Addiction (2001) 96, 1279–1287

RESEARCH REPORT

Randomized controlled trial of briefcognitive-behavioural interventions amongregular users of amphetamine

AMANDA BAKER, TRACY G. BOGGS & TERRY J. LEWIN

Centre for Mental Health Studies, University of Newcastle, Callaghan, New South Wales,Australia

AbstractAims. To identify whether brief cognitive-behavioural interventions are feasible among regular users ofamphetamine, to assess the effectiveness of intervention overall and to pilot two- and four-session interven-tions. Design. Subjects were assigned randomly to individually receive a cognitive-behavioural intervention(n 5 32) of either two or four sessions’ duration or a self-help booklet (control condition; n 5 32). Setting.Subjects were volunteers recruited from needle exchange schemes and treatment centres in Newcastle,Australia. Participants. Regular (at least monthly) users of amphetamine were recruited. Intervention.Either four sessions of cognitive-behaviour therapy, consisting of a motivational interview and skills trainingin avoidance of high-risk situations, coping with craving and relapse prevention, or two sessions consisting ofa motivational interview and discussion of skills. Measurements. The Opiate Treatment Index was themain measure at pre-treatment and 6-month follow-up. Findings. There was a signi� cant reduction inamphetamine use among the sample as a whole, with inconclusive differences between intervention subgroups.There was a moderate overall intervention effect, with the intervention group reporting over twice thereduction in daily amphetamine use as the control group. Signi� cantly more people in the cognitive-behavioural intervention condition abstained from amphetamine at 6-month follow-up compared to thecontrol condition. Conclusion. Brief cognitive-behavioural interventions appear feasible among regularusers of amphetamine. A larger randomised controlled trial of the effectiveness of such interventions appearswarranted.

IntroductionAmphetamine is the second most commonlyused illicit drug in Australia (Makkai & McAllis-ter, 1998) and the United Kingdom (Klee,Wright & Morris, 1999) and its use is increasingin the United States (Proudfoot & Teesson,2000). The most recent Australian National

Drug Strategy household survey indicated thatamphetamine has been widely available duringthe last decade and signi� cant numbers of peo-ple have used the drug regularly (Makkai &McAllister, 1998). In 1999, amphetamine wasthe most commonly injected drug in the state ofQueensland, Australia (Illicit Drug Reporting

Correspondence to: Amanda Baker, Centre for Mental Health Studies, University of Newcastle, University Drive,Callaghan, New South Wales 2308, Australia. Tel: 61 2 49246610; fax: 61 2 49246608; e-mail: amanda.baker@newcastle.edu.au

Submitted 27th September 2000; initial review completed 9th January 2001; � nal version accepted 6th March2001.

ISSN 0965–2140 print/ISSN 1360–0443 online/01/091279–09 Ó Society for the Study of Addiction to Alcohol and Other Drugs

Carfax Publishing, Taylor & Francis Limited

DOI: 10.1080/09652140120070337

1280 Amanda Baker, Tracy G. Boggs & Terry J. Lewin

System, 1999). Heavy and continued use of am-phetamine can induce psychosis, rages and vio-lent behaviour and is associated with serioushealth problems such as stroke (Proudfoot &Teesson, 2000).

Despite the popularity of amphetamine andincreasing regular use of the drug and its associ-ated problems, there is a paucity of researchevaluating the effectiveness of interventions forregular users of amphetamine. Apart from a re-cent randomised controlled trial of the feasibilityof monitoring controlled prescribing of dexam-phetamine (Shearer et al., 1999), the authors areunaware of any other randomized controlled tri-als evaluating non-pharmacological interventionstargeted speci� cally for regular users of am-phetamine. In their review of available interven-tions for psychostimulant users, Kamieniecki etal. (1998) concluded that the non-pharmacologi-cal interventions which have demonstrated themost ef� cacy in treating psychostimulant (pri-marily cocaine) users are cognitive-behavioural,particularly relapse prevention for heavy users.They suggested that given the much higher useof amphetamine than cocaine in Australia, andthe lack of studies on the use of relapse preven-tion techniques in treating amphetamine users,further studies of relapse prevention interven-tions should target amphetamine users. Hando,Topp & Hall (1997) have addressed the questionas to whether an amphetamine-speci� c treat-ment is needed, given that polydrug use is almostuniversal among users of amphetamine. Theyanswered af� rmatively, pointing out that an am-phetamine dependence syndrome is a commonharm associated with regular use of the drug,prompting users to seek treatment.

