This is Archived Content

This content is available for historical purposes only. It may not reflect the current state of science or language from the National Institute on Drug Abuse (NIDA). To view the latest NIDA Notes visit

Cite this article

NIDA. (2015, November 19). Digital Addiction Therapies Affirm Promise in Replication and Large Trial. Retrieved from

press ctrl+c to copy
November 19, 2015
Person sitting in front of a computer screen that shows options for learning to deal with drug cravings.

Two computerized programs improved outcomes when they were used to supplement or partially replace in-person behavioral therapy for drug addiction in recent NIDA-sponsored trials. In one trial, adding Computer-Based Training for Cognitive Behavioral Therapy (CBT4CBT) to standard methadone-maintenance therapy increased abstinence from cocaine among patients who were addicted to both opioids and cocaine. In the other trial, outpatients who interacted with the Therapeutic Education System (TES) in place of attending some in-person therapy sessions had almost twice the odds of abstaining from all drugs and alcohol in the last 4 weeks of a 12-week trial period.

Many researchers are developing and testing computerized approaches to substance abuse prevention, treatment, and aftercare. The goals are to increase access to patient care, enhance patient engagement, and reap additional advantages, such as the ability to collect data and deliver personalized therapeutic content directly to patients in their communities. The treatment programs evaluated in these two studies are among those that have advanced furthest in testing to date.

Screen Test for a Proven Therapy

Created by Dr. Kathleen Carroll and colleagues from the department of psychiatry at Yale University School of Medicine, CBT4CBT uses a variety of interactive techniques, games, quizzes, and short films, to teach patients how to recognize triggers for substance use, cope with drug cravings, and counter negative thoughts and challenging situations.

“CBT4CBT teaches people skills and strategies in an entertaining way,” says Dr. Carroll. Each of the CBT4CBT movies depicts a character in a high-risk situation for substance use. Alternative endings contrast a negative outcome, such as a relapse to drug use, with a positive one that is the direct result of a skill a patient has just learned. “This delivers the powerful impression of seeing someone successfully using the new skill rather than just hearing about it in an abstract way from a clinician,” says Dr. Carroll.

CBT4CBT also focuses on decision-making skills, problem-solving, and other cognitive, affective, and behavioral self-control strategies. A typical homework assignment might have the participant practice thinking through each small decision he makes in the course of a day, to build awareness of his decision-making processes. “CBT4CBT’s teaching model for coping with cravings involves encouraging people to ‘go with,’ and learn to tolerate, cravings and other strong feelings and to prevent acting impulsively on them,” says Dr. Carroll. “It trains a person’s skills in cognitive control.”

Dr. Carroll and colleagues’ test of CBT4CBT is the first replication of a randomized clinical trial of a computer-assisted therapy for addiction. In the earlier, smaller trial (see Computer-Based Interventions Promote Drug Abstinence), patients reduced their use of a range of substances when the program was added to the standard therapy in a community treatment center.

In the new trial, conducted at the APT Foundation in New Haven, Connecticut, 101 adult outpatients dually addicted to cocaine and opioids received daily methadone-maintenance treatment and weekly group therapy. All the patients also met twice a week with a research assistant who monitored their drug use and collected urine specimens to be tested for cocaine and other drugs. At these visits, roughly half (54 patients) also had access to CBT4CBT on a dedicated computer.

During 8 weeks of treatment, the patients who used CBT4CBT were twice as likely as those who did not to achieve 3 consecutive weeks of abstinence from cocaine use (36 percent versus 17 percent). They also provided a higher percentage of completely drug-free urine samples, although this difference was statistically significant only among patients who completed the full course of treatment.

“This difference in abstinence is really a big deal,” says Dr. Carroll. “It points to how people are going to fare in the future.” Bearing out this assessment, follow-up interviews and urine screens disclosed that the patients in the CBT4CBT group reduced their cocaine use more than those in the control group for 6 months after the end of treatment (see Figure 1).

The researchers acknowledge that participants continued to use cocaine at high rates throughout the treatment period, such that 75 percent of urine samples submitted by CBT4CBT patients, and 81 percent of those submitted by control patients, tested positive for the drug. The researchers note, however, that cocaine addiction is a highly refractory condition among methadone patients, and that the results obtained with CBT4CBT compare favorably to those reported with other evidence-based therapies.

