mentioned, an outcome study does not enable direct comparison with other interventions, including those of known efficacy. In one sense this does not matter. If relatively low cost interventions such as telephone counselling are associated with significant clinical gains, that is sufficient information for some purposes. However, what if the outcomes are not better than they would have been if clients had not sought help? From prospective general population studies it is apparent that gambling problems fluctuate considerably over time and that recovery, without professional intervention, is not uncommon. However, relapse is also common, particularly among people with more serious problems (Abbott & Clark, 2007). Psychotherapy research, generally, as well that relating to gambling, provides convincing evidence that most people improve during therapy and these improvements are often sustained (Lambert & Ogles, 2004). In part this is because people who seek help for gambling and other problems have usually reached a crisis point and are at a stage where they are more open to behaviour change. At this time, if a client engages with a counsellor or therapist who is perceived to be credible and the counsellor delivers an intervention that he or she believes is likely to be effective, positive outcomes are common. Clinical trials with problem gambling and other mental health disorders have often found little or no difference in outcome between treatments. In part, this is because of shared ‘non-specific' placebo effects. Westphal and Abbott (2006) have noted that problem gambling trials typically appear to have high rates of non-specific response, apparently even higher than rates found with trials involving other mental health disorders. The major clinical interest, apart from finding ways to enhance the impact of non-specific effects, is to develop specific, defined interventions that further enhance effectiveness, and to better match clients to particular therapies and therapy components. This information is required to inform the development of services that reach more people at an earlier stage, deliver better outcomes and are cost-effective. As Helpline treatment was included in the randomised controlled trial (RCT), it is possible to compare it with interventions that had previously been studied in RCTs that included a wait-list control group. While follow-up was relatively short, for ethical reasons, participants allocated to treatment groups did significantly better than those who were waiting to access treatment. Given that there were no significant differences between helpline treatment and these interventions on gambling outcome measures, it is highly likely that Helpline counselling is contributing to the marked reductions in gambling and related problems reported and discussed in the present study. Currently, for the first time, there are large scale-prospective general population studies of gambling being conducted (in Sweden, Australia, Canada and New Zealand). Among many other things these studies will provide detailed information on the natural history of problem gambling, including problem onset, change over time, help-seeking, ‘natural' recovery and relapse. Information from these studies will enable comparisons of rates of change with and without therapy to be made. Little is known about the long-term effects of problem gambling treatment, beyond one or two years. Longer term tracking of the present clinical sample and other clients in the RCT would enable large clinical and non-clinical samples of problem and at-risk gamblers to be studied in parallel. While the addition of formal professional counselling to Helpline treatment did not generally lead to better outcomes this does not mean that such services are without value. Had clients first presented to these other services, as the majority of clients do, it is likely that similar improvements would have occurred. Further, it may well be that supplementing these interventions with Helpline engagement would have made little or no difference to these clients. However, it would be helpful to know what interventions are being delivered by face- Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
to-face services across the country and to what extent they correspond to those interventions that have been found to be more efficacious with problem gamblers. It would also be helpful to know which clients do as well or better from a relatively low-cost, brief intervention as they would from more intensive face-to-face interventions and which do better with the latter. This will require further investigation of face-to-face interventions using both prospective cohort and RCT study designs. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
REFERENCES