The present study was designed with the pri-mary aim of testing the feasibility of conductingand evaluating cognitive-behavioural interven-tions for regular users of amphetamine. As thiswas the � rst trial of this nature, it was necessaryto ascertain whether regular users of am-phetamine would present themselves to the re-searchers and be recruited into the study,retained in treatment and available for follow-up.It was hypothesized that a cognitive-behaviouralintervention would be more effective than a self-help booklet control condition in reducing am-phetamine use and other drug-related harms.The net bene� t of treatment was assessed. Twointerventions were piloted: a four-session cogni-tive-behavioural intervention involving motiva-

tional interviewing and skills training and atwo-session intervention involving motivationalinterviewing and discussion of skills. The presentstudy sought to examine whether regular am-phetamine users would attend an intervention aslengthy as four sessions. It was hypothesized thatif four sessions were acceptable to subjects thena four-session intervention would be more effec-tive than a two-session intervention in reducingamphetamine use and related harms.

MethodsDesignSubjects were allocated randomly to either anintervention group (two or four sessions of cog-nitive-behaviour therapy in addition to a self-help booklet) or to a control group (self-helpbooklet alone). The self-help booklet was devel-oped by the National Drug and Alcohol Re-search Centre (NDARC, 1997). Assessmentswere scheduled at pre-treatment and 6 monthsfollowing the pre-treatment assessment. Follow-up assessments were conducted by an inter-viewer blind to the subject’s group allocation.

SubjectsThe subjects were 64 regular users of am-phetamine (at least monthly) and were recruitedin the Newcastle region of New South Wales(150 km north of Sydney). Polydrug users andpeople enrolled in methadone maintenancetreatment (MMT) were not excluded from thestudy provided they reported regular use of am-phetamine. This is consistent with the approachtaken in other recent studies of amphetamineusers (e.g. Shearer et al., 1999; Gossop, Marsden& Stewart, 2000). All subjects were volunteersand were paid a nominal amount ($20) for at-tendance at each assessment session. Theamount was assumed to be small enough not toin� uence response to the intervention but ad-equate to reduce non-compliance caused by anyinconvenience in attending sessions.

ProcedureSubjects were recruited between July and De-cember 1998 by means of notices placed withinvarious agencies, cafes and treatment centres andan inner-city needle-exchange scheme in New-castle, as well as through word of mouth. Notices

Brief interventions among regular amphetamine users 1281

stated that university researchers needed the helpof regular users of amphetamine by participatingin interviews for which they would be reim-bursed $20. Interviews took 45–60 minutes tocomplete. When a subject volunteered, the pur-pose and design of the study were described.They were also informed that on giving writtenconsent to participate in the study, they wouldbe assigned randomly to either a control group(assessment plus self-help booklet), two sessionsof counselling or to four sessions of counselling.Subjects were assured that all information wasstrictly con� dential, that researchers were inde-pendent of the agencies in which the interviewswere being conducted, and that refusal to par-ticipate would not affect their relationship withthe clinic in any way. Half of the subjects(n 5 32) were allocated randomly to the controlgroup and half were allocated randomly to anintervention condition (16 subjects to two ses-sions and 16 subjects to four sessions).

MeasuresPre-treatment measures were documented in anearlier paper (Baker, Boggs & Lewin, 2001)and are described here only brie� y. Data col-lected at the pre-treatment assessment included:demographic characteristics; history of any treat-ment for drug dependence; current drug use;exposure to blood-borne virus infections; andhistory of imprisonment. In addition, the follow-ing instruments were administered to all sub-jects: the Opiate Treatment Index (OTI; Darkeet al., 1992), the amphetamine version of theSeverity of Dependence Scale (SDS; Gossop etal., 1995) and the Contemplation Ladder (Bi-ener & Abrams, 1991). The OTI measured am-phetamine and other drug use, HIV risk-takingbehaviour (Darke et al., 1991), drug-relatedhealth outcomes, social functioning, psychologi-cal health via the 28-item General Health Ques-tionnaire (GHQ-28) (Goldberg & Hillier, 1979)and crime. All measurements on the OTI relatedto the 1 month period prior to interview exceptfor the Social Scale which assessed the 6 monthsprior to interview. Higher scores indicate higherlevels of dysfunction on all scales. Scores on theContemplation Ladder referred to stage ofchange for reducing amphetamine use. Becausesubjects’ responses were clustered around the� ve rungs with verbal anchors, the categorieswere subsequently collapsed as follows: 0, “No

