The purpose of the trial is to investigate the effect of specialized treatment of cannabis abuse among young people with psychosis. The specialized treatment consists of manualized individual treatment and group therapy. It will be compared with the standard treatment, which consists of non-specialized, non-manualized treatment. 140 patients will be randomised to one of the two treatments, and the investigator(s) will be blinded to the treatment received.

PURPOSE OF THE TRIAL:

The randomized clinical trial aims to examine the effect of specialized treatment of cannabis abuse amongst young people with psychosis. Specialized addiction treatment is compared with treatment as usual. Treatment as usual is a non-specialized, non-manualized treatment.

In order to examine the effect of the two treatments, patients are randomized to either specialized addiction treatment or treatment as usual.

INTRODUCTION:

Clinically, consensus exists about the fact that cannabis can cause psychotic symptoms that cannot be distinguished from schizophrenia, and that abuse of cannabis among patients with psychosis can maintain and worsen the psychotic symptoms. This has also been found in scientific studies [Linszen et al., 1994]. Several studies show that use of cannabis increases the risk of development of schizophrenia-like symptoms, especially in young men disposed to developing psychosis [Zammit et al., 2002; Arseneault et al., 2002; Arseneault et al., 2004; Henquet et al., 2005]).

Comorbid schizophrenia and substance abuse (dual-diagnosis) have been associated in several reviews with lack of compliance to treatment. A systematic examination of hospitalized patients with schizophrenia found that lack of compliance to the treatment was related to substance abuse and lack of insight (Kamali et al., 2001). A follow-up study of patients with schizophrenia and cannabis abuse found significantly more rehospitalizations, worse psychosocial functioning, higher levels of thought disorders and more pronounced paranoia compared to non-abusing patients with schizophrenia (Caspari, 1999).

In the "Reference Programme for Schizophrenia" [Sekretariatet for reference programmer, 2004], the Danish National Board of Health makes the following recommendation:

"Since patients with schizophrenia and concomitant cannabis abuse have a poorer prognosis, the treatment system should therefore develop treatment methods that are effective at alleviating cannabis abuse."

Evidence for the effect of treatment for cannabis abuse:

There is little evidence concerning the treatment of cannabis abuse, but the literature suggests that it can be treated using methods that are effective for other types of abuse. Several randomized trials show an effect of cognitive behavioural therapy [Waldron and Kaminer, 2004]; however, treatment programmes that combine motivational interviewing, family involvement, and cognitive behavioural treatment seem to be more effective [McRae et al., 2003; Carroll, 2005].

Evidence for the effect of treatment of patients with dual diagnosis:

A Cochrane review from 2002 concludes that insufficient evidence exists to show that any intensive treatment method is superior to others [Ley et al., 2000]. It is recommended not to offer the treatment separately but as a part of treatment programmes [Linszen et al., 1994b; Drake et al., 2004]. A randomized trial showed that the combination of cognitive behavioural therapy, motivational interviewing, and family involvement had a significant positive effect on level of functioning, psychotic symptoms, and duration of periods without abuse, compared with regular treatment [Barrowclough et al., 2001]. Two reviews conclude that there is positive evidence for integrated treatment with motivational interviews, cognitive behavioural therapy (individual or group-based), 12-step treatment, and a harm-reduction approach [Ziedonis, 2004; Rachbeisel et al., 1999].

Evidence for the effect of group-based intervention for cannabis abuse:

There is no evidence to show that group-based interventions are superior to individual treatment (McRoberts et al., 1998; Greene, 2002). Treatment in groups is less expensive. In a literature review, Weiss et al. conclude that specialized group therapy can reinforce the effect of the existing treatment (Weiss et al., 1992). A study from 1999 that robustly implemented a manualized behavioural group intervention showed a significant positive effect on symptoms, level of functioning, and lower costs for supportive services [Jerrell and Ridgely, 1999].

There is evidence for the efficacy of a stepwise or phase-specific treatment that is successively based on engagement, motivation, coping with symptoms, and preventing relapse [Drake et al., 2004].

Conclusion:

A review of the literature shows that there is a lack of randomized trials that can ensure that the treatment of patients with dual diagnosis is evidence-based. This trial will assist in solving this problem by building on best practice. Therefore, we plan a trial in which group-based, combined treatment, such as motivational interviewing, psychoeducation, cognitive behavioural therapy, and social skills training is compared with treatment as usual.

INTERVENTIONS:

1. The experimental intervention: Cannabis and Psychosis (CapOpus) - the specialized treatment programme:

   The patient is connected to a case manager who is offered education and supervision by one of the two addiction consultants employed in the trial-project.

   The addiction consultants are both directly and indirectly involved in the treatment of the patient. During the month prior to the group intervention, one of the addiction consultants is in contact with the patients once or twice a week. A meeting is also held with the patient's family and fortnightly contact with the case manager is established.

   During the three months of group intervention, one of the addiction consultants has weekly individual contacts with the patient and two meetings with the family. In addition, the patient follows the weekly group intervention (12 meetings of 1½ hours) and has fortnightly consultative contacts to the case manager.

   During the two months following the group intervention, the addiction consultants are in weekly contact with the patient and the family is invited to a meeting. The addiction consultants contact the patient's case manager every three weeks.

   For the purpose of creating alliance and motivation, the treatment starts with motivational interviewing [Miller, 1983; Miller and Rollnick, 1991]. There is good evidence for the efficacy of motivational interviews in short-term treatment [Hettema et al., 2005; Burke et al., 2003].

   The patient formulates individual goals for the treatment and is offered the group intervention. The groups consist of 6 to 8 patients, with the addiction consultants as trainers. Each group runs for three months with weekly sessions lasting 1½ hours. The group is conducted at a fixed time on a weekday and the structure of the agenda is the same at all meetings.

   The group intervention is followed by two months during which the individual weekly meetings with the patient continue. The two addiction consultants offer consultative assistance to the patient's case manager, who also has the possibility of involving the consultants directly in the treatment of the patient. The entire specialized treatment programme, in which the addiction consultants are involved, amounts to six months.

   A manualized treatment programme based on the Australian EPPIC manual [Hinton et al., 2002] especially developed for first-episode psychotic patients with cannabis abuse is used. This method is well described and incorporates methods with a high degree of evidence. Motivational interviewing with analyses of advantages and drawbacks of continued abuse are used. Instruction is given in coping skills in relation to craving and situations that usually trigger abuse, and in developing personal strategies for avoiding or handling these situations. Furthermore, strategies for handling withdrawal symptoms and for preventing relapse are facilitated.

   General coping skills are introduced, such as handling unpleasant emotions, stress-management, social skills training, and relaxation techniques.

   An element of contingency management is introduced to enhance motivation for participation in the group intervention. In a Cochrane Review from 2006, Denis et al. conclude that contingency management treatment may enhance outcomes combined with CBT or motivational enhancement [Denis et al., 2006] Contingency management in the CapOpus trial is solely connected to positive reinforcement of attendance in the group intervention. It has no connection to whether or not the use of cannabis is decreