Online Cognitive Behavioural Therapy (CBT) for Drug or Alcohol Dependence
Prepared by Catriona Matheson for CERGA, April 2010.
Research and evaluation of online CBT is growing rapidly at present.
Findings to date indicate online CBT can be effective at reducing both drug and
There is too much variability between models studied to identify key significant
factors at this stage.
There is insufficient quality data regarding specific target group e.g. rural or
Online treatment services generally are accessed by a higher proportion of young,
employed, white and female users.
The use of the internet to deliver easily accessible, cost effective treatment and
support for substance misusers i.e. drugs, alcohol and nicotine has been recognized
for some time. This phenomenon is not new. The use of online psychotherapies
generally has been widely applied in mental health and previous research has been
conducted in panic disorder, obesity, post-traumatic stress disorder, depression and
anxiety (Copeland & Martin, 2004). Of all the psychotherapies cognitive behavioural
therapy (CBT) best lends itself to an interactive web-based format because it does
not necessarily require a therapist. The remit of this brief topic review was to assess
the literature on the use of online CBT specifically, in substance misuse broadly.
At an early stage of reviewing the literature it became clear there was a considerable
volume of research on web-based smoking cessation. To identify whether CBT in
particular was used required sifting through a large number of abstracts which was
time consuming. Therefore it was decided to limited the review to online CBT for
illicit drug and alcohol use.
The Medline database was searched from 1996 to March 2010 using the following
keywords: Substance abuse, heroin use, alcohol (and nicotine) combined with
cognitive therapy, behavior therapy, cognitive behavioural therapy and combined
with: internet, online systems and computer communication networks. In addition a
search of the Drug and Alcohol findings website was conducted which revealed one
The Medline search identified 25 papers relating to alcohol and 16 on general
substance misuse. Scanning of the alcohol titles excluded many papers that were
not relevant e.g. due to covering the workforce, opinion based papers, letters, or
were not using CBT. Five evaluative studies comparing effectiveness were identified,
one study covered service user characteristics and one study covered service users
as well effectiveness.
3.1 The Characteristics of Online Service Users
A novel US study attempted to characterize online service users of internet recovery
services (IRS) for drug and alcohol use (Hall et Tidwell, 2003). Data was collected
using a survey link sent to all known online recovery service providers. There were
significantly more females than expected and regarding ethnicity, significantly more
respondents were characterized as ‘white’. Regarding age, this was normally
distributed with a range of 16-80 years. There was an age sex correlation with more
females in younger age groups and more males in older age groups. Respondents
came from 20 different countries.
An Australian study also characterized (Swann & Tyssen, 2009) service users of an
online counseling service called ‘CounsellingOnline’. A client survey was condusted
(n=277). Data was compared to a state-wide conventional counselling dataset
(Alcohol and Drug Information System; n=19,283) and a statewide telephone
counselling database (DirectLine; n=59,863). CounsellingOnline clients differed from
conventional and telephone counselling clients, representing much higher proportions
of young, employed and female clients. Notably, the majority of CounsellingOnline
sessions occurred outside standard business hours.
3.2 The Effectiveness of Online Services for Drugs and Alcohol
Five RCTs of online therapy based on CBT were identified (Carroll, et al 2008;; Alemi
et al, 2010; Riper et al 2007; Finfgeld-Connett & Madsen, 2008; Leiberman 2006)
and a protocol for a further ongoing study was found (Blankers et al, 2009). Two of
these RCTs covered drug misuse and two covered alcohol misuse. One cohort study
and one cross section survey was identified.
Randomised Controlled Trials
Carroll et al (2008) evaluated the efficacy of a computer-based version of cognitive-
behavioral therapy (CBT) for substance dependence. 77 individuals seeking
treatment for substance dependence at an outpatient community setting were
randomly assigned to standard treatment or standard treatment with biweekly access
to computer-based training in CBT (CBT4CBT) skills (intervention). Participants
assigned to the intervention demonstrated significantly less drug use demonstrated
by negative urine samples for any type of drugs and tended to have longer
continuous periods of abstinence during treatment. The CBT4CBT program received
positive feedback from participants. In the intervention group outcome was more
strongly associated with treatment engagement than in treatment as usual;
furthermore, completion of ‘homework’ in CBT4CBT was significantly correlated with
outcome and a significant predictor of treatment involvement. This study indicates
CBT4CBT is an effective adjunct to standard outpatient treatment for substance
dependence. A key feature of this study was that online therapy was in addition to
Alemi et al (2010) examined the effect of online counseling on drug use among what
they describe as ‘underserved’ patients i.e. a hard to access, socially disadvantaged
group (low income, undereducated, unemployed, court involved, or diagnosed with
co-occurring psychiatric disorders). Participants were recruited from an Indian
Reservation in South Dakota; a family court in New Jersey; a probation office in
Virginia; and a co-occurring disorders treatment clinic in Washington, DC. There
were 79 volunteering participants that were randomly assigned to either a control
group or intervention group. Both groups were given an Internet-ready computer
and 1 year of Internet service. The intervention group had access to online
counseling intervention. Drug use was measured using a combination of self-usage
reporting and supervised urine tests. Self-usage reports or urine test results were
available from 70% of subjects. The difference in the rates of drug use between
control and intervention groups was not significantly suggesting that online
counseling had not led to a reduction in substance use. However this study may be
under powered. The key features of this study were the targeting of hard to reach
groups and the lack of any other therapy.
Riper et al (2007) conducted a trial in the Netherlands in 2003-2004. The
intervention and trial were conducted online. 261 adult problem drinkers were
selected from the general population Participants were randomized to either the
experimental drinking less (DL) condition (intervention) or to the control condition
which only received access to an online psychoeducational brochure on alcohol use
(PBA). The drink less intervention was a web-based, interactive self-help intervention
based on cognitive-behavioural and self-control principles without therapist guidance
which lasted 6 weeks. Outcomes measured at 6-month follow-up were (i) the
percentage of participants who had reduced their drinking levels to within the
normative limits of the Dutch guideline for low-risk drinking; and (ii) the reduction in
mean weekly alcohol consumption. At follow-up, 17.2% of the intervention group
participants had reduced their drinking successfully to within the guideline norms; in
the control group this was 5.4% [odds ratio = 3.66; (CI) 1.3-10.8; P = 0.006;
number needed to treat (NNT) = 8.5]. Intervention participants decreased their
mean weekly alcohol consumption significantly more than control participants, with a
difference of 12.0 standardized units (95% CI 5.9-18.1; P < 0.001; standardized
mean difference 0.40). This intervention appeared to be highly effective at reducing
problem drinking in the community. A key feature of this was the lack of therapist
Finfgeld-Connett & Madsen (2008) conducted an RCT to evaluate a web-based, self-
guided alcohol treatment program. Eligible women were randomized to standard care
or an online treatment program. Web-based treatment components included gender-
specific reference modules and decision making modules, an asynchronous bulletin
board, and a synchronous chat feature. It was not clear whether or not CBT was an
underlying component (only the abstract was available for this paper). At 3 months
follow up both groups had decreased their drinking but there were no significant
differences. Again the relatively small numbers may have resulted in the trial being
underpowered. The key feature of this trial was that it targeted women. The same
researchers are undertaking a further study with women but full results are not yet
available (Finfgeld-Connett D, 2009).
Since the quality of the therapeutic alliance is believed to effect the outcome of
psychotherapy and psychopharmacologic interventions, Leiberman (2006) aimed to
develop a personified guide designed to stimulate reactions similar to those
experienced in a therapeutic relationship, and evaluate the effect of the guide on
adherence to and satisfaction with an online alcohol use program. Standard
questionnaires used to evaluate problem drinking, 288 subjects were randomized to
receive feedback on their results in text form or via a multimedia condition involving
a personified guide. Those who received feedback via the guide demonstrated
greater levels of program adherence by completing more modules. Reported
satisfactions levels did not differ between the two conditions. It appears that
encouraging the personification of an automated alcohol intervention may lead to
greater levels of engagement. The key feature of this study was the attempt to
simulate a therapeutic relationship.
Swann & Tyssen explored whether ‘Counselling Online’ enhanced alcohol and drug
service accessibility for clients. This model used a ‘online chat’ based system. It was
not explicitly stated whether CBT was one on the underlying models. They examined
service utilisation data (n=2004) and client survey data (n=277) to evaluate service
responses, service impact, client characteristics and client satisfaction (reported
earlier). Positive service responses, such as fast counsellor connection speed, low
client attrition and anonymous login, were reported. Authors conclude
CounsellingOnline has showed enhanced AOD service accessibility through its service
responsiveness, high level of after-hours service utilisation, and appeal to a client
group that differs from those seen in conventional and telephone AOD counselling
One UK based cohort was identified. This pilot study assessed the usefulness of a
web based intervention called down your drink (DYD). The DYD website was a 6-
week manual based programme, based on motivational approaches and cognitive
behavioural therapy. Site visitors whose responses to the Fast Alcohol Screening Test
were positive, and those indicating excessive alcohol consumption, were encouraged
to register. Those registered were asked to complete alcohol dependence and mental
health questionnaires before the programme, and a drinking diary at each of the
weekly sessions. Questionnaires were sent electronically to those who completed the
programme, or who missed >three of the six sessions. Over a 6-month period there
were 7581 visits to the site indicating a large potential interest. Of these 1319
registered and of those, 61.8% completed week 1, but only 6.0% completed the
programme. Whilst this 6% provided encouraging feedback the drop out rate is very
high. From the minimal information collected from those that dropped out it appears
the programme was simply too time-consuming. A key feature of this study was that
it was UK based.