thought of quitting or cutting down” or 1,“Think I need to consider quitting someday”(precontemplation); 2, “Think I should quit butnot quite ready” (contemplation); 3, “Starting tothink about how to change my use patterns”(preparation); and 4, “Taking action to quit orcut down” (action). At the 6-month follow-upassessment the OTI was readministered, whichincluded measures of amphetamine and otherdrug use, HIV risk-taking behaviour, drug-related health outcomes, social functioning, psy-chological health (GHQ-28) and crime.

Four-session cognitive-behavioural interventionThese sessions were conducted individually andlasted 30–60 minutes. Each session focused onthe acquisition of different skills aimed at helpingthe subject to reduce amphetamine use. Sessionswere guided by a therapist manual and involvedrole-plays. In addition to a self-help booklet onreducing amphetamine use and related harms(NDARC, 1997), subjects were given wallet-sized pamphlets to refer to when practising newskills for homework between sessions. The � rstsession occurred immediately following the pre-treatment assessment and involved a motiva-tional interview (Miller & Rollnick, 1991) whichaimed to raise motivation to reduce am-phetamine use. The following three sessions fo-cussed on cognitive-behavioural coping strategiesand relapse prevention, using techniques devel-oped by Marlatt & Gordon (1985). In the se-cond session the focus was on assistingidenti� cation of high-risk situations for am-phetamine use. The third session addressed theissue of craving. Subjects were taught how toreduce craving with progressive muscular relax-ation and coping self-talk. A relaxation tape wasprovided for practice between sessions. Thefourth session focused on coping with lapses andpreparation of a coping drill for use in futurehigh-risk situations and following lapses.

Two-session cognitive-behavioural interventionThe procedure and content of the � rst sessionwas the same as described above for the longerintervention. Cognitive-behavioural copingstrategies re� ecting some of the content of ses-sions two, three and four of the longer interven-tion were discussed during the second session.

1282 Amanda Baker, Tracy G. Boggs & Terry J. Lewin

The same self-help booklet (NDARC, 1997) andpamphlets were given to subjects.

Control groupSubjects were given the same self-help bookleton reducing amphetamine use and related harms(NDARC, 1997) as the intervention conditions.

TherapistsThe therapists were two research assistants with4 years’ training in psychology. A therapist man-ual developed for the project by the � rst authorwas used in both initial training and weeklysupervision.

Data analysisIn view of attendance patterns among the inter-vention group, and as detailed below, these sub-jects were reassigned to three interventionsubgroups according to the number of sessionsattended (one, two or three–four sessions). Datawere analysed using SPSS for Windows (version9) and PSY2000 (Bird, Hadzi-Pavlovic & Isaac,2000). Categorical outcome variables (e.g. absti-nence vs. continued use) were analysed usingchi-square tests. For the continuous outcomevariables (e.g. amphetamine use), analyses ofvariance (ANOVA) were performed usingplanned contrasts, in which we examined allcomparisons between the subgroups as well asoverall differences between the control and thecombined intervention subgroups. Repeatedmeasures ANOVAs were conducted to ascertainthe relationship between treatment and the fol-lowing variables: level of amphetamine and otherdrug use; level of injecting risk-taking behaviour;level of psychopathology; health; social function-ing and criminal involvement. As a partial con-trol for the number of statistical tests, thethreshold for signi� cance was set at p , 0.01.

ResultsPatterns of participationThere were 32 control subjects. Of the 16 sub-jects assigned to the two-session intervention, 11attended both sessions (11/16, 68.8%). Of the16 subjects assigned to the four-session con-dition, eight attended all four sessions and oneattended three sessions (9/16, 56.3%), with the

remainder attending only the � rst session. Analy-ses of data on key prognostic variables, age,gender, years of education, duration of am-phetamine use, stage of change for amphetamineuse and level of amphetamine use during themonth prior to the initial interview indicated thatthere were no signi� cant differences betweenintervention subjects who completed the sessionsto which they were assigned (n 5 20) and thosewho did not (n 5 12). Accordingly, brief inter-vention subjects were reassigned to the followingsubgroups on the basis of the number of sessionsactually attended: B1, one session, B2, two ses-sions or B3, three/four sessions.

