Patterns and correlates of treatment: findings of the 2000 – 2001 NSWminimum dataset of clients of alcohol and other drug treatmentservices

JAN COPELAND1& DEVON INDIG

2

1Senior Lecturer, National Drug and Alcohol Research Centre, University of New South Wales, Sydney and 2Manager, Drug

Programs Bureau, NSW Health Department, Australia

AbstractThe aim of this study was to provide an overview of the first year of the NSW Minimum Dataset for Alcohol and Other DrugTreatment Services data collection, including describing the patterns and correlates of people having received treatment in NewSouth Wales. All closed treatment episodes for the 2000 – 2001 financial year were included for descriptive, univariate andmultivariate analyses. There were 33 459 closed episodes of care in New South Wales in the 2000/2001 financial year. Themajority of clients (69%) were male and the mean age was almost 34 years. The majority of treatment is sought for problemsrelated to alcohol (37%) and heroin (33%) use. More than a third (40%) of clients were new to drug and alcohol treatment. Halfthe clients had a history of injecting drug use with 6.3% of those with heroin as their principal drug of concern, never havinginjected. The most common main service provided was in-patient withdrawal (26%). Multivariate logistic regression revealedthat being older, not homeless, non-indigenous and having heroin as the principal drug of concern predicted receiving out-patientwithdrawal management. Analyses of length of stay in residential treatments and number of service contacts in non-residentialtreatments are reported. The NSW MDS AODTS is a critical information source for policy development, service planning andsurveillance. The results of this paper illustrate the utility of the data collection for identifying emerging issues in the patterns ofdrug use and service delivery for clients with alcohol and other drug problems. [Copeland J, Indig D. Patterns and correlates oftreatment: findings of the 2000 – 2001 NSW minimum dataset of clients of alcohol and other drug treatment services.Drug Alcohol Rev 2004;23:185 – 194]

Key words: alcohol, cannabis, heroin, injecting drug use, treatment.

Introduction

The global burden of disease attributable to alcohol

and illicit drug use has been estimated at 884 000 lives

in 1990 with a worst case prediction of 30 962 000

lives by 2020 [1]. With an increase over the last

decade in the proportion of Australians ever having

tried a range of illicit drugs, including the proportions

ever having injected an illicit drug [2], the need for

comprehensive data on the patterns of treatment being

received for alcohol and other drug problems is

growing. The only data source available in Australia

to date has been the series of National Clients of

Treatment Service Agencies Census’ held in 1990,

1992, 1995 and 2001, which showed a significant

increase in presentations for illicit drug problems over

that period [3].

1 July 2000 saw the commencement of collection of

the New South Wales Minimum Dataset for Alcohol

and Other Drug Treatment Services (NSW MDS

AODTS). A subset of this dataset contributes to the

National Minimum Dataset for Alcohol and Other

Drug Treatment Services. The first year of collection of

the national dataset is mainly registration-based and

incomplete (does not include Queensland) [4]. The

NSW MDS began as an episode-based collection and,

therefore, captures more data on the client’s treatment

experience than the national collection.

Given the costs associated with harmful drug use, the

resources required to address these harms, and the

Received 12 May 2003; accepted for publication 23 October 2003.

Jan Copeland PHD, Senior Lecturer, National Drug and Alcohol Research Centre, University of New South Wales, Sydney, NSW, Australia 2052;Devon Indig, Manager, Drug Programs Bureau, NSW Health Department, Australia. Correspondence to Jan Copeland PHD, Senior Lecturer,National Drug and Alcohol Research Centre, University of New South Wales, Sydney, NSW, Australia 2052. Tel: 61 2 9385 0333;Fax: 61 2 9385 0222; E-mail: J.Copeland@unsw.edu.au

Drug and Alcohol Review (June 2004), 23, 185 – 194

ISSN 0959-5236 print/ISSN 1465-3362 online/04/020185–10 # Australian Professional Society on Alcohol and Other Drugs

DOI: 10.1080/09595230410001704172

rapid development of the treatment sector, the NSW

Health Department aimed to develop a relevant and

timely data collection to inform policy development,

service utilization, resource allocation and strategies for

the sector. The NSW Minimum Dataset consists of 26

separate items to be collected at the beginning, during

and upon cessation of treatment. The dataset consists

of a broad range of items describing administrative,

social, demographic, drug-related and service-related

information. The dataset has been developed in

conjunction with treatment providers to ensure that

data items are useful not only on a national or state level

but also to individual agencies needing consistent and

accurately defined information with which to inform

service development and planning.

