Statewide Gambling Therapy Service South Australia Flinders Centre - - PowerPoint PPT Presentation

Statewide Gambling Therapy Service South Australia Flinders Centre for Gambling Research

www.sagamblingtherapy.com.au Malcolm Battersby Peter Harvey

Funding

• Office of Problem Gambling, Department of Families and Social

Inclusion

• Casino, SA Clubs and Aust Hotels Association via Office of

Problem Gambling

• Flinders University
• Acknowledgements:
• Jane Oakes, Ben Riley, Sharon Harris, Dave Smith, Amii

Larsen,

A woman has been jailed for at least two years for stealing more than $800,000 from her employers to feed her gambling habit

http://www.abc.net.au/news/2012-07-13/l

Gaming machines take over peoples’ minds

Larry (PA)

“The damn machines take your mind over, your body and soul over, and that’s probably the reason why people commit suicide, because it just takes control. But it’s – what would you call it, it’s being delirious.”

The urge and the "Zone" – loss of critical thinking

Simon (SGTS) “It’s the urge. Once you start it’s almost like you’re in the "Zone" and you ignore all random thoughts in your head that say ‘stop, get out, you’re just going to lose it’’. ‘I also can become hyper- focused where you’re so focused on

• ne machine you can’t leave it”

"Zone" unable to meet basic demands

altered awareness

Janet (PA)

“You just didn’t get a drink. You just

didn’t go to the toilet you didn’t get anything to eat. You had absolutely nothing all the hours that you were there’”.

Statewide Gambling Therapy Service

• Flinders Therapy Service for Problem Gamblers established in 1996
• Expanded in 2007 to become Statewide Gambling Therapy Services
• Inpatient / outpatient
• Consumer Consultants
• The SGTS is run through the Southern Adelaide Health Service (FMC) and

Flinders University funded by SA government with contributions from the industry

• Comprehensive assessment and treatment using mental health and Cognitive

Behaviour Therapy (CBT) urge reduction program for people with gambling problems

• Training opportunities for students (social work, mental health sciences,

psychology)

Well documented that ease of access increases problems…

EGMS most addictive form of gambling…Why?

• availability & access
• attraction of sights, sounds
• increasing speed and betting rates
• near misses
• dopamine pathways / reward system
• perates like a substance addiction

312.31 Gambling Disorder (DSM V )

Non substance related disorder - not better accounted for by a manic episode At least 4 of the following criteria

• Preoccupation
• Tolerance
• Loss of control
• Withdrawal
• Escape negative affect
• Chase losses
• Lying
• Risk significant relationship or job
• Relies on bailout

Theory: the development of problem gambling

• Initially: Classical conditioning – pairing excitement (arousal)

induced by the reward of gambling with neutral stimuli

• external eg the machine colours, sounds, the venue,

money,

• internal: positive mood, low mood, anger, boredom create

a gambling urge These stimuli then become triggers to gambling arousal without gambling occurring

External Triggers: Classical Conditioning

Gambling triggers - internal

Conditioning

• Then; Operant conditioning of gambling behaviour
• Positive - intermittent variable ratio reinforcement ie increase

in gambling behaviour mediated by arousal Unpredictable reward timing and amount Eg near miss Classical and operant conditioning lead to development and increase of the urge As the arousal becomes heightened it becomes aversive ie an ‘urge’ http://www.youtube.com/watch?v=I_ctJqjlrHA

Variable Ratio Reinforcement

Operant Conditioning – reward learning

• Money: actual or fantasy
• Cheap / free drinks and food
• Excitement, social interaction

Conditioning

Finally: Negative reinforcement via tension relief ‘a behaviour is reinforced or strengthened by avoiding or removing a negative outcome or stimulus’. ie the act of gambling reduces the urge temporarily but has the effect of increasing the strength of the urge the next time the person comes across a gambling trigger or cue

Why Cue Exposure?

• If for most problem gamblers an urge or arousal exists and

conditioning provides a reasonable explanation of the development of the urge then we should use a de-conditioning approach to extinguish the urge

• Gaming machines are designed using a range of reinforcement

schedules and provide the ideal experimental paradigm to test a treatment which aims to reverse the psychological process which created the behaviour (as compared to anxiety disorders which are not ‘created’ )

Why cue exposure?