Fifty-two subjects (81.3%) were successfullyfollowed-up and there was no signi� cant differ-ence in retention rates across groups (controls,n 5 28, 87.5%; intervention subgroups: B1,n 5 8, 66.7%; B2, n 5 9, 81.8%; B3, n 5 7;77.8%; v 2(3) 5 2.57, NS). Analysis of data onthe same key prognostic variables above indi-cated that there were no signi� cant differencesbetween subjects who were followed up (n 5 52)and subjects who were not followed-up (n 5 12).

Sample characteristicsOverall pre-treatment sample characteristics andpatterns of drug use have been reported previ-ously (Baker et al., 2001). The main pre-treatment characteristics of the control andintervention groups are shown in Table 1. Analy-sis of data on the same key prognostic variablesabove indicated that there were no signi� cantdifferences. The majority of the subjects weremale, with long histories of amphetamine use,and more than one-third were currently enrolledin MMT. Approximately one-� fth (21.9%) ofthe sample were at the pre-contemplation stagefor quitting or reducing amphetamine use,37.5% were at the contemplation stage, 23.4%were at the preparation stage and 17.2% were atthe action stage.

Changes in drug useMean pre-treatment, 6-month follow-up andchange scores are reported in Table 2 for am-phetamines, the two most commonly used drugs,cannabis and tobacco and polydrug use. Stan-dardized change scores and abstinence rates forthese outcome measures are also reported in

Brief interventions among regular amphetamine users 1283

Table 1. Pre-treatment sample characteristics by group

Control group Intervention group(n 5 28)a (n 5 24)a

Demographic characteristics% Male 57.1% (16) 66.7% (16)Mean age (years) 30.57 (8.57) 32.79 (8.76)Mean education (years) 10.66 (1.86) 10.50 (1.74)

Amphetamine useMean duration of regular use (years) 10.25 (7.03) 10.92 (7.84)Mean daily level of amphetamine

use (OTI) 0.83 (1.03) 1.20 (1.65)Mean stage of change for

amphetamine use (0–4) 2.39 (1.07) 2.17 (1.40)% Enrolled in methadone maintenance

treatment (MMT) 39.3% (11) 33.3% (8)

a Tabled values are percentages (and frequencies) or mean scores (with standarddeviations).

Table 2; the former facilitate “effect size” com-parisons with other intervention studies.

Overall, the repeated measures (treatmentgroup 3 time) ANOVAs revealed signi� cantmain effects for time, but no signi� cant groupmain effect contrasts or any signi� canttime 3 group interaction contrasts. Am-phetamine use fell signi� cantly for the sample asa whole [F(1,48) 5 17.80, p , 0.001] and therewas a (non-signi� cant) tendency for this fall tobe greater among the B2 (two-session) interven-tion group compared to the control group[F(1,48) 5 6.80, p , 0.05]. While not statisticallysigni� cant [relevant interaction contrast:F(1,48) 5 2.36, p , 0.15], mean daily occasionsof use of amphetamine fell 0.44 units among thecontrol group versus 1.02 units among the inter-vention group as a whole. Expressed in effect sizeunits (0.40 vs. 0.93), this represents a differenceof over half a standard deviation, a moderateeffect size. There was a signi� cant associationbetween group membership and abstinence fromamphetamines at follow-up [ v 2(3) 5 11.66,p , 0.01]. The overall difference in abstinencerates between the control condition subjects (6/28, 21.4%) and the intervention group (14/24,58.3%) [ v 2(1) 5 7.43, p , 0.01] was due largelyto differences between the control and B3 condi-tions (three/four sessions: 6/7, 85.7%)[ v 2(1) 5 10.27, p , 0.01].