The aim of this paper is to provide an overview of the

NSW MDS AODTS and provide an overview of the

patterns and correlates of treatment being received in

New South Wales in the first year of the collection.

Method

Unit of measurement

The unit of measurement for the NSW MDS AODTS

is a ‘treatment episode’. A treatment episode is defined

as ‘a period of contact, with a defined date of

commencement and cessation between a patient/client

and a provider or team of providers that occurs in one

setting and in which there is no major change in either

the goal of intervention or the predominant treatment

activity.’ The ‘main service provided’ element of the

collection corresponds to the predominant treatment

activity of the treatment episode. The items collected

for the 2000 – 2001 data collection and the timing of

collection are set out in Table 1. These data are

reported as treatment episodes and differ slightly,

therefore, from the National Minimum Dataset for

AODS, which uses client registration as the unit of

measurement [4].

The data collection is an ongoing monthly data

collection, which is managed by NSW Health on a

financial year basis from 1 July to 30 June. All agencies

covered by the data collection submit a complete

dataset for every treatment episode completed within

the data submission period.

Coverage

Agencies included within the coverage of the NSW

MDS AODTS are all government treatment agencies

and non-government agencies in receipt of government

funding within NSW that provide services specifically

aimed at reducing drug-related harm for individuals.

Opioid maintenance pharmacotherapies, clients in

prison treatment programmes and Aboriginal Health

Services are not currently within the coverage of the

collection, but are proposed for inclusion in the future.

Data quality

Extensive data checking and validation procedures were

implemented to ensure the quality of the 2000 – 2001

data. These checks were implemented at an agency

level, an Area Health Service level and, finally, at a state

level. Any errors in the data were attempted to be

rectified at the original source of the information.

Data analysis

The analyses include, unless stated otherwise, primary

and secondary clients where data are collected for both

groups such as demographics and principal drug of

concern. The analyses were primarily descriptive in

nature and were performed using SPSS for Windows

(version 11.0). Means and medians for highly skewed

data are reported for continuous data. Categorical

variables are described in percentages. When compar-

isons were carried out, t-tests were used for compar-

isons between normally distributed continuous data, F-

Table 1. NSW minimum data set items

Commencementof treatment

Cessation oftreatment

Agency code Date of cessation oftreatment episode

Agency location Reason for cessationof treatment

Client code Referral to anotherservice

Date of birthSexIndigenous statusCountry of birthPreferred languagePrincipal source of incomeLiving arrangementUsual accommodationClient typePrincipal drug of concernOther drugs of concernMethod of use for principal drug ofconcernInjecting drug useService delivery settingDate of commencement of treatmentepisodeSource of referral to treatmentPrevious treatmentMain service providedOther services provided

During treatmentService contact dates

186 Jan Copeland & Devon Indig

tests for group differences and odds ratios (OR) with

corresponding 95% confidence intervals (CI) for

categorical data. Multivariate logistic regression was

used to identify independent associations between

demographic and drug use variables and the choice of

an in-patient withdrawal management, as this was the

most common main service provided.

Results

The 2000 – 2001 collection included 33 459 episodes of

care for approximately 27 394 registered clients. The

calculation of individual clients is based on the

Australian Institute of Health and Welfare’s national

counting rule [4]. This counting rule determines the

number of clients by identifying those that had the same

agency and client identification code, age, gender and

principal drug of concern. Using this rule it was

ascertained that 83% of clients received only one

episode of care, 12% received two episodes, and only

0.7% received five or more episodes (range 1 – 41).

This treatment was delivered by 181 agencies, 70% of

which were government agencies.

Demographic characteristics

The mean age of the sample was 33.6 (SD 11.8) years

with a range of 9 – 88 years, where those aged under 9

years were excluded. The majority of clients in 2000 –

2001 was male (69%), with less than 0.1% (n=14)

where gender was not stated or described inadequately.

Among the under-20 years group females represented

39.6% compared with 31% of the overall sample (w2

10.1; p5 0.001). Overall, 87.9% of the sample

identified themselves as not indigenous; 5.4% identi-

fied as Aboriginal but not Torres Strait Islander origin,

0.2% as Torres Strait Islander but not Aboriginal and

0.5% as Aboriginal and Torres Strait Islander with

6.4% not reporting their indigenous status.

There was a total of 126 countries of birth

represented with 1.5% not stated or described inade-

quately. The majority of the sample was born in

Australia (85.1%). The five next most common

countries of birth were England (2.6%), New Zealand

(2.1%), Vietnam (0.9%), Ireland (0.8%) and Scotland

(0.5%). All other countries accounted for 0.4% or less

of the clients in treatment.