• Problem Gambling in fact provides one of the rare occurrences in

mental disorders where a theoretical model of causation can directly support the selection of an intervention

• Large evidence base for use of exposure in anxiety disorders

where arousal or anxiety is linked to triggers for all 6 anxiety disorders and brief (2-12 sessions) even single session exposure (phobias) can lead to major improvements or even ‘cure’ of anxiety disorders

Exposure and response prevention

• Obsessive compulsive disorder provides the closest paradigm

to problem gambling

• Once the anxiety/urge has been overcome the avoidances,

distractions and modifiers become unnecessary

(Oakes et al, 2008,2010,2011)

Function

Sometimes solutions create more of the problem…

DIG!

DIG!

Costs of Avoidance

• Loss of independence
• Frustration
• Social isolation/loneliness
• Restricted lifestyle
• Conflict over money

Graded Cue-exposure with Response Prevention

Habituation

Screening Assessment

• Cognitive behavioural Assessment
• Mental state
• Risk assessment
• Suitability (exclude acute psychosis, suicide plans,

recreational drug use, benzodiazepine use, alcohol ) The problem is predictable and observable.

• Treatment Rationale

STATEWIDE GAMBLING THERAPY SERVICE Gambling Screening Assessment

Assessing Therapist Full Name, *Signature *Screening Date PATIENT LABEL Patient Profile *Age, *Marital Status *Occupation, *Suburb lived in What Is the problem at the moment? Where Does the client gamble or not? When Does the client gamble or not? Why/Triggers Does the client gamble? With Whom Does the client gamble and when gambling, does the client socialise?

Specific Incident Autonomic What happens physically when having an urge to gamble eg; sweating and palpitations? Behaviour How does the client place bets, any superstitious habits, take a break, smoke or drink? Cognitive Thoughts/Imager y of gambling, self talk erroneous beliefs? Before During After

Excitement Anxiety Restlessness Rushing I shouldn’t go Only $20 I might win! My lucky day! Money Hotels Bills Sadness Anger Happiness Gambling relieves the urge temporarily

Uncontrollable urge

(Battersby et al., 2008)

Gambling episodes within pattern of problem gambling

Gambling reduces urge: reinforces gambling; trigger maintains urge- eliciting power. No reappraisal.

Exposure - Habituation

1 2 3 4 5 6 7 8 Time

URGE Don't avoid H a b i t u a t i

• n

Principles of Exposure

• Graded: tackle the easiest trigger first
• Prolonged: Remain with the trigger until your urge has reduced

by at least 50% - you must feel the urge go away

• Repeated: repeat or practice the task at least 5 times per week

preferably daily

• Focussed: don’t distract yourself, allow the gambling thoughts

to stay and don’t replace with correcting thoughts

Graded Cue-exposure with Response Prevention A typical hierarchy for EGM problem

Client achieves extinction

• Picture

Start

• Music
• Simulator/DVD
• Outside venue (no $)
• Inside venue at lounge area (no $)
• Outside venue with small amount of $
• Sitting at EGM with small amount of $
• Gradually increase amounts of $ until goal reached

Imaginal exposure (guided imagery)

• 30-45 mins in session
• A step by step account in the here and now fully immersed in all

the sensory facets of the gambling scenario

• Use the 4 principles of exposure to guide the content and

duration of the imagery task

• Complete the scenario up to the point of gambling but not the

act of gambling

• Audio tape or video the client description of the session
• Replay at home for homework using the 4 exposure principles

Therapy session plan

1. CBT Assessment and rationale, baseline measures, money management 2. Full psychiatric and psychosocial history,

1. Family involvement 2. Review rationale 3. Review money management 4. Exposure tasks agreed, use urge monitoring record and treatment diary

3. Homework reviewed, cash management reviewed, new exposure tasks in hierachy

1. Urge surfing for spontaneous urges

Therapy sessions

• 4. review homework,

1. Repeat measures 2. In session cue exposure 3. Guided imagery exercise (optional)

• 5. Use sounds and DVDs of gambling cues for exposure

Site visit – if required

Sessions 6-12 – follow steps up the hierachy until client able to sit in venue with coins or credits in the machine and not press the button and have no urge Repeat measures every 4 weeks

Stimulus Control (money management)