There was no signi� cant change in cannabis[F(1,31) 5 3.54, NS] or tobacco usage[F(1,44) 5 2.67, NS] over time nor any differen-

tial change by group. Similarly, there was noassociation between group membership and ab-stinence from cannabis [ v 2(3) 5 2.28, NS] ortobacco [ v 2(3) 5 5.49, NS] at follow-up.Re� ecting the signi� cant reduction in am-phetamine use among the sample as a whole,there was a signi� cant overall reduction in poly-drug use over time [F(1,48) 5 12.71, p , 0.01].There was also a (non-signi� cant) tendency forthe reduction in polydrug use to be greateramong the B1 (one-session) intervention groupcompared to the B2 (two-session) interventiongroup [F(1,48) 5 5.15, p , 0.05].

Despite the evidence suggesting some treat-ment bene� ts, it is worth noting that amongthose who received an intervention, the meanOTI scores at follow-up (Table 2) indicate thatthe typical subject was using amphetamines atleast weekly, cannabis three times a day, 19cigarettes per day, and three-and-a-half classes ofdrugs per month.

A 2 3 2 3 (2) repeated-measures ANOVAwas conducted to assess the contribution ofstage of change to patterns of amphetamine use(i.e. action stage of change vs. early stage ofchange 3 control vs. intervention 3 time). Therewere no signi� cant main or interaction effectsinvolving stage of change. A similar analysisexamining the effects of duration of regularamphetamine use (8 years or less vs. more than8 years) revealed no signi� cant main or interac-tion effects involving duration of amphetamineuse.

1284 Amanda Baker, Tracy G. Boggs & Terry J. Lewin

Tab

le2.

Sel

ecte

dO

TI

pre-

trea

tmen

t,fo

llow

-up,

chan

ge,

and

abst

inen

cein

dice

sby

trea

tmen

tgr

oup

Mea

nO

TI

scor

e(S

D)

Sta

nd

ard

ised

%A

bst

inen

tO

TI

dru

gu

se6

mon

thch

ange

at6

mon

thca

tego

ryT

reat

men

tgr

oupa

nbP

re-t

reat

men

tfo

llow

-up

Cha

nge

(eff

ect

size

un

its)

cfo

llow

-up

Am

phet

amin

esC

:C

ontr

ol28

0.83

(1.0

3)0

.39

(0.6

2)

0.4

4(1

.28)

0.40

21.4

%B

:B

rief

inte

rven

tion

241.

20(1

.65)

0.1

8(0

.52

)1

.02

(1.2

3)0.

9358

.3%

B1

:1

sess

ion

80.

84(0

.93)

0.0

6(0

.13

)0

.78

(0.9

6)0.

7162

.5%

B2

:2

sess

ion

s9

2.10

(2.3

3)0

.43

(0.8

0)

1.6

7(1

.63)

1.53

33.3

%B

3:

3–4

sess

ion

s7

0.46

(0.2

9)0

.01

(0.0

1)

0.4

5(0

.29)

0.41

85.7

%O

vera

ll52

1.00

(1.3

5)0

.29

(0.5

8)

0.7

1(1

.28)

0.65

38.5

%

Can

nab

isC

:C

ontr

ol19

7.43

(8.9

6)4

.94

(5.6

8)

2.4

9(7

.59)

0.36

26.3

%B

:B

rief

inte

rven

tion

165.

93(7

.53)

3.0

0(4

.36

)2

.93

(6.6

4)0.

426.

3%B

1:

1se

ssio

n6

10.7

6(1

0.3

4)4

.89

(5.8

6)

5.8

7(9

.86)

0.84

16.7

%B

2:

2se

ssio

ns

42.

76(2

.74)

1.1

8(1

.88

)1

.57

(1.6

8)0.

230.

0%B

3:

3–4

sess

ion

s6

3.22

(3.7

6)2

.33

(3.6

1)

0.9

0(4

.09)

0.13

0.0%

Ove

rall

356.

75(8

.25)

4.0

6(5

.14

)2

.69

(7.0

7)0.

3917

.1%

Tob

acco

C:

Con

trol

2718

.09

(7.7

7)16

.36

(8.7

8)

1.7

3(5

.62)

0.14

0.0%

B:

Bri

efin

terv

enti

on21

22.9

0(2

0.3

8)19

.43

(11.

21)

3.4

7(1

4.27

)0.

284.

8%B

1:

1se

ssio

n8

17.5

6(8

.28)

17.5

0(7

.07

)0

.06

(2.1

5)0.