Abbott, M.W. (2001). Problem and non-problem gamblers in New Zealand: A report on Phase Two of the 1999 National Prevalence Study. Wellington: Department of Internal Affairs. Abbott, M., Bellringer, M., Vandal, A., Hodgins, D., Palmer Du Preez, K., Landon, J., Sullivan, S., & Feigin, V. (2012). Effectiveness of problem gambling brief interventions: A randomised controlled trial. Auckland: Gambling and Addictions Research Centre, Auckland University of Technology. Abbott, M.W, & Clark, D. (2007). Prospective problem gambling research: Contribution and potential. International Gambling Studies, 7(1), 123-144. Abbott, M.W., Williams, M. & Volberg, R. (1999). Seven years on: A follow-up study of frequent and problem gamblers living in the community. Wellington, New Zealand: Department of Internal Affairs. Abbott, M.W., Williams, M. & Volberg, R.A. (2004). A prospective study of problem and regular non-problem gamblers living in the community. Substance Use and Misuse, 39 (6), 855-884. Abbott, M.W., & Volberg, R. (1992). Frequent and problem gambling in New Zealand. Research Series No. 14. Wellington: Department of Internal Affairs. Abbott, M.W. & Volberg, R.A. (1996). The New Zealand National Survey of Problem and Pathological Gambling. Journal of Gambling Studies, 12(2), 143-159. Abbott, M.W., & Volberg, R.A. (2000). Taking the pulse on gambling and problem gambling in New Zealand: A report on Phase One of the 1999 National Prevalence Survey. Report Number Three of the New Zealand Gaming Survey. Wellington: Department of Internal Affairs. Abbott, M.W., & Volberg, R.A. (2006). The measurement of adult problem and pathological gambling. International Gambling Studies, 6, 175-200. Amey, B. (2001). People's participation in and attitudes to gambling, 1985–2000: Final results of the 2000 survey. Wellington: Department of Internal Affairs. Anderson, G., & Brown, R.I.F. (1984). Real and laboratory gambling, sensation seeking and arousal: Towards a Pavlovian component in general theories of gambling and gambling addictions. British Journal of Psychology, 75, 401 - 411. Battersby, M., Oakes, J., Tolchard, B., Forbes, A. & Pols, R. (2008). Cognitive behavioural therapy for problem gamblers. In M. Zangeneh, A. Blaszczynski, N. Turner (Eds.), In the Pursuit of Winning (pp 179-197). New York: Springer. Black, D. W. (2004). An open-label trial of bupropion in the treatment of pathological gambling. Journal of Clinical Psychopharmacology, 24, 108-110. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Blackman, S., Simone, R.V., & Thoms, D.R. (1989). The Gamblers Treatment Clinic of St. Vincent's North Richmond Community Mental Health Center: Characteristics of the clients and outcome of treatment. The International Journal of the Addictions, 24(1), 29-37. Brewin, C.R. (1996). Theoretical foundations of cognitive-behavior therapy for anxiety and depression. Annual Review of Psychology, 47, 33-57. Brown, R.I.F. (1985). The effectiveness of Gamblers Anonymous. In W. R. Eadington (Ed.) The gambling studies: Proceedings of the sixth national conference on gambling and risk taking (pp. 259 - 284), Reno, NV: Bureau of Business and Economic Research, University of Nevada, Reno. Christoffel, P. (1992). People's participation in and attitudes towards gambling. Wellington: Department of Internal Affairs. Cocco, K.M., & Carey, K.B. (1998). Psychometric properties of the Drug Abuse Screening Test in psychiatric outpatients. Psychological Assessment, 10, 408-414. Delfabbro, P.H. (2012). Australasian gambling review (1992–2011) 5th Edition. Adelaide: Independent Gambling Authority. Department of Internal Affairs. (2008). People's participation in, and attitudes to, gambling, 1985–2005. Wellington: Department of Internal Affairs. Department of Internal Affairs website. (2012). Various statistics: http://www.dia.govt.nz/ Resource-material-Information-We-Provide-Gaming-Statistics?OpenDocument Dickerson, M.G., & Weeks, D. (1979). Controlled gambling as a therapeutic technique for compulsive gamblers. Journal of Behavior Therapy and Experimental Psychiatry, 10, 139 - 141. Emshoff, J., Perkins, A., Zimmerman, L., Mooss, A. & Zorland, J. (2007). Pathological gambling treatment literature review. Second Edition. Georgia State University, Atlanta, GA. Ferris, J., & Wynne, H. (2001). The Canadian Problem Gambling Index: Final report. Ottawa: Canadian Centre on Substance Abuse. Freidenberg, B.M., Blanchard, E.B., Wulfert, E. & Malta, L.S. (2002). Changes in physiological arousal to gambling cues among participants in motivationally enhanced cognitive-behavior therapy for pathological gambling: A preliminary study. Applied Psychophysiology and Biofeedback, 27(4), 251-260. Gaboury, A., & Ladouceur, R. (1989). Erroneous perceptions and gambling. Journal of Social Behavior and Personality, 4, 411 - 420. Gamblers Anonymous website: http://www.12steps.co.nz/Home_Page.php Gambling Helpline. (2012). Report for national statistics to 31 December 2011. Auckland: Gambling Helpline New Zealand. Gerstein, D., Volberg, R.A., Murphy, S., Toce, M. et al. (1999). Gambling impact and behavior study. Report to the National Gambling Impact Study Commission. Chicago: National Opinion Research Center at the University of Chicago. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Griffiths, M.D., & MacDonald, H.F. (1999). Counselling in the treatment of pathological gambling: An overview. British Journal of Guidance & Counseling, 27(2), 179 - 190. Hapai Te Hauora Tapui website: http://www.hapai.co.nz/issues/problem-gambling. Health Sponsorship Council. (2007). 2006/07 Gaming and Betting Activities Survey: New Zealanders' knowledge, views and experience of gambling and gambling-related harm. Auckland: Health Sponsorship Council. Hodgins, D.C., Currie, S.R., Currie, G. & Fick, G.H. (2009). Randomized trial of brief motivational treatments for pathological gamblers: More is not necessarily better. Journal of Consulting and Clinical Psychology, 77(5), 950–960. Hodgins, D.C. & Peden, N. (2005). Natural course of gambling disorders: Forty-month follow-up. Journal of Gambling Issues, 14, 1-15. Hollander, E., DeCaria, C.M., Mari, E., Wong, C.M., Mosovich, S., Grossman, R. & Begaz, T. (1998). Short-term single-blind fluvoxamine treatment of pathological gambling. American Journal of Psychiatry, 155, 1781-1783. Jackson, A.C., Thomas, S.A., Thomason, N., Borrell, J., Crisp, B.R., Ho, W., Holt, T.A. & Smith, S. (2000). Longitudinal evaluation of the effectivenss of problem gambling counselling services, community education strategies and information products - Volume 2: Evaluation of the effectiveness of the counselling interventions at BreakEven Problem Gambling Counselling Services. Melbourne: Victorian Department of Human Services. Jimenez-Murcia, S., Aymami, N., Gómez-Peña, M., Santamaria, J.J., A lvarez-Moya, E., Fernández-Aranda, F., Granero, R., Penelo, E., Bueno, B., Moragas, L., Gunnard, K. & Menchón, J.M. (2012). Does exposure and response prevention improve the results of group cognitive-behavioural therapy for male slot machine pathological gamblers? British Journal of Clinical Psychology, 51, 54–71. Kessler, R., & Mroczek, D. (1994). Final versions of our Non-Specific Psychological Distress Scale. Written communication-memo dated 10/3/94. Ann Arbour (MI), Surrey Research Center of the Institute for Social Research, University of Michigan. Kobak, K.A., Taylor, L.H., Dottl, S.L., Greist, J.H., Jefferson, J.W., Burroughs, D., Mantle, J.M., Katzelnick, D.J., Norton, R., Henk, H.J., & Serlin, R.C. (1997). A computer-administered telephone interview to identify mental disorders. Journal of the American Medical Association, 278, 905-910. Korn, D.A. & Shaffer, H.J. (2004). Massachusetts's department of public health practice guidelines for treating gambling related problems: An evidenced based treatment guide for clinicians. Boston, MA: Massachusetts Council on Compulsive Gambling. Kuentzel, J.G., Henderson, M.J., Zambo, J.J., Stine, S.M., & Schuster, C.R. (2003). Motivational interviewing and fluoxetine for pathological gambling disorder: A single case study. North Americans Journals of Psychology, 5(2), 229-248. Ladouceur, R., Sylvain, C., Letarte, H., Giroux, I., & Jacques, C. (1998). Cognitive treatment of pathological gamblers. Behaviour Research and Therapy, 36, 1111-1119. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Ladouceur, R., & Walker, M. (1996). A cognitive perspective on gambling. In P.M. Salkovskis (Ed.) Trends in Cognitive and Behavioural Therapies, 89-120. New York: John Wiley and Sons. Lambert, M. J., & Bergin, A. E. (1994). The effectiveness of psychotherapy. In A.E. Bergin & S.L. Garfield (Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 143-189). New York: Wiley. Lambert, M. J., & Ogles, B. M. (2004). The efficacy and effectiveness of psychotherapy. In M.J. Lambert (Ed.), Bergin and Garfield's handbook of psychotherapy and behavior change (5th ed., pp. 139-193). New York: Wiley. Lester, D. (1980). The treatment of compulsive gambling. The International Journal of the Addictions 15, (2), 201-206. Marks, I.M. (1987). Fears, phobias and rituals. London: Oxford University Press. MacDonald, G., Sheldon, B. & Gillespie, J. (1992). Contemporary studies of the effectiveness of social work. British Journal of Social Work, 22(6), 615-643. Martin, D.J., Garske, J P. & Davis, M.K. (2000). Relation of the therapeutic alliance with outcome and other variables: A meta-analytic review. Journal of Consulting and Clinical Psychology, 68(3), 438-450. McConaghy, N., Armstrong, M.S., Blaszczynski, A., & Allcock, C. (1983). Controlled comparison of aversive therapy and imaginal desensitization in compulsive gambling. British Journal of Psychiatry, 142, 366-372. McConaghy, N., Armstrong, M.S., Blaszczynski, A., & Allcock, C. (1988). Behavior completion versus stimulus control in compulsive gambling: Implications for behavioral assessment. Behavior modification, 12, 371-384. McConaghy, N., Blaszcynski, A., & Frankova, A. (1991). Comparison of imaginal desensitisation with other behavioural treatments of pathological gambling. A two- to nine-year follow-up. British Journal of Psychiatry, 159, 390-393. Miller, W.R., & Rollnick, S. (2002). Motivational interviewing: Preparing people for change (2nd ed.). New York: Guilford