The most common source of income was temporary

benefit (41.8%) and 22.9% pension, with 14% em-

ployed full-time and 5.9% part-time. Seven-and-a-half

per cent of the sample data for this variable was not

stated, not known or described inadequately. Most

commonly clients rented a public or privately owned

house or flat (45.3%), with 26.1% living in a privately

owned home, and 8.9% described themselves as

homeless/no usual residence. Clients lived most com-

monly with their parents (22%) or alone (21%), with

13.5% living with their spouse/partner, 8% with partner

and children and 4% alone with children.

Drug use data items

The majority (94.1%) of clients presented for their own

drug use, 3.3% for the drug use of another and the

remainder attending for their own drug use as well as

that of another person.

Principal and other drugs of concern. The most common

principal drug of concern was alcohol (37.3%) followed

by heroin (33.4%) and cannabis (10.4%). The drug

class with the lowest mean age on treatment entry was

cannabis (mean 28.1, SD 9.4 years) and the highest was

alcohol with a mean of 40.8 (SD 12.0) years. See Table

2 for a complete list. Among the three most commonly

injected drugs, heroin was most likely to be injected

with a rate of 86.7%. See Table 3 for more information.

Homeless clients presented most commonly with

alcohol as their principal drug of concern (14.4%),

followed by heroin (10.3%), amphetamines (8.8%) and

cannabis (4.1%). Alcohol was the most commonly

reported principal drug of concern for all accommoda-

tion types, with the exception of ‘other’ and ‘not stated’

for whom heroin was most common with 45.1%

(n=408) and 36.9% (n=1105), respectively.

Other drugs of concern. More than half (58.7%) of

clients had no other drug of concern. This data element

differs from polydrug use as a principal drug of

concern, as that may be nominated only where the

client’s problem arises from the practice of using a mix

of different drugs, and there is no single drug that can

be identified as constituting a major problem in its own

right. The three most common other drugs of concern

were cannabis (14.0%), alcohol (8.1%) and nicotine

(5.1%). The most commonly cited other drug of

concern for clients with alcohol as their principal drug

of concern was cannabis (14.3%) and nicotine (8.5%),

with 66.6% noting no other drug of concern. The most

commonly cited other drug of concern for clients with

heroin as their principal drug of concern was cannabis

(16.7%) and alcohol (7.8%), with 57% noting no other

drug of concern. The most commonly cited other drug

of concern for clients with cannabis as their principal

drug of concern was alcohol (24.6%) and ampheta-

mines (9.8%), with 51.1% noting no other drug of

concern.

Injecting drug use. More than a third of clients (36.3%)

had never injected a drug, with 38.7% being current

Alcohol and Drug Treatment in NSW 187

injectors and 13.7% having not stated or described

inadequately injecting drug use data. A full breakdown

is provided in Table 4. Of those with heroin as their

principal drug of concern 77.6% were current injectors,

with comparable rates for amphetamines 75.6%,

cocaine 66.9% and benzodiazepines 31.5%. Among

clients with heroin as their principal drug of concern

6.3% reported never having injected with comparable

rates for amphetamines of 8.3%, cocaine of 15.9% and

benzodiazepines of 30.6%. It is interesting to note the

different patterns in injection for benzodiazepines

between those presenting with that as their principal

drug of concern and those who report using benzodia-

zepines in general (31.5% versus 6.6%). Of these 6.6%

the majority (82.5%) nominated heroin as their

principal drug of concern. Of those presenting with

alcohol as their principal drug of concern, 11.8%

reporting injecting drug use. Similarly, 5% of those

presenting with cannabis as their principal drug of

concern reported injecting drug use at some time.

Service provision data items

Source of referral to treatment. The most common source

of referral to treatment was self (36.8%), followed by

non-residential alcohol and other drug treatment agency

(10.8%) and a general medical practitioner (7.5%).

Table 5 provides the complete breakdown of source of

referral to treatment. Self-referral remained the most

common source of referral across sex, indigenous and

English-speaking background comparisons.

Self-referral remains the most popular form of referral

across the three main drug classes, with cannabis having

the lowest rate of self-referral (32.7%), followed by

alcohol (34.2%) and heroin (41.9%). The highest rate of

referral from the criminal justice system was for heroin

(10.8%), followed by cannabis (10.6%) and ampheta-

mines (10.6%), with 8.3% of referrals for alcohol

problems being from this source. Aboriginal and/or

Torres Strait Islander client had higher rates of police or

court diversion at 0.9 versus 0.6%.