000.

0%B

2:

2se

ssio

ns

831

.67

(30

.33)

24.1

3(1

1.67

)7

.54

(22.

26)

0.60

0.0%

B3

:3–

4se

ssio

ns

517

.40

(9.9

1)15

.00

(15.

07)

2.4

0(9

.24)

0.19

20.0

%O

vera

ll48

20.1

9(1

4.6

9)17

.70

(9.9

3)

2.4

9(1

0.24

)0.

202.

1%

Pol

ydru

gC

:C

ontr

ol28

5.00

(1.2

2)4

.32

(1.6

8)

0.6

8(1

.61)

0.46

0.0%

(ou

tof

11)

B:

Bri

efin

terv

enti

on24

4.38

(1.2

8)3

.54

(1.4

4)

0.8

3(1

.40)

0.56

0.0%

B1

:1

sess

ion

84.

88(0

.99)

3.3

8(1

.41

)1

.50

(1.4

1)1.

010.

0%B

2:

2se

ssio

ns

83.

56(0

.88)

3.6

7(1

.32

)2

0.1

1(1

.27)

20.

070.

0%B

3:

3–4

sess

ion

s7

4.86

(1.5

7)3

.57

(1.8

1)

1.2

9(0

.95)

0.86

0.0%

Ove

rall

524.

71(1

.27)

3.9

6(1

.61

)0

.75

(1.5

1)0.

500.

0%

aC

:Con

trol

—as

sess

men

ton

ly;B

:Bri

efin

terv

enti

on,w

ith

thre

esu

bgr

oup

s:B

1:1

sess

ion

,mot

ivat

ion

alin

terv

iew

(MI)

;B2

:2se

ssio

ns,

MI1

chan

gest

rate

gies

(CS

);B

3:3–

4se

ssio

ns,

MI1

CS

1m

ore

det

aile

dre

lap

sep

reve

nti

on.

bE

xclu

din

gn

on-u

sers

(at

pre

-tre

atm

ent)

and

thos

ew

ho

did

not

com

ple

tebo

thas

sess

men

toc

casi

ons.

cU

sin

gas

are

fere

nce

poi

nt

the

gran

dst

and

ard

dev

iati

onfo

rth

ere

leva

nt

OT

Id

rug

use

cate

gory

(i.e

.ac

ross

bot

has

sess

men

toc

casi

ons)

.

Brief interventions among regular amphetamine users 1285

Changes in amphetamine related harmsSimilar repeated-measures (treatment group 3time) ANOVAs were conducted for the am-phetamine-related harm variables. There were nodifferential changes in OTI crime scores acrossthe groups; however, there was a signi� cant re-duction in crime for the sample as a whole, froma mean score of 1.87 to 0.79 [F(1,48) 5 8.19,p , 0.01]. Similarly, OTI health scores also im-proved for the sample as a whole, from a meanscore of 20.15 to 16.04 [F(1,48) 5 8.43,p , 0.01]. While there were no differentialchanges across groups, the B2 (two-session) in-tervention group had signi� cantly better (i.e.lower) overall health scores than the controls[12.56 vs. 19.23, F(1,48) 5 7.47, p , 0.01] andthe B1 (one-session) intervention group [12.56vs. 21.00, F(1,48) 5 7.43, p , 0.01]. There wasno signi� cant change in levels of injecting risk-taking behaviour; however, the control group hadsigni� cantly higher injecting risk-taking scoresoverall compared to the intervention groups [9.02vs. 5.34, F(1,48) 5 9.24, p , 0.01], particularlywhen compared to the B2 (two-session) interven-tion group [9.02 vs. 3.89, F(1,48) 5 11.37,p , 0.01]. There were no signi� cant differencesbetween groups or changes over time in GHQ-28scores or OTI social functioning.

DiscussionThe results of the present study indicate thatfurther studies of brief cognitive-behavioural in-terventions among regular users of amphetamineare feasible and warranted. As documented previ-ously (Baker et al., 2001), the current samplecomprises a group of regular amphetamine userswith high levels of: dependence on amphetamine,injecting risk-taking behaviour, psychopathology,social dysfunction and crime. Although only17.2% of the initial sample were at the actionstage for reducing amphetamine use, 81.3% (52/64) were retained at 6-month follow-up. Morethan half (59.4%, 19/32) of all subjects assignedto intervention conditions attended all sessions.Thus, regular users of amphetamine, many ofwhom are ambivalent about change, can be re-cruited, treated and retained for follow-up evalu-ation.