Press. Miller, W.R., Zweben, A., DiClemente, C.C., & Rychtarik, R.G. (1992). Motivational enhancement therapy manual: A clinical research guide for therapists treating individuals with alcohol abuse and dependence (Project MATCH Monograph Series 2). Rockville, MD: National Institute on Alcohol Abuse and Alcoholism. Ministry of Health. (2006). Problem gambling in New Zealand: Analysis of the 2002/03 New Zealand Health Survey. Wellington: Ministry of Health. http://www.moh.govt.nz/moh.nsf/ pagesmh/5025?Open. Ministry of Health (2009). A focus on problem gambling. Results of the 2006/07 New Zealand Health Survey. Wellington: Ministry of Health. Ministry of Health (2012) website: http://www.health.govt.nz/our-work/preventative-health-wellness/problem-gambling/service-user-data/intervention-client-data. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Moyers, T.B., Martin, T., Manuel, J.K., Hendrickson, S.M.L., & Miller, W.R. (2004). Assessing competence in the use of motivational interviewing. Journal of Substance Abuse Treatment, 28(1), 19-26. Oakes, J., Battersby, M.W., Pols, R.G. & Cromarty, P. (2008). Exposure therapy for problem gambling via videoconferencing: A case report. Journal of Gambling Studies, 24, 1074-1118. Oakes, J., Gardiner, P., McLaughlin, K. & Battersby, M. (2012). A pilot group cognitive behavioural therapy program for problem gamblers in a rural Australian setting. International Journal of Mental Health and Addiction, 10, 490-500. Oasis website: http://salvationarmy.org.nz/here-to-help/addictions/oasis-centres/. Petry, N.M. (2003). Patterns and correlates of Gamblers Anonymous attendance in pathological gamblers seeking professional treatment. Addictive Behaviors, 28, 1049-1062. Petry, N.M. & Armentano, C. (1999). Prevalence, assessment, and treatment of pathological gambling: A review. Psychiatric Services, 50 (8), 1021-1027. PGF website: http://www.pgfnz.org.nz/Uploads/PDFDocs/ Information_Booklet.PDF. PGF website: Help seeking - http://www.pgfnz.org.nz/Get-Help/0,275,1532,00.html. Problem Gambling Research and Treatment Centre (PGRCT) (2011). Guideline for screening, assessment and treatment in problem gambling. Clayton: Monash University. Productivity Commission. (2010). Gambling, Report No. 50. Canberra: AusInfo. Project MATCH Research Group. (1997). Matching alcoholism treatments to client heterogeneity: Project MATCH posttreatment drinking outcomes. Journal of Studies on Alcohol, 58, 7-29. Rankin, H. (1982). Case histories and shorter communications. Behavior Research and Therapy, 20, 185-187. Reid, K., & Searle, W. (1996). People's participation in and attitudes towards gambling: Final results of the 1995 survey. Wellington: Department of Internal Affairs. Reinert, D.F., & Allen, J.P. (2002). The Alcohol Use Disorders Identification Test (AUDIT): A review of recent research. Alcoholosim: Clinical and Experimental Research, 26(2), 272-279. Russo, A.M., Taber, J.I., McCormick, R.A. & Ramirez, L.A. (1984). An outcome study of an inpatient treatment program for pathological gamblers. Hospital and Community Psychiatry, 35, (8), 823-827. Salmond, G. (2005). Cited in Ministry of Health (2005). The New Zealand Health Survey 2002/03. Wellington: Author. Saunders, J.B., Aasland, O.G., Babor, T.F., De La Fuente, J.R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption–II. Addiction 88,791–804. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Schmidt, S., Muhlan, H., & Power, M. (2005). The EUROHIS-QOL 8-item index: Psychometric results of a cross-cultural field study. European Journal of Public Health, 16(4), 420-428. Shaffer, H.J. & Hall, M.N. (2001). Updating and refining prevalence estimates of disordered gambling behaviour in the United States and Canada. Canadian Journal of Public Health, 92, 168-172. SHORE & Whariki. (2008). Assessment of the social and economic impacts of gambling in New Zealand. Auckland: Centre for Social and Health Outcomes Research and Evaluation (SHORE) & Te Ropu Whariki, Massey University. Skinner, H.A. (1982). The Drug Abuse Screening Test. Addictive Behaviors, 7, 363-371. Slutske, W. (2006). Natural recovery and treatment-seeking in pathological gamblers. American Journal of Psychology, 163(2), 297-302. Slutske, W., Jackson, K.M. & Sher, K.J. (2003). The natural history of problem gambling. Journal of Abnormal Psychology, 112, 263-274. Smith, M.L., Glass, G.V., & Miller, T.I. (1980). The benefits of psychotherapy. Baltimore: Johns Hopkins University Press. Smith, D., Harvey, P., Battersby, M., Pols, R., Oakes, J. & Baigent, M. (2010). Treatment outcomes and predictors of drop out for problem gamblers in South Australia: A cohort study. Australia New Zealand Journal of Psychiatry, 44 (10), 911-920. Smith, S.A., Thomas, S.A. & Jackson, A.C. (2004). An exploration of the therapeutic relationship and counselling outcomes in a problem gambling counselling service. Journal of Social Work Practice, 18(1), 99-112. Sobell, L.C., & Sobell, M.B. (1992). Time-line follow-back: A technique for assessing self-reported alcohol consumption. In R.Z. Litten and J.P. Allen (Eds.), Measuring alcohol consumption: Psychological and bio-chemical methods (pp. 41-72). Totowa, NJ: Humana Press. Spitzer, R.L., Williams, J.B.W., Gibbon, M., & First, M.B. (1992). The Structured Clinical Interview for DSM-III-R (SCID), I: History, rationale, and description. Archives of General Psychiatry, 49, 624-629. Spitzer, R.L., Williams, J.B., Kroenke, K., Linzer, M., deGruy, F.V., 3rd, Hahn, S.R., Brody, D., & Johnson, J.G. (1994). Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study, Journal of American Medical Association, 272(22), 1749-56. Stewart, R.M., & Brown, S.F. (1988). An outcome study of Gamblers Anonymous. British Journal of Psychiatry, 152, 284-288. Stiles, W.B., Shapiro, D., & Elliot, R. (1986). Are all psychotherapies equivalent? American Psychologist, 41, 165-180. Stinchfield, R. & Winters, K.C. (2001). Outcome of Minnesota's gambling treatment programs. Journal of Gambling Studies, 17(3), 217-245. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Suurvali, H., Hodgins, D., Toneatto, T., & Cunningham, J. (2008). Treatment seeking among Ontario problem gamblers: Results of a population survey. Psychiatric Services, 59(11), 1343-1346. Symes, B.A., & Nicki, R.M. (1997). A preliminary consideration of cue-exposure and response prevention treatment for pathological gambling behaviour: Two case studies. Journal of Gambling Studies, 13, 145-157. Taber, J.I., McCormick, R.A., Russo, A.M., Adkins, B.J., & Ramirez, L.F. (1987). Follow-up of pathological gamblers after treatment. American Journal of Psychiatry, 144, (6), 757-761. Toneatto, T. & Dragonetti, R. (2008). Effectiveness of community-based treatment for problem gambling: A quasi-experimental evaluation of cognitive-behavioral vs. twelve-step therapy. The American Journal on Addictions, 17, 298-303. Toneatto, T., & Sobell, L.C. (1990). Pathological gambling treated with cognitive behavior therapy: A case report. Addictive Behaviors, 15, 497-501. Toneatto, T. & Gunaratne, M. (2009). Does the treatment of cognitive distortions improve clinical outcomes for problem gambling? Journal of Contemporary Psychotherapy, 39, 221-229. Walker, M.B. (1993). Treatment strategies for problem gambling: A review of effectiveness. In W.R. Eadington and J. Cornelius (Eds.) Gambling behavior and problem gambling (pp. 533-566), Reno, NV: Institute for the Study of Gambling and Commercial Gaming. Westphal, J.R., & Abbott, M.W. (2006). Models for multi-site problem gambling clinical trials. International Gambling Studies, 6(2), 129-145. Williams, R.J. & Volberg, R.A. (2010). Best practices in the population assessment of problem gambling. Ontario Problem Gambling Research Centre. Guelph, Ontario, Canada. Wither, A. (1987). Taking a gamble: A survey of public attitudes towards gambling in New Zealand. Wellington: Department of Internal Affairs. Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 1
Ethical approval
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 2
Descriptive statistics
Table 2.1: Socio-demographics by assessment point Assessment point
Baseline
12 months
Variable
Marital Status
Age Group
Ethnicity - Priority
Ethnicity - Any
Employment status
Disabled/illness/sick leave Highest educational
qualification achieved
None
Secondary school qualification Trade or technical certificate Professional qualification Undergrad. Dip. or Cert. University degree & above Gross family income in last 12
months
<$20,000
$20,001 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.1: Socio-demographics by assessment point - continued Assessment point
Baseline
12 months
Variable
Area of residence
Waikato/Coromandel East Coast (Bay of Plenty/ Lakes/Hawkes Bay) Primary problem gambling mode
Casino gaming machines Casino table games Club gaming machines Pub gaming machines Primary mode - EGM
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.2: Baseline socio-demographics by follow-up data availability and formal assistance accessed in first three months Baseline data
Follow-up data
Y/N to formal help at
available
available
3-months
Variable
N Missing Marital Status
N Missing Age Group
N Missing Ethnicity - Priority
N Missing Ethnicity - Any
N Missing Employment status
23
17
15
Disabled/illness/sick leave Illness/sick leave 9
9
8
N Missing Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.2: Socio-demographics - continued Baseline data
Follow-up data
Y/N to formal help at
available
available
3-months
Variable
Highest educational
qualification achieved
None
Secondary school qualification Trade or technical certificate Professional qualification Undergrad. Dip. or Cert. University degree & above Degree - undergraduate Dip./Cert. - postgrad. Degree - postgraduate 11
10
8
N Missing Gross family income in last
12 months
<$20,000
$20,001 - $30,000 $30,001 - $50,000 $50,001 - $100,000 $100,001 - $200,000 14
13
11
N Missing Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.3: Area of residence Baseline data
Follow-up data
Y/N to formal help
available