Table 2. Principal drug of concern

Principal drug of concern % (n) Mean age (SD) at episode commencement

Alcohol 37.3 (12448) 40.8 (12.0)Heroin 33.4 (11177) 28.8 (8.2)Cannabis 10.4 (3464) 28.1 (9.4)Amphetamines 8.0 (2666) 28.3 (8.3)Benzodiazepines 2.2 (744) 36.0 (11.1)Prescribed methadone 1.6 (542)Nicotine 1.3 (437)Opiates (organic derivatives) 1.2 (395) –Polydrug use 0.9 (310) –Cocaine 0.7 (235) 30.4 (7.4)Other amphetamine-related substances 0.3 (104) –Non-prescribed methadone 0.2 (78) –Hallucinogens 0.0 (13) –Anaesthetics 0.0 (12) –Volatile solvents 0.0 (12)Caffeine 0.0 (7) –Anabolic – androgenic steroids 0.0 (2)Volatile inhalants 0.0 (1*) –Other/inadequately described 2.5 (800) –

Table 3. Method of use of principal drug of concern

Heroin Amphetamines Cocaine Benzodiazepines Total all drugsMethod % (n) % (n) % (n) % (n) % (n)

Ingest 2.3 (257) 9.2 (246) 0.8 (2) 91.4 (680) 44.1 (14, 754)Smoke 9.0 (1002) 1.1 (29) 4.6 (11) 0.0 (0) 14.3 (4,778)Inject 86.7 (9696) 81.1 (2161) 73.6 (176) 6.6 (49) 37.3 (12, 479)Sniff (powder) 0.3 (29) 5.3 (142) 18.8 (45) 0.0 (0) 0.7 (221)Inhale (vapour) 0.0 (0) 0.1 (2) 0.0 (0) 0.0 (0) 0.1 (26)Not stated/inadequately described 1.7 (193) 3.2 (86) 2.1 (5) 2.0 (15) 3.65 (1201)

188 Jan Copeland & Devon Indig

The three most common sources of referrals for an

episode involving assessment-only as the main service

provided were self (36.4%), GP (8.3%) and non-

residential alcohol and other drug treatment agency

(6.7%). Comparable figures for a counselling episode

were self (31.9%), criminal justice (14.9%) and general

practitioner (GP) (9.6%). For episodes of out-patient

withdrawal, self-referral was most common at 45.1%

followed by GPs (18.3%) and non-residential alcohol

and other drug treatment agency (6.7%). Sources of

referral for in-patient withdrawal services were self

(44.6%), non-residential alcohol and other drug treat-

ment agency (6.1%) and residential alcohol and other

drug treatment agency (6.1%). The patterns of referrals

to residential rehabilitation services was self (29.2%),

followed by criminal justice (17.9%) and residential

alcohol and other drug treatment agency (20.2%).

Previous treatment. The three most common forms of

treatment self-reported as previously being accessed by

the clients were counselling (27.7%), in-patient/resi-

dential withdrawal management (18.5%) and out-

patient withdrawal management (3%). Table 6 pro-

vides a breakdown of the two most commonly

nominated previous treatment types accessed.

More than a third of clients (40.3%) had not

previously attended alcohol and other drug treatment.

These treatment neophytes were significantly (t=6.9,

df 33361, p5 0.001) older than those previously in

treatment (34.1 years versus 33.2 years) and signifi-

cantly (w2 21.9, p5 0.001) more likely to be male (70.4

versus 68.1%). While all other variables were tested,

none other reached statistical significance. More than

half (55.4%) of cannabis treatment clients had never

previously entered treatment compared with 44.3% of

alcohol and amphetamine treatment clients and 29.7%

of cocaine and 28.7% of heroin treatment clients.

Main service provided. The most common main service

provided was in-patient/residential withdrawal manage-

ment (25.9%), followed by assessment only (23%),

counselling (22.2%), residential rehabilitation activities

Table 4. Injecting drug use by principal drug of concern

Injecting drug use % (n) Most common principal drug of concern

Last injected within the previous 3 months (current) 38.7 (12938) Heroin (67.0%)Last injected 4 3 months but 5 12 months ago (recent) 4.0 (1333) Heroin (41.4%)Last injected 4 12 months ago (injecting history) 7.4 (2460) Alcohol (36.3%)Never injected 36.3 (12151) Alcohol (68.5%)Not stated/inadequately described 13.7 (4577)

Table 5. Source of referral to treatment

Referral source % (n)

Self 36.8 (12302)Alcohol and other drug treatment agency—non-residential 10.8 (3599)General practitioner 7.5 (2494)Family member/friend 6.7 (2226)Alcohol and other drug treatment agency—residential 6.6 (2212)Medical officer/specialist 5.1 (1711)Other hospital 4.2 (1422)Other correctional/criminal justice setting 4.5 (1518)Court diversion 4.2 (1420)Other community service agency 3.1 (1047)Other 2.2 (730)Community health centre—non-residential 2.1 (717)Community mental health centre—non-residential 1.2 (408)Family and child protection service 1.1 (353)Psychiatric hospital 1.0 (333)Education institution 0.5 (151)Police diversion 0.3 (112)Workplace 0.3 (111)Not stated/inadequately described 1.8 (590)

Alcohol and Drug Treatment in NSW 189

(9.1%) and out-patient withdrawal management

(8.3%). See Table 7 for a full description.

Multivariate logistic regression using backwards

stepwise removal revealed that being older, not home-

less, having heroin as the principal drug of concern and

being non-indigenous predicted receiving an out-

patient withdrawal service rather than a residential

withdrawal service, with only gender dropping out of

the model.

The most common main service provided for

clients with alcohol as their principal drug of concern

was in-patient/residential withdrawal management

(28.5%), for heroin was in-patient/residential with-

drawal management (31.4%), for cannabis was

counselling (39%), for amphetamines was assessment

only (25.5%) and for cocaine was in-patient/residen-

tial withdrawal management (30.5%). Comparisons

are provided in Fig. 1.

Figure 2 provides a comparison of main service

provided by age categories. Clients under 20 years were

significantly more likely to have received assessment

only as their main service provided (OR 1.35, 95% CI

1.2 – 1.52) than did clients aged more than 50 years

(24.2% versus 19.1%). This age group was also

significantly more likely to have received residential

rehabilitation as their main service provided (OR 5.25,

95% CI 4.3 – 6.4) than did clients aged more than 50

years (15.9% versus 3.5%). Conversely, clients aged

more than 50 years were significantly more likely to

have received in-patient withdrawal management as

their main service provided (OR 2.7, 95% CI 2.4 – 3.1)

than did clients aged less than 20 years (32.9% versus

15.9%).

Of those clients in counselling treatment 30.6% had

self-reported that they had previously received counsel-

ling and 48.9% had received no previous treatment.

Almost half (49.1%) of clients in out-patient with-

drawal services had received no previous treatment and

15.5% had received a previous in-patient withdrawal.

Among clients in in-patient withdrawal services less

than a third (31.3%) had received no previous

treatment and 29.1% had experienced a previous in-

patient withdrawal. Only around one-quarter (25.9%)

of residential rehabilitation clients had reported that

this was their first treatment, with 25.1% having had a

previous in-patient withdrawal but only 7.5% having

previously entered residential rehabilitation.

Other services provided. Around half (49.4%) of clients

received no other service in addition to their main

Table 6. Previous treatment

Treatment type 1st % (n) 2nd % (n)

No previous treatment 40.3 (13354) -Counselling 27.7 (9175) 1.7 (196)In-patient/residential withdrawal 18.5 (6119) 43.7 (4973)Out-patient withdrawal management 3.0 (999) 9.9 (1122)Residential rehabilitation activities 2.4 (781) 17.6 (2007)Methadone 2.4 (779) 7.2 (817)Assessment only 2.1 (706) 3.2 (362)Other 1.1 (359) 1.1 (124)Information and education 0.9 (309) 5.6 (641)Out-patient consultation 0.4 (136) 0.9 (105)Naltrexone maintenance 0.4 (121) 3.1 (348)Day programme rehabilitation services 0.4 (117) 1.0 (119)In-patient consultation 0.3 (85) 0.7 (79)Acamprosate 0.2 (50) 0.6 (70)Other maintenance pharmacotherapies 0.1 (34) 0.5 (52)Buprenorphine maintenance 0.0 (6) 0.1 (10)Disulfiram 0.0 (3) 0.1 (10)Buprenorphine 0.0 (6) 0.1 (10)LAAM maintenance 0.0 (0) 0.2 (22)

Table 7. Main service provided

Main service provided % (n)

In-patient/residential withdrawal management 25.9 (8667)Assessment only 23.0 (7711)Counselling 22.2 (7417)Residential rehabilitation activities 9.1 (3037)Out-patient withdrawal management 8.3 (2774)Information and education 4.3 (1142)Other 4.1 (1152)Consultation 2.3 (760)Day programme rehabilitation activities 0.8 (280)

190 Jan Copeland & Devon Indig

service. The most common other services provided

were information and education (23.9%) and counsel-

ling (17.3%).