As this was a small feasibility study and groupswere formed as a result of both intended assign-ment to the two- and four-session interventionand actual patterns of attendance, the results

regarding intervention effectiveness should beinterpreted with caution. Furthermore, use of a� yer to attract study participants and reimburse-ment for attendance at assessment interviews mayhave attracted more pre-contemplators comparedto a clinical setting. With these caveats, the� ndings are suggestive of a moderate overallintervention effect. The differential reduction inmean daily amphetamine use among the inter-vention versus control groups (1.02 vs. 0.44)represents a moderate difference in effect sizeunits (0.93 vs. 0.40), which is of a clinicallysigni� cant magnitude. The difference in absti-nence rates between the control and interventionconditions attained statistical signi� cance, lend-ing support to the comparative effectiveness ofthe intervention.

Brief cognitive-behavioural interventions maybe bene� cial for many amphetamine users withinexisting drug treatment services, includingMMT. In a recent study (Gossop et al., 2000),psychostimulant users were found to improvefollowing treatment in existing services, yet it wassuggested that the provision and effectiveness ofservices should be improved. Amphetamine userswith co-morbid psychological problems may needmore intensive interventions addressing bothmental health and drug issues. Future researchshould address the contribution of psychologicalfunctioning (e.g. as assessed by the GHQ) tointervention outcome.

The failure to detect a signi� cant differencebetween groups on the OTI amphetamine usescore, while detecting a signi� cant difference be-tween control and intervention groups in theproportion of subjects reporting abstinence,raises several methodological issues. First, wemay simply have too small a sample size toconsistently detect intervention effects. Secondly,the discrepancy in � ndings may call into questionthe sensitivity of the OTI in assessing am-phetamine as opposed to opioid use. Shearer et al.(1999) reported inconsistency between OTI am-phetamine scores and reductions in injecting fre-quency and the proportion of subjects reportingabstinence in their study of amphetamine substi-tution therapy. They point out that subjects whoactually reduce their mean daily level of am-phetamine use and who use less regularly in theform of a binge may receive in� ated OTI scoresas the OTI score estimates mean use during thelast binge and extrapolates this to the entiremonth. Shearer et al. (1999) suggest that a 30-day

1286 Amanda Baker, Tracy G. Boggs & Terry J. Lewin

diary of use could be completed by subjects. Athird methodological issue is that, within thecontext of this small study, random assignmentresulted in an uneven distribution of heavy ver-sus less heavy users of amphetamine. For exam-ple, the comparatively lower OTI amphetamineuse scores at pre-treatment among the B3 group(see Table 2) could have resulted in a “� oor”effect, limiting the amount of change that wasable to be demonstrated. Conversely, and associ-ated with their comparatively lower initial am-phetamine use scores, abstinence may have beenan easier outcome for the B3 group to achieve atfollow-up. By comparison, the B2 group hadhigher initial OTI amphetamine scores, weremore able to demonstrate sizeable improvement,and had the lowest abstinence rates at follow-up(see Table 2). The B2 group also had the lowestOTI polydrug use scores, which is consistentwith the � nding that they had better overallhealth (compared to C and B1) and injectingrisk-taking scores (compared to C). If this studyis replicated with a larger sample size, then amore even distribution of pre-treatment am-phetamine scores is likely and the capacity todemonstrate intervention effects would be moreequivalent. The general decrease in am-phetamine use for all groups is likely to partiallyre� ect regression to the mean effects, in thatsubjects were recruited at a time of high am-phetamine use. The non-speci� c effects of beingidenti� ed as a regular user of amphetamine andthe use of a self-help booklet may also havecontributed to improvement.

A signi� cant reduction in crime and asigni� cant improvement in health detectedamong the sample as a whole is consistent withthe overall reduction in amphetamine use amongthe sample. This attests to the community andpublic health bene� ts that accompany an overallreduction in amphetamine use among a cohortof regular amphetamine users.