3.3 Miscellaneous Related Studies of Interest
Since one of the advantages on online CBT for treatment is the lack of trained
therapists/counselors an interesting study tested the use of online technology as a
cost effective method of training a geographically diverse group of therapists.
Weingardt et al, 2009 conducted an RCT comparing training outcomes obtained by
147 substance abuse counselors who completed eight self-paced online modules on
cognitive-behavioral therapy (CBT) and attended a series of four weekly group
supervision sessions using Web conferencing software. Participants were randomly
assigned to two conditions (low fidelity or high fidelity) that varied in the degree to
which they promoted adherence to the CBT protocol. Counselors in both conditions
demonstrated similar improvements in CBT knowledge and self-efficacy at follow up.
Those in the low-fidelity condition demonstrated greater improvement on one of
three measures of job-related burnout when compared to the high-fidelity condition.
The authors concluded it was feasible to implement a technology-based training
intervention with a geographically diverse sample of practitioners, that two training
conditions applied to these samples of real-world counselors do not produce
statistically or clinically significant differences in knowledge or self-efficacy.
4 Reflection on Methods
Although several RCTs were found a couple of these may have been underpowered
as numbers were small and it was not clear whether sample size calculations were
conducted. Studies had been conducted in a range of countries which increases
The models of online provision varied considerable, some used a mixture of
approaches of which CBT was only one. The information given on the details of each
programme was variable.
The volume of good quality research to evaluate online CBT is growing. Based on
the literature available there is some evidence that online CBT can reduce drug and
alcohol use. Furthermore online systems can be an effective tool for delivering CBT
training for therapists/counsellors. The strongest evidence came from the findings of
Carroll et al (2008). Their study varied from others in that online CBT was in
addition to face to face standard, weekly counseling. However the only other study
that had significant findings (Riper et al, 2007) was entirely web-based. Thus no
conclusion can be drawn about the relative merits of an entirely web-based model of
delivery versus online methods being used as an adjunct to standard treatment.
The characteristics of the programme are likely to be important but at present there
has been too much variability between the interventions being tested, combined with
an insufficient volume of research, to identify what characteristics are important.
The ongoing study by Blanker et al (2009) should help tease out some of these
Although groups under-represented in standard treatment have been targeted these
studies did not demonstrate significant findings so no conclusion can be drawn.
If a program were developed that was to be entirely web-based it should account for
the potential user profile in which there is a higher representation of young,
Alemi F. Hack M. Nemes S. Harge A. Baghi H. Impact of online counseling on drug use: a pilot
study. Quality Management in Health Care. 19(1):62-9, 2010 Jan-Mar.
Blanker M, Koeter M., Schippers G. (2009) Evaluating real-time internet therapy and online
self-help for problematic alcohol consumers: a three arm RCT protocol. BMC public health
open access vol 9.
Carroll, K, Ball, SA , Martino, S. et al (2008). Computer-Assisted Delivery of Cognitive-
Behavioral Therapy for Addiction: A Randomized Trial of CBT4CBT Am J Psychiatry 2008;
Copeland J., Martin Greg. (2004) Web-based interventions for substance use disorders: A
qualitative review. J Substance abuse treatment; 26:109-116.
Finfgeld-Connett D. Madsen R. Web-based treatment of alcohol problems among rural women.
Journal of Psychosocial Nursing & Mental Health Services. 46(9):46-53, 2008 Sep.
Finfgeld-Connett D. Web-based treatment for rural women with alcohol problems: preliminary
findings. Computers, Informatics, Nursing. 27(6):345-53, 2009 Nov-Dec.
Hall M., Tidwell W (2003) Internet recovery for substance abuse and alcoholism: an
exploratory study of service users. J Substance Abuse Treatment 24 (2003)161-167.
Lieberman, D Z. Effects of a personified guide on adherence to an online program for alcohol
abusers. Cyberpsychology & Behavior. 9(5):603-7, 2006 Oct.
Riper H. Kramer J. Smit F. Conijn B. Schippers G. Cuijpers P. Web-based self-help for problem
drinkers: a pragmatic randomized trial. Addiction. 103(2):218-27, 2008 Feb.
Swan AJ. Tyssen EG. Enhancing treatment access: evaluation of an Australian Web-based
alcohol and drug counselling initiative. Drug & Alcohol Review. 28(1):48-53, 2009 Jan.
Weingardt KR. Cucciare MA. Bellotti C. Lai WP. A randomized trial comparing two models of
web-based training in cognitive-behavioral therapy for substance abuse counselors. Journal of
Substance Abuse Treatment. 37(3):219-27, 2009 Oct.