Service contacts. Among non-residential services the

mean number of service contacts per treatment episode

was 2.1 (SD 3.0, range 1 – 92). More than two-thirds

(67.8%) of clients had only one service contact per

episode. The mean number of service contacts for

clients with alcohol as their principal drug of concern

receiving a non-residential main service was 2.1 (SD

5.42, range 1 – 64), for heroin clients was a mean of 1.9

(SD 2.7, range 1 – 46), for cannabis clients was a mean

of 2.5 (SD 3.0, range 1 – 34), for amphetamine clients

was a mean of 2.0 (SD 2.6, range 1 – 42) and for

cocaine was a mean of 1.9 (SD 2.7, range 1 – 25).

Counselling clients received a mean of 3.5 (SD 3.9,

range 1 – 48) in-person service contacts compared with

a mean of 3.4 service contacts (SD 5.3, range 1 – 92) for

out-patient withdrawal management. Clients in day

programmes received a mean of 4.9 (SD 4.8, range 1 –

38) service contacts and clients receiving out-patient

consultations not including withdrawal management

received a mean of 2.1 (SD 2.5, range 1 – 26) service

contacts per episode.

Women received significantly more service contacts,

across non-residential treatment types, than did men (t

7 9.7, df 16171.1, p5 0.001), with a mean of 2.4 (SD

3.5) compared with 1.9 (SD 2.7) service contacts per

Figure 1. Five most common main services provided by major drug classes.

Figure 2. Main service provided by age categories.

Alcohol and Drug Treatment in NSW 191

treatment episode. There were no significant differ-

ences in the number of service contacts received by age

or indigenous status.

Length of stay. The mean length of stay for clients

receiving a residential main service was 25.8 days (SD

46.6, range 0 – 362 days). Clients of in-patient with-

drawal services stayed a mean of 5.6 days (SD 12.4,

range 0 – 223, mode of 1 day) and clients of residential

rehabilitation services a mean of 32.5 days (SD 40.3,

range 0 – 266).

Among clients of residential withdrawal services,

those with alcohol as their principal drug of concern

spent significantly more days (t7 7.2, df 6925.4,

p5 0.001) in treatment than did clients with heroin

as their principal drug of concern (mean of 6.2, SD

12.8 versus 4.2, SD 11.1). Clients with cannabis as

their principal drug of concern had significantly more

days in withdrawal management treatment (t7 5.5, df

509.9, p5 0.001) than did clients seeking heroin

withdrawal (mean of 7.3, SD 10.7 days). The homeless

spent significantly less days in residential withdrawal

services (t7 7.2, df 6299.7, p5 0.001) than those with

other accommodation, (mean 4.3, SD 6.6 versus mean

of 5.9, SD 13.6 days). There were no age, gender or

indigenous status differences in the length of time in

residential withdrawal services.

Among clients in residential rehabilitation services,

those aged over 50 years had significantly longer

treatment episodes (t7 4.2, df 158.8, p5 0.001) than

those aged under 20 years (mean 39.6, SD 47.5 versus

a mean of 20.6, SD 31.7 days). Clients with heroin as

their principal drug of concern stayed significantly

longer (t 3.1, df 678.6, p5 0.002) than those with

cannabis (mean 32.2, SD 43.3 versus a mean of 25.5,

SD 29.9 days), while there were no significant

differences in the length of stay between clients with

alcohol and heroin as their principal drug of concern.

There were no gender, indigenous status or home-

lessness status differences in the length of treatment

episodes in residential rehabilitation services.

Reason for cessation of treatment episode. The most

common reason for cessation of treatment was that

the episode was completed (44.7%). The combined

rate of adverse events such as left without notice,

against advice or discharge for non-compliance was

27.5%. Table 8 provides a breakdown of the reason for

treatment cessation for the entire sample.

Referral to another service. More than half the clients

(52%) received no referral to another service. A referral

in this data collection is defined as a formal referral

process that results in a letter or telephone call to the

agency that the client is being referred for the

continuation of their treatment needs. The most

common referral was to a residential alcohol and other

drug treatment agency (16.2%) and a non-residential

alcohol and other drug treatment agency (13.7%).