The main conclusion from the present study isthat brief cognitive-behaviour therapy appears tobe feasible and moderately effective among reg-ular users of amphetamine. The results of thisstudy suggest that a larger randomized con-trolled trial of the effectiveness of cognitive-behavioural therapy should be conducted.

AcknowledgementsThis work was funded by a Research Manage-ment Committee (RMC) grant from the Univer-

sity of Newcastle. Paul Constable, Holly Devirand Rachel Garrett assisted with interviewing.We wish to thank the study participants andagencies from which participants were recruitedand those at which interviews were conducted.

ReferencesBAKER, A., BOGGS, T. G. & LEWIN, T. J. (2001) Char-

acteristics of regular amphetamine users and impli-cations for treatment, Drug and Alcohol Review, 20,49–56.

BIENER, L. & ABRAMS, D. B. (1991) The Contem-plation Ladder: validation of a measure of readinessto consider smoking cessation, Health Psychology, 10,360–365.

BIRD, K., HADZI-PAVLOVIC, D. & ISAAC, A. (2000)PSY2000 program (Sydney, School of Psychology,University of NSW).

DARKE, S., HALL, W., HEATHER, N., WARD, J. &WODAK, A. (1991) The reliability and validity of ascale to measure HIV risk-taking behaviour amongintravenous drug users, AIDS, 5, 181–185.

DARKE, S., HALL, W., HEATHER, N., WODAK, A. &WARD, J. (1992) Development and validation of amultidimensional instrument for assessing outcomeof treatment among opioid users: the Opiate Treat-ment Index, British Journal of Addiction, 87, 733–742.

GOLDBERG, G. & HILLIER V. F. (1979) A scaled versionof the General Health Questionnaire, PsychologicalMedicine, 9, 139–145.

GOSSOP, M., MARSDEN, J. & STEWART, D. (2000)Treatment outcomes of stimulant misusers: one yearfollow-up results from the National TreatmentOutcome Research Study (NTORS), AddictiveBehaviours, 25, 509–522.

GOSSOP, M., DARKE, S., GRIFFITHS, P., HANDO, J.,POWIS, B., HALL, W. & STRANG, J. (1995) The Sever-ity of Dependence Scale (SDS): psychometricproperties of the SDS in English and Australiansamples of heroin, cocaine and amphetamine users,Addiction, 90, 607–614.

HANDO, J., TOPP, L. & HALL, W. (1997) Am-phetamine-related harms and treatment preferencesof regular amphetamine users in Sydney, Australia,Drug and Alcohol Dependence, 46, 105–113.

ILLICIT DRUG REPORTING SYSTEM (1999) Drug TrendsBulletin, October (Canberra, Commonwealth De-partment of Health and Aged Care).

KAMIENIECKI, G., VINCENT, N., ALLSOP, S. &LINTZERIS, N. (1998) Models of Intervention and Carefor Psychostimulant Users, Monograph Series no. 32(Canberra, Commonwealth of Australia).

KLEE, H., WRIGHT, S. & MORRIS, J. (1999) Am-phetamine users in treatment: factors associated withsustained abstinence from street drugs, AddictionResearch, 7, 239–265.

MAKKAI, T. & MCALLISTER, I. (1998) Patterns of DrugUse in Australia 1985–95 (Canberra, AustralianGovernment Publishing Service).

MARLATT, G. A. & GORDON, J. R. (1985) Relapse

Brief interventions among regular amphetamine users 1287

Prevention: maintenance strategies in the treatment ofaddictive behaviours (New York, Guilford Press).

MILLER, W.R. & ROLLNICK, S. (1991) MotivationalInterviewing: preparing people to change addictivebehaviour (New York, Guilford Press).

NDARC (1997) A User’s Guide to Speed (Sydney:NDARC, University of New South Wales).

PROUDFOOT, H. & TEESSON, M. (2000) Investing inDrug and Alcohol Treatment, NDARC Technical Re-

port no. 91 (Sydney, National Drug and AlcoholResearch Centre).

SHEARER, J., WODAK, A., MATTICK, R. P., VAN BEEK, I.,LEWIS, J., HALL, W. & DOLAN, K. (1999) A Ran-domised Controlled Trial of Monitoring Controlled Pre-scribing of Dexamphetamine, NDARC TechnicalReport no. 75 (Sydney, National Drug and AlcoholResearch Centre).