• The urge is often uncontrollable and money is a

powerful trigger

• Clients are advised to remove or limit access to

cash during early stages of exposure

• Allows effective grading of tasks

Relapse Prevention

• Harm minimisation
• Cash restrictions
• Support person
• Early warning signs
• Stress vulnerability
• Negative affect
• Internal triggers

Other forms of gambling

> TAB > Internet http://www.youtube.com/watch?v=sJFXGJ3GBoU

Outcome measures

• VGS
• CPGI
• Gambling behaviours
• Gambling urge scale
• Gambling cognitions
• Work and social adjustment
• Kessler 10
• Audit

Name Post code Age Gender Therapist Screening appt Attended Y/N Suitable for inpatient/group/individual - Accepted treatment Y/N Diagnosis 1 (DSM-IV) Diagnosis 2 (DSM-IV)

Problem

A When alone, stressed and have money I gamble on EPM causing financial, emotional and relationship problems . It has also lead to an inability to carry cash

Screening Self Therapist

Assess 8 8 Mid 6 6

Post 2 2

1MFU

3MFU 6MFU 1YFU

Goal 1 I will be able to take $50 into a venue sit at the Indian Dreaming machine , stay 30 minutes and leave without playing 3 times a week for a month

Self Therapist

8 8 6 6

Goal 2 I will be able to save $50 a week towards my debts and then a holiday to Bali

Self Therapist

8 8 4 4

EGM hours 30 Other Professional Involvement professional Details Other Gambling Hours VAGS 30

4

CPGI 21

• KESSLER (K10)

31

20 17

• WSAS Home Management

8

4 2

WSAS Social Leisure 6

2

WSAS Private Leisure 2

2

WSAS Relationships 8

6 2

WSAS Work 2 Goldney SI Scale 4 AUDIT 8

• GRCS

76

23 23 23

Urge 21

3

Statewide Gambling Therapy Service Data Summary Sheet

Future Project Options

48

Screening Outpatient Program Weekly Therapy Inpatient Program 2 Week Admission Discharged to Follow-Up Follow-up Program

1, 3, 6, 12, 24, and 36 months

Research

Longitudinal Outcomes Rationale

• controlled studies are designed to test key elements of an

intervention or to compare one approach to treatment with another ie CBT and general counselling

• in the day to day operation of a treatment service, client progress

through treatment is haphazard, so naturalistic study designs may provide a more realistic assessment of service effectiveness over time

• it is important to know what actually happens to clients on their

journey through treatment not just what is intended to happen or what happens to ‘ideal’ clients

51

Approach to managing gambling outcome data

• individuals are tracked over time and changes in outcome

indicators modelled using observed and fitted estimates

• data from the included population is then analysed using

time as a continuous variable using random effects modelling

• note…ANOVA and other forms of analysis with fixed data

points require imputed or replaced data otherwise cases missing time points are dropped from the analysis

52

Context

• clients consent to 3 year follow-up when engaging in

treatment

• key clinical and gambling activity data are routinely

collected at baseline and every 4 sessions as well as on discharge and at 1,3,6 & 12 months with 2 & 3 year follow-up

• we focused on VGS, K10 and WSAS scores over time for

a population of 664 unique clients registered for treatment

53

This approach accepts that:

• people may stop and re-start treatment
• ther treatments may coincide with the focus treatment
• life events interfere with the treatment process
• motivation to engage may change
• therapists change
• clients accept some treatments and not others
• service provision is not laboratory controlled
• time intervals for treatment and data collection vary
• data is not always available / provided
• longer term follow-up is difficult

54

evaluating service treatment outcomes: a risk-adjusted model

Baseline socio-demographics and clinical characteristics of n = 664 problem gamblers

Variable Value

Socio-demographic data Age (years)

43.02 (13.38)

Male

367 (55.11)

Relationship

married/de facto 250 (38.17) separated/divorced 173 (26.41) never married 194 (29.62) widowed 22 (3.36)

• ther

16 (2.44) Employed 377 (57.91)

Duration of gambling problem

< 1 year 79 (12.44) 1 - 5 years 202 (31.81) 6 -10 years 145 (22.83) > 10 years 209 (32.91)

Primary form of gambling

gaming machines 528 (80.73) horse/dog racing 86 (13.15)

• ther

40 (6.12)

Outcome measures VGS

38.89 (11.81)

K10

28.58 (9.93)

WSAS

14.68 (10.12) Abbreviations: VGS, Victorian Gambling Screen harm to self sub-scale; K10, Kessler 10 Scale; WSAS, Work and Social Adjustment Scale; Data are mean (SD), or n (%) unless otherwise indicated.