Table 9 provides the breakdown of referral to other

service for the entire sample. The homeless were 1.6

times more likely to receive a referral to another service

than were those in other living situations (OR 1.6, 95%

CI 1.3 – 2.0). Those aged more than 50 years were

almost one-and-a-half times more likely to receive a

Table 8. Reason for cessation of treatment episode

Reason % (n)

Treatment completed 44.7 (14942)Left without notice (LWN) 16.3 (5460)Transferred/referred to another service 15.1 (5063)Left against advice (LAA) 10.3 (3404)Involuntary discharge (non-compliance) 5.0 (1672)Other 3.9 (1310)Not stated/inadequately described 2.8 (930)Moved out of area 1.0 (330)Imprisoned, other than through courtsanction

0.4 (146)

Sanctioned by drug court/court diversionprogramme

0.2 (66)

Ceased treatment upon expiation 0.2 (61)Released from prison 0.1 (37)Died 0.1 (23)

Table 9. Referral to another service

Referral service % (n)

No referral 52.0 (17400)Alcohol and other drug treatment agency—residential

16.2 (5408)

Alcohol and other drug treatment agency—non-residential

13.7 (4567)

Other 5.6 (1889)General practitioner 3.6 (1203)Other community service agency 1.8 (586)Community health centre—non-residential 1.7 (585)Medical officer/specialist 1.2 (397)Community mental health centre—non-residential

1.0 (348)

Other hospital 0.9 (301)Other correctional/criminal justice setting 0.7 (220)Court diversion 0.6 (209)Psychiatric hospital 0.4 (125)Family and child protection service 0.3 (96)Workplace 0.2 (59)Education institution 0.1 (34)Police diversion 0.1 (17)

192 Jan Copeland & Devon Indig

referral to another service than those aged less than 20

years (OR 1.45, 95% CI 1.3 – 1.6). Clients with a

principal drug of concern other than cannabis were 3.5

times more likely to receive a referral to another service

than were cannabis clients (OR 3.5, 95% CI 2.8 – 4.4).

There were no differences in the referral rates by gender

or indigenous status.

Discussion

The NSW Minimum Dataset for Alcohol and Other

Drug Treatment Services was developed to enable

better monitoring of the characteristics of those

individuals receiving treatment and of the treatments

they were provided. These data assist treatment

providers and funding bodies to plan service models

and resource allocation with greater precision and to

monitor client demographics, drug use and treatment-

related trends over time.

The results of the first year of the collection

illustrate the characteristics and treatment experiences

of people attending drug and alcohol agencies in

NSW. Over two-thirds (69%) of clients were male,

with women being significantly over-represented in the

under-20 age group. The mean age of participants was

almost 34 years, with alcohol clients being significantly

older at a mean age of 41 years at the commencement

of the treatment episode. Just over 6% of the clients

identified as Aboriginal or Torres Strait Islander. This

represents a much higher burden of drug and alcohol

problems in this community that only makes up 2.1%

of the NSW population as at the 2001 Census [5].

The majority of clients received some form of

temporary benefit or pension (65%), illustrating the

social disadvantage this population suffers and the cost

to the community of their remaining outside of

treatment.

The majority of treatment is sought for problems

related to alcohol and heroin use. As this dataset also

does not include opioid maintenance pharmacothera-

pies the trend of increasing presentations for heroin

treatment are consistent with the findings of the 1-day

censuses conducted in the past [3]. This reflects

changes in patterns of drug use and the need for

treatment services to respond to these changes in client

demographics and intervention needs.

This dataset also aids in the monitoring of public

health issues related to individuals presenting for

treatment. For example, among those seeking treat-

ment for heroin-related problems 9.0% nominate

smoking as their method of use. This would suggest

that non-injecting routes of administration are still

placing individuals at risk of developing dependence.

The frequency of current injecting amongst the

cohort (38.7%), with half having ever injected,

highlights the need for awareness of blood-borne

virus risk-taking behaviour among clients of alcohol

and other drug treatment service even when present-

ing with alcohol or cannabis as their principal drug of

concern.

More than a third (40.3%) of clients were new to

treatment. Among this important group those present-

ing for cannabis, alcohol and amphetamine treatment

were over-represented. Those new to treatment were

significantly more like to be male, older and be of

Aboriginal and/or Torres Strait Islander descent. This

suggests that the emerging concerns of amphetamine

type substances and cannabis are being reflected in

treatment populations.

Age interacts with the principal drug such that a

greater proportion of young people were seeking

treatment for heroin while older people were repre-

sented more highly among those seeking treatment for

alcohol. Younger people were also more likely to

receive residential rehabilitation, while older people

were more likely to receive counselling as their main

service provided, regardless of the principal drug of

concern.