See text description below Figure 1. Computerized Cognitive Behavioral Therapy, CBT4CBT, Benefits Patients With Co-Occurring Opioid and Cocaine Dependence Patients who used the CBT4CBT program in addition to attending in-person group therapy sessions reduced their cocaine use faster and further than patients who received in-person group therapy alone. The differences were still significant 6 months after the end of a 2-month course of treatment.
Text Description of Figure 1

The figure shows a line graph comparing cocaine use in the CBT4CBT group with that in the treatment-as-usual group. The vertical (y)-axis shows days of cocaine use per month, and the horizontal (x)-axis shows the month of self-reported cocaine use. At month zero, patients in both groups had similar levels of self-reported cocaine use of about 15 days per month; cocaine use then gradually declined in both groups over the 8-month period studied, with a somewhat more pronounced decline in the CBT4CBT than in treatment-as-usual group. By month 8, use had decreased to about 4 days per month among patients in the CBT4CBT group and to about 6 days among patients in the treatment-as-usual group.

Signs of Success With TES

The TES program is based on the community reinforcement approach for treating drug addiction. As such, it covers a broader range of topics than CBT4CBT. For example, some of its 65 interactive modules, which include short films and self-assessment quizzes, aim to teach patients relationship and employment skills to enable them to elicit and enjoy social support and reintegrate into their communities. A patient’s homework might simply involve taking their family to a movie and reporting back on the experience.

“Our study was an opportunity to take TES and examine it in a clinical trial with a take-all-comers population of treatment-seeking men and women,” says Dr. Aimee Campbell, who, along with Dr. Edward Nunes and colleagues at the New York State Psychiatric Institute based at Columbia University Medical Center, examined TES’ effectiveness in a multisite trial with 10 outpatient treatment program partners affiliated with NIDA’s National Drug Abuse Treatment Clinical Trials Network. The trial participants were 507 men and women who started treatment between June 2010 and August 2011 for addiction to alcohol, cocaine, opioids, marijuana, or other substances.

All of the trial participants received their particular clinic’s standard outpatient therapy, which was usually group therapy. Half (255 patients) spent all of their therapy time in clinician-led sessions, and half (252 patients) substituted working independently with TES for approximately 2 hours per week of clinician-led sessions. To enhance patients’ motivation, TES incorporates contingency management (CM), with rewards ranging from $1 to (rarely) $100 for completing modules and submitting negative urine screens (for more on CM, see Training Workshops Boost Approval of Contingency Management).

The patients who used TES had a 60 percent higher rate of biologically confirmed abstinence from 10 different substances at the end of the 12-week treatment period. Among participants who were still using substances when they entered the study, the TES-based intervention nearly doubled the likelihood of achieving abstinence compared with the controls. However, among those who had achieved abstinence at the start of the study, abstinence rates with TES and treatment as usual did not differ significantly (see Figure 2).

See text description below Figure 2. Therapeutic Education System (TES) Reduces Demand on Clinicians, Increases Abstinence Trial participants who used the TES system in lieu of attending some clinician-led therapy sessions had increased abstinence if they were using substances at the start of the trial. At the 3-month follow-up, no persistent advantage of TES was found.
Text Description of Figure 2

The figure shows a line graph comparing abstinence from various substances in the TES + treatment-as-usual group with that in the treatment-as-usual-only group; the graph represents two datasets: one for patients who were abstinent already at baseline and one for patients not abstinent at baseline. The vertical (y)-axis shows the proportion of abstinent patients (in percent), and the horizontal (x)-axis shows time points in the study represented by half-week visits and as two follow-up points at 3 and 6 months. For patients already abstinent at baseline, both treatment groups showed very similar levels of abstinence over the course of the 6-month period assessed, with abstinence of just under 90 percent at the first half-week visit gradually decreasing to about 55 percent in both groups at 6 months. For the patients not abstinent at baseline, among those in the TES + treatment-as-usual group, abstinence gradually rose from just over 20 percent at the first half-week visit to about 40 percent at the 24th half-week visit, and among those in the treatment-as-usual-only group, abstinence rose from just over 10 percent at the first half-week visit to about 22 percent the 24th half-week visit. The differences in abstinence between the two groups narrowed thereafter, with about 36 percent and 30 percent of the patients in the TES + treatment-as-usual and treatment-as-usual-only groups, respectively, being abstinent at the 3-month follow-up, and with about 32 and 30 percent of the patients in the TES + treatment-as-usual and treatment-as-usual-only groups, respectively, being abstinent at the 6-month follow-up.