available
at 3-months
Residential Area
5
5
5
47
37
34
Waikato/Coromandel 8
8
8
East Coast (Bay of Plenty/Lakes/Hawkes Bay) 25
23
23
Tarankai/Manawatu/ Palmerston North 20
17
13
19
18
16
8
7
7
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.4: Primary problem gambling mode Baseline data
Follow-up data
Y/N to formal
available
available
help at 3-months
Variable
Primary problem gambling mode
Casino gaming machines Casino table games Club gaming machines Pub gaming machines N Missing Primary mode - EGM
N Missing # Three participants reported multiple primary problem gambling modes for electronic gaming machines Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.5: Trends Assessment point
Baseline
12 months
days gambled
per month
N MISSING money lost per
N MISSING PGSI-12 (12
month time
N MISSING PGSI-12 (12
Non-problem N (%) month time
Moderate risk N (%) Problem gambler N (%) N MISSING PGSI-3 (3
month time
N MISSING Control over
gambling
behaviour
(scale 1 to 10)
N MISSING Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.4: Trends - continued Assessment point
Baseline
12 months
Work affected
in past month
(scale 1 to 10)
N MISSING Social life
affected in past
month (scale 1
N MISSING Family/home
affected in past
month (scale 1
N MISSING Physical health
affected in past
month (scale 1
N MISSING Received
assistance in
past 3 months
Any (formal + informal) N (%) N MISSING Quit all forms N (%) gambling goal
Quit some forms N (%) Control gambling N (%) Maintain abstinence N (%) N MISSING Kessler-10
Low (score 10 - 15) N (%) Medium (score 16 - 29) N (%) High (score 30 - 50) N (%) N MISSING PRIME-MD
depressive
disorder
N MISSING Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 2.5: Trends - continued Assessment point
Baseline
12 months
PRIME-MD
depressive
disorder
N MISSING PRIME-MD
Dysthymia
N MISSING PRIME-MD
disorder
N MISSING N MISSING Current tobacco
N MISSING No problem N (%) Substantial N (%) N MISSING Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 3
Predictors of utilisation of formal treatment services
Table 3.1: Univariate odds ratios for uptake of services by socio-demographic characteristics Variable
Category
Uptake of
Univariate odds ratio
services
Marital status
Age group
Prioritised ethnicity
Ethnicity - any
European
Asian & Other
Employment status
Disabled/illness/sick leave Highest educational
Secondary school qual. achieved
Trade/technical certificate Professional qualification Undergrad. Deg./Dip./Cert. University degree & above Gross family income
in last 12 months
$20,000 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Area of residence
Waikato/Coromandel Taranaki/Manawatu/ Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 3.2: Univariate odds ratios for uptake of services by baseline gambling and related behaviours Variable
Category
Uptake of
Univariate odds ratio
services
EGMs as gambling
type, dichotomised
Self-reported days
gambled per month in
past 2 months
(quartiles)
Self-reported amount
of money lost per day
in past 2 months
Level of motivation to
overcome problem
(scale 1-10)
Current goal,
Belief in treatment
Lower than median success, dichotomised
Higher than median Perceived level of
difficulty in
overcoming problem
(scale 1-10) (quartiles)
Length of problem
duration (months)
(quartiles)
Number of days since
last gamble
Current assistance for
gambling prob.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 3.3: Univariate odds ratios for uptake of services by other baseline covariates Variable
Category
Uptake of
Univariate odds ratio
services
PGSI-12 (12 month
time frame) (quartiles)
PGSI-3 (3 month time
(quartiles)
Kessler-10
(quartiles)
Audit-C, dichotomised
DAST, dichotomised
Suicide ideation
Prime MD - Major
depressive disorder
Prime MD -
Dysthymia
Prime MD - Minor
depressive disorder
Tobacco - Current
(quartiles)
(quartiles)
Treatment - drugs/
alcohol in last year
How was work
affected in past
month? (10 point
scale) (quartiles)
How was social life
affected in past
month? (10 point
scale) (quartiles)
How was family/ home
affected in past
month? (10 point
scale) (quartiles)
How was health
affected in past
month? (10 point
scale) (quartiles)
Legal problems in past
12 months
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 4
Predictors of successful problem gambling outcomes: PGSI
Table 4.1: PGSI-12 change at 12-month assessment by socio-demographic characteristics Variable
Category
Estimated least
Standard
squares mean diff.
Marital status
Age group
Prioritised
ethnicity
Ethnicity - any
European
Asian & Other
educational
Secondary school qual. Trade/technical certificate achieved
Professional qualification Undergrad. Deg./Dip./Cert. University degree & above Gross family
income in last 12
$20,000 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Area of residence
Waikato/Coromandel Taranaki/Manawatu/ Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 4.2: PGSI-12 change at 12-month assessment by baseline gambling and related behaviours Variable
Category
Estimated least
Standard
squares mean diff.