The most common main service provided was in-

patient/residential withdrawal. Multivariate analysis

revealed that older, more socially stable, non-indigen-

ous clients seeking heroin withdrawal were more likely

to be provided an out-patient than a residential with-

drawal. These demographics are consistent with those

recommended for out-patient withdrawal management;

however, this model of withdrawal has been found to be

the least successful for those seeking heroin withdrawal

[6]. This finding is worthy of further research attention.

The vast majority of clients sought treatment for their

own drug use and refer themselves into treatment. In

residential treatment services the average stay for

withdrawal services was 5 days. The longest average

withdrawal was for cannabis at 7.3 days. The average

length of stay for rehabilitation services was 32 days.

Given that many residential services offer programmes

of 6 – 12 months or longer, this suggests the need to

structure programmes to take into account this average

length of stay rather than the optimal length of stay for

that treatment model. It must be remembered, how-

ever, that length of stay may be constrained artificially

by local policy and funding decisions that affect the

programme length.

The average number of service contacts for out-

patient clients was two; however, more than two-thirds

of clients received only one service contact per episode.

Once again, cannabis clients received the greatest

number of service contacts per episode followed by

alcohol, amphetamines, heroin and cocaine. This

suggests that the average episode of out-patient care is

extremely brief and that intervention development

should take this into account. It is also interesting to

note that once again clients attending treatment for

Alcohol and Drug Treatment in NSW 193

cannabis-related problems are requiring the most

intensive intervention or are more willing to be engaged

into treatment.

The treatment was successfully completed in almost

half of all treatment episodes, which is the main

measure of treatment effectiveness in the collection. It

is concerning that almost a third had left treatment

under adverse circumstances, such as being unwilling,

or unable, to participate meaningfully. In addition, less

than half of clients received a referral following

treatment cessation, particularly young people and

those with cannabis as their principal drug of concern.

This suggests that more research is needed into

facilitation of treatment entry, engagement and con-

tinuing care for various client groups.

The findings of the NSW MDS AODTS have had,

and will continue to have, a significant impact on policy

development, service planning and surveillance at a

state-wide level. The discovery that more than 10% of

all treatment episodes are for clients concerned with

their cannabis use has led the NSW Health Department

to develop a cannabis policy. These data have also

informed the development of a psychostimulant policy

and the development of the Aboriginal Alcohol and

Drug Substance Misuse Plan. In addition to policy

development, the availability of agency and Area Health

Service data has improved the service planning process

at both a state and local level. The data have also been

utilized in surveillance reports such as the Illicit Drug

Reporting System [7] and for NSW Health’s monitor-

ing of trends in heroin overdose.

As the 2000 – 2001 NSW MDS AODTS was in its

first year of collection, it has certain limitations with

regard to data quality. Extensive procedures have been

put into place to develop data dictionaries and guide-

lines for the collection, to train service providers, to

provide an electronic data collection platform and to

provide quarterly reports back to the agencies about the

quality of their data. The quality of the data collection is

sure to be even greater in future years. These first year

findings, however, have already highlighted important

service delivery issues and the need for further research

on a range of treatment related factors.

References

[1] Murray CJL, Lopez AD. The global burden of disease. A

comprehensive assessment of mortality and disability from

diseases, injuries, and risk factors in 1990 projected to 2020.

Boston: The Harvard School of Public Health, 1996.

[2] Miller M, Draper G. Statistics on drug use in Australia 2001.

Drug Statistics Series no. 8, 2001. Canberra: Australian

Institute of Health and Welfare, 2001.

[3] Shand F, Mattick RP. Clients of treatment service agencies:

May 2001 census findings. National Drug Strategy Mono-

graph Series No. 47. Canberra: National Drug Strategy.

[4] Australian Institute of Health and Welfare. Alcohol and other

drug treatment services in Australia 2000 – 2001. First report

on the National Minimum Dataset. Canberra: Australian

Institute of Health and Welfare, 2002.

[5] Australian Bureau of Statistics. Australian demographic

statistics 3101.0, 30/3/03.

[6] Mattick RP, Hall W. Are detoxification programmes

effective? Lancet 1996;347:97 – 100.

[7] Roxburgh A, Degenhardt, Breen C, Barker B. New South

Wales drug trends 2002: findings from the Illicit Drug

Reporting System (IDRS). NDARC Technical Report no.

14. Sydney: NDARC, 2003.

194 Jan Copeland & Devon Indig