55

PGSI profile for standard cohort (n = 491)

56

Age distribution

57

I ndigenous client numbers increasing

58

final model

• all outcomes showed a statistically significant reduction (improvement) over

time (P < 0.001) e.g. on average, VGS decreased by 4.81 units for each increase in time (months) when holding all other variables constant

59

longitudinal outcomes VGS

Predictive margins of gambling duration status with 95% confidence intervals 10 20 30 40 VGS 3 6 9 12 Time (months) 60

longitudinal outcomes K10

Predictive margins of gambling duration status with 95% confidence intervals 10 15 20 25 30 K10 3 6 9 12 Time (months) >10 years 5-10 years 1-5 years < 1 year 61

longitudinal outcomes WSAS

Predictive margins of gambling duration status with 95% confidence intervals 5 10 15 WSAS 3 6 9 12 Time (months) >10 years 5-10 years 1-5 years < 1 year 62

interactive effects VGS

Predictive margins of gender status with 95% confidence intervals. 10 15 20 25 30 VGS 18 28 38 48 58 68 78 Age (years) female male 63

summary

• time in treatment is a key element in outcome achievement
• this work provides evidence for modifying the treatment

regimen to improve compliance and completion rates

• Some relapse and clients stop and start therapy due to

changing circumstances and motivation

64

Relapse Study (Battersby et al 2010)

Significant predictors of problem gambling and relapse with VGS as the outcome measure: Relapse and continuing to gamble: Gambling urge Gambling related cognitions

Cognitive versus behavioural (exposure-based) therapy:

• Parallel, two group randomised trial
• Superiority design – BT will be superior to CT
• Single site
• Manualised (CT:R Ladouceur; ET: SGTS)
• Therapy fidelity checks
• all therapy sessions audio recorded
• 20% randomly sampled & checked
• Primary outcome VGS (baseline, Tx-end, 1m, 3m, & 6m)

Participant eligibility

• ≥ 18 years of age
• Treatment seeking for problem gambling with EGMs
• Not involved in concurrent gambling treatment program
• Not received psychological treatment for problem gambling in

the previous 12 months

• SOGS ≥ 5
• Not suicidal or experiencing mental distress

Randomization

• 1:1 allocation ratio.
• Block sizes 2,4,6
• Stratification:
• age 18 ‐ 42 years / 43+ years
• gender
• gambling severity (SOGS: 5 – 11 / 12+)
• Allocation by clinical trials centre

Participant flow

Participant flow (con’t)

Implementation

• For participants who started intervention (n=87):
• median CT sessions 8.5 (IQR, 4 ‐ 11.5)
• Median ET sessions 5 (IQR, 3 ‐ 9) (P = 0.046)
• Mean session duration:
• CT (51.9 minutes, SD=16.3)
• ET (43.3 minutes, SD=20.9) (P < 0.001)
• Median number of weeks participants engaged in treatment:
• CT (Median = 13.5; IQR, 6.9 – 21.6)
• ET (Median = 9.6; IQR, 2.7 ‐ 20.7) (P = 0.316)

Implementation (con’t)

• On therapist’s judgement, 41% (36/87) therapy drop‐outs:
• 31.8% (14) for CT
• 51.2% (22) ET
• 66.7% attended 1 ‐ 3 sessions
• No difference in session numbers for completers (9.5 vs 9.0)
• No difference in therapy duration:
• CT (Median = 16.6; IQR, 11.9 ‐ 24.1)
• ET (Median = 18.1; IQR, 12.0 ‐ 28.7) (P = 0.893)
• Therapy fidelity: CT 98.5% (SD=4.4%); ET 99.5% (SD=2.8%)

Preliminary results

Individual response profiles for Victorian Gambling Screen (VGS) scores by treatment completion status, treatment group and time.a

20 40 60 20 40 60 Base Tx end 1 m 3 m 6m Base Tx end 1 m 3 m 6m

Treatment completers (n=30) Treatment completers (n=21) Treatment drop-outs (n=14) Treatment drop-outs (n=22)

VGS scores Time

Cognitive therapy (N=44) Exposure therapy (N=43) Lower scores indicate a reduction (improvement) in gambling symptom severity. Note: a Horizontal line is VGS cut score of 21+ (indicative of problem gambler).