Self-reported days
gambled per month in
past 2 months
(quartiles)
Self-reported amount
of money lost per day
in past 2 months
Control over gambling
Level of motivation to
overcome problem
(scale 1-10)
Current goal,
Belief in treatment
Lower than median success, dichotomised
Higher than median Perceived level of
difficulty in
overcoming problem
(scale 1-10) (quartiles)
Length of problem
duration (months)
(quartiles)
Current assistance for
gambling prob.
Prev. assist. for
gambling prob.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 4.3: PGSI-12 change at 12-month assessment by other baseline covariates Variable
Category
Estimated least
Standard
squares mean diff.
Kessler-10 (quartiles)
Audit-C, dichotomised
DAST, dichotomised
Suicide ideation
Prime MD - Dysthymia
Tobacco - Current
WHOQoL (quartiles)
Treatment - drugs/
alcohol in last year
How was work affected
in past month? (10 point
scale) (quartiles)
How was social life
affected in past month?
(10 point scale)
(quartiles)
How was family/home
affected in past month?
(10 point scale)
(quartiles)
How was health affected
in past month? (10 point
scale) (quartiles)
Legal problems in past
12 months
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 5
Predictors of successful problem gambling outcomes: days gambled
Table 5.1: Days gambled change by socio-demographic characteristics Variable
Category
Estimated least
Standard
squares mean diff.
Marital status,
Age group
Prioritised
ethnicity
Ethnicity - any
European
Asian & Other
Employment
Disabled/illness/sick leave educational
Secondary school qual. Trade/technical certificate achieved
Professional qualification Undergrad. Deg./Dip./Cert. University degree & above Gross family
income in last 12
$20,000 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Area of residence
Waikato/Coromandel Taranaki/Manawatu/ Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 5.2: Days gambled change by baseline gambling and related behaviours Variable
Category
Estimated least
Standard
squares mean diff.
EGMs as gambling
type, dichotomised
PGSI-12 (12 month
time frame) (quartiles)
Self-reported amount
of money lost per day
in past 2 months
Control over gambling
Level of motivation to
overcome problem
(scale 1-10)
Current goal,
Belief in treatment
Lower than median success, dichotomised
Higher than median Perceived level of
difficulty in
overcoming problem
(scale 1-10) (quartiles)
Length of problem
duration (months)
(quartiles)
Number of days since
last gamble
Current assistance for
gambling prob.
Prev. assist. for
gambling prob.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 5.3: Days gambled change by other baseline covariates Variable
Category
Estimated least
Standard
squares mean diff.
Kessler-10 (quartiles)
Audit-C, dichotomised
DAST, dichotomised
Suicide ideation
Prime MD - Major
depressive disorder
Prime MD - Minor
depressive disorder
Prime MD - Dysthymia
Tobacco - Current
WHOQoL (quartiles)
NZDI (quartiles)
Treatment, mental
health last year
Prescription, mental
health last year
Treatment - drugs/
alcohol in last year
How was work affected
in past month? (10 point
scale) (quartiles)
How was social life
affected in past month?
(10 point scale)
(quartiles)
How was family/home
affected in past month?
(10 point scale)
(quartiles)
How was health affected
in past month? (10 point
scale) (quartiles)
Legal problems in past
12 months
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 6
Predictors of successful problem gambling outcomes: money lost
Table 6.1: Money lost change by socio-demographic characteristics Variable
Category
Estimated least
Standard
squares mean diff.
Marital status
Marital status,
Age group
Prioritised
ethnicity
Ethnicity - any
European
Asian & Other
Employment
Disabled/illness/sick leave educational
Secondary school qual. Trade/technical certificate achieved
Professional qualification Undergrad. Deg./Dip./Cert. University degree & above Gross family
income in last 12
$20,000 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Area of residence
Waikato/Coromandel Taranaki/Manawatu/ Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 6.2: Money lost change by baseline gambling and related behaviours Variable
Category
Estimated least
Standard
squares mean diff.
EGMs as gambling
type, dichotomised
Self-reported days
gambled per month in
past 2 months
(quartiles)
Control over gambling
Level of motivation to
overcome problem
(scale 1-10)
Current goal,
Belief in treatment
Lower than median success, dichotomised
Higher than median Perceived level of
difficulty in
overcoming problem
(scale 1-10) (quartiles)
Length of problem
duration (months)
(quartiles)
Number of days since
last gamble
Current assistance for
gambling prob.
Prev. assist. for
gambling prob.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 6.3: Money lost change by other baseline covariates Variable
Category
Estimated least
Standard
squares mean diff.