“The impact of the TES treatment was greatest among those participants who had a more severe problem,” says Dr. Nunes. It stands to reason, he explains, that the participants who were abstinent at the beginning of the study did well regardless of the treatment they received. “They were on a good prognosis trajectory to begin with,” he says.

TES also improved treatment adherence. At the end of the treatment period, almost 50 percent of the participants assigned to receive TES were still participating, whereas that proportion had dropped to about 40 percent among the participants assigned to receive only the treatment-as-usual program (see Figure 3).

At the 3- and 6-month follow-ups, the TES participants no longer showed significantly higher levels of abstinence than the treatment-as-usual group. This lack of a long-term effect suggests the need for providing access to the web-based interventions over longer periods of time and a more intense focus on relapse prevention, says Dr. Nunes.

See text description below Figure 3. TES Retains Patients in Treatment Patients who used the TES program instead of attending some clinician-led therapy sessions remained in treatment longer than those who exclusively attended clinician-led sessions.
Text Description of Figure 3

The figure shows a Kaplan-Meier plot showing treatment retention for the TES + treatment-as-usual group and the treatment-as-usual-only group over a 12-week treatment period. The vertical (y)-axis shows the proportion of patients remaining in treatment (in percent), and the horizontal (x)-axis shows the number of weeks in treatment. At time zero, both groups had 100 percent of patients in treatment. Over time, treatment retention gradually decreased in both groups, but less so among the patients in the TES + treatment-as-usual group, which on average showed a treatment retention of 5 to 10 percentage points higher than that of the treatment-as-usual-only group during weeks 2 to 12. By the end of the study at week 12, the percentage of patients remaining in treatment had dropped to about 50 percent in the TES + treatment-as-usual group and to about 40 percent in the treatment-as-usual-only group.

Invoking Synergy To Increase Abstinence

We’re envisioning and testing the TES intervention as something that clinicians can prescribe to patients to augment and synergize the face-to-face therapy that is part and parcel of most treatment programs for drug abuse,” says Dr. Nunes.

Researchers have noted that it’s challenging to train clinicians to deliver cognitive behavioral interventions in a consistent manner. Sometimes “drift” occurs, says Dr. Nunes, in which therapists may stray from the scripted material. “The computer always delivers the same content faithfully,” he says. Where computers sometimes fall short is being able to tailor content to the particular needs of individual patients, something that clinicians are often better at. “The strengths of computer-delivered therapy and those of the human therapist are complementary, and this synergy benefits the patient,” says Dr. Nunes.

Dr. Lisa Onken, former chief of NIDA’s Behavioral and Integrative Treatment Branch, says, “Dr. Nunes and colleagues’ findings suggest that interventions such as TES might be used to expand access and improve addiction treatment outcomes.” She comments, however, that further research is needed to determine how much each of TES’ many components contributes to its effectiveness. She speculates, for example, that much of the reduction in drug use observed in the study could be due to the incorporation of CM, which in itself is a proven effective treatment, rather than the content of the modules. If that were so, it could help explain the lack of persistent impact in the trial’s follow-up period.

“Efficacious behavioral treatments for drug addiction are sorely needed, but there have been many obstacles to developing them,” says Dr. Onken. “These two studies add to the literature suggesting that behavioral treatments can be successfully computerized and added to treatment as usual with substantial benefits to those struggling with addiction.”

These studies were supported by NIH grants DA015969 and DA09241 (Dr. Carroll and colleagues), and DA013035, DA015831, DA013034, DA013720, DA013732, DA020024, DA013714, DA015815, and DA022412 (Dr. Nunes and colleagues).


Carroll, K.M.; Kiluk, B.D.; Nich, C. et al. Computer-assisted delivery of cognitive-behavioral therapy: efficacy and durability of CBT4CBT among cocaine-dependent individuals maintained on methadone. American Journal of Psychiatry 171(4):436-44, 2014. Full text

Campbell, A.N.C.; Nunes, E.V.; Matthews, A.G. et al. Internet-delivered treatment for substance abuse: a multisite randomized controlled trial. American Journal of Psychiatry 171(6):683-90, 2014. Full text