Results

VGS (Victorian Gambling Screen)

• Average assessments per individual 2.9 (1 ‐ 5)
• No group difference in rate of change in scores over time (P =

0.477).

• Significant reduction (improvement) within groups (P < 0.001).

For each one week increase in time VGS score decreased by:

• 1.93 (95% CI: 1.65 – 2.22) in CT participants
• 1.87 (95% CI: 1.60 – 2.13) in ET participants.

change over time – RCT (VGS)

(n=43)(n=43) (n=31)(n=25) (n=22) (n=14) (n=19) (n=17) (n=22)(n=18)

Problem gambling Non-problem gambling 10 20 30 40 50 Mean VGS score Baseline Tx end 1 m 3 m 6 m Time Cognitive therapy Exposure therapy

76

change over time – (GUS)

(n=44)(n=43) (n=39)(n=34) (n=24) (n=15) (n=21) (n=17) (n=25) (n=17)

5 10 15 20 Mean GUS Baseline Tx end 1 m 3 m 6 m Time Cognitive therapy Exposure therapy

77

Findings (con’t)

(n=43)(n=43) (n=39) (n=34) (n=24)(n=15) (n=21)(n=17) (n=25)(n=17)

20 40 60 80 Mean GRCS score Baseline Tx end 1 m 3 m 6 m Time Cognitive therapy Exposure therapy

Lower scores indicate a reduction (improvement) in gambling related cognitions.

Observed Gambling Related Cognitions Scale (GRCS) scores by time and treatment group.

change over time - RCT pilot (K10)

(n=44)(n=43) (n=31)(n=25) (n=22)(n=14) (n=21)(n=17) (n=25) (n=18)

Severe Moderate Mild 10 15 20 25 30 35 Mean K10 score Baseline Tx end 1 m 3 m 6 m Time Cognitive therapy Exposure therapy

79

Qualitative

• 2. Specific effects for CT participants

>Yeah, the strategies, you know, like she was saying ‘You have to be able to recognise what’s a gambling thought and what’s not a gambling thought’. Before, I used to drive past the casino and I used to go, ‘Oh I’ll stop and play’ and now I say to myself, ‘Oh do you really want to go and play? Are you really thirsty?’ So you actually question your ideas of why you want to go in there. (CT04,COM)

Qualitative Specific effects for ET participants

>…Once you’ve got that urge being a gambler, it controls you more than anything else. You thought I was always going with it, whereas now the urge isn’t there, I’ve just, I’ve lost interest, it’s just nothing, it doesn’t, hasn’t got that magnet to it to pull me in anymore. (ET03,COM

Study 2 - A component analysis of cognitive- behavioural therapies versus treatment as usual

Randomised controlled trial – BT (exposure‐based) – CT developed by Robert Ladouceur and colleagues – CBT a balanced and flexible application of cognitive and cue‐exposure – TAU (control) based on Max Abbott, David Hodgins approach

Outcomes for an inpatient group (n = 53)

83

Victorian Gambling Screen (VGS) (n = 53)

84

Kessler 10 (K10) (n = 53)

85

Ongoing Research

• adapting tools & treatments for Aboriginal and Vietnamese communities (eg

CPGI validation for Aboriginal people)

• cultural impacts of gambling for Vietnamese people
• Department of Corrections treatment programme for prisoners with gambling

problems

• Family violence and gambling ARC linkage – Allun Jackson, Nicki Dowling et al
• Self-help on-line interventions (Deakin Uni, Turning Point) VPGF
• non-inferiority study of CBT/ET service efficiency vs treatment efficacy

superiority studies

86

progress to date is predicated on…

• consistent team & core business focus
• effective business model and site variations satisfying

the expectations of our funders

• quality improvement linked to research
• training and staffing model via MMHSc
• improved data management system
• Incorporating research findings into practice

SGTS

http://www.sagamblingtherapy.com.au/

88

THANK YOU

Malcolm.battersby@flinders.edu.au Peter.harvey@health.sa.gov.au