Kessler-10 (quartiles)
Audit-C, dichotomised
DAST, dichotomised
Suicide ideation
Prime MD - Minor
depressive disorder
Prime MD - Dysthymia
Tobacco - Current
WHOQoL (quartiles)
NZDI (quartiles)
Treatment, mental
health last year
Prescription, mental
health last year
Treatment - drugs/
alcohol in last year
How was work affected
in past month? (10 point
scale) (quartiles)
How was social life
affected in past month?
(10 point scale)
(quartiles)
How was family/home
affected in past month?
(10 point scale)
(quartiles)
How was health affected
in past month? (10 point
scale) (quartiles)
Legal problems in past
12 months
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 7
Predictors of successful problem gambling outcomes: control over gambling
Table 7.1: Control over gambling change by socio-demographic characteristics Variable
Category
Estimated least
Standard
squares mean diff.
Marital status
Age group
Prioritised
ethnicity
Ethnicity - any
European
Asian & Other
Employment
Disabled/illness/sick leave educational
Secondary school qual. Trade/technical certificate achieved
Professional qualification Undergrad. Deg./Dip./Cert. University degree & above Gross family
income in last 12
$20,000 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Area of residence
Waikato/Coromandel Taranaki/Manawatu/ Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 7.2: Control over gambling change by baseline gambling and related behaviours Variable
Category
Estimated least
Standard
squares mean diff.
EGMs as gambling
type, dichotomised
PGSI-12 (12 month
time frame) (quartiles)
Self-reported days
gambled per month in
past 2 months
(quartiles)
Self-reported amount
of money lost per day
in past 2 months
Level of motivation to
overcome problem
(scale 1-10)
Current goal,
Length of problem
duration (months)
(quartiles)
Number of days since
last gamble
Current assistance for
gambling prob.
Prev. assist. for
gambling prob.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 7.3: Control over gambling change by other baseline covariates Variable
Category
Estimated least
Standard
squares mean diff.
Audit-C, dichotomised
DAST, dichotomised
Suicide ideation
Prime MD - Minor
depressive disorder
Tobacco - Current
Prescription, mental
health last year
Treatment - drugs/
alcohol in last year
How was work affected
in past month? (10 point
scale) (quartiles)
How was social life
affected in past month?
(10 point scale)
(quartiles)
How was family/home
affected in past month?
(10 point scale)
(quartiles)
Legal problems in past
12 months
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
APPENDIX 8
Predictors of successful problem gambling outcomes: treatment success
Table 8.1: Univariate odds ratios for treatment success by socio-demographic characteristics Variable
Category
Univariate odds ratios
Marital status
Marital status,
Age group
Prioritised
(<0.001, 17.03) ethnicity
Ethnicity - any
European
Asian & Other
Employment
Disabled/illness/sick leave educational
Secondary school qual. Trade/technical certificate achieved
Professional qualification Number of observations too small Undergrad. Deg./Dip./Cert. University degree & above Gross family
income in last 12
$20,000 - $30,000 $30,001 - $50,000 $50,001 - $100,000 Area of residence
Waikato/Coromandel Taranaki/Manawatu/ Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 8.2: Univariate odds ratios for treatment success by baseline gambling and related behaviours Variable
Category
Univariate odds ratios
EGMs as gambling
type, dichotomised
PGSI-12 (12 month
time frame) (quartiles)
Control over gambling
Level of motivation to
overcome problem
(scale 1-10)
Current goal,
Belief in treatment
Lower than median success, dichotomised
Higher than median Perceived level of
difficulty in
overcoming problem
(scale 1-10) (quartiles)
Length of problem
duration (months)
(quartiles)
Number of days since
last gamble
Current assistance for
gambling prob.
Prev. assist. for
gambling prob.
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013
Table 8.3: Univariate odds ratios for treatment success by other baseline covariates Variable
Category
Univariate odds ratios
Kessler-10 (quartiles)
Audit-C, dichotomised
DAST, dichotomised
Suicide ideation
Prime MD - Major
depressive disorder
Prime MD - Minor
depressive disorder
Prime MD - Dysthymia
Tobacco - Current
WHOQoL (quartiles)
NZDI (quartiles)
Treatment, mental
health last year
Prescription, mental
health last year
Treatment - drugs/
alcohol in last year
How was work affected
in past month? (10 point
scale) (quartiles)
How was social life
affected in past month?
(10 point scale)
(quartiles)
How was family/home
affected in past month?
(10 point scale)
(quartiles)
How was health affected
in past month? (10 point
scale) (quartiles)
Legal problems in past
12 months
Effectiveness of problem gambling brief telephone interventions: An uncontrolled outcome study
Provider No: 467589, Contract No: 326673/00 and 326673/01
Gambling and Addictions Research Centre, Auckland University of Technology
Final Report, re-issued 18 July 2013

Source: http://www.nzct.org.nz/assets/Uploads/Gaming-research/effectiveness-brief-telephone-intervention-final-report-July-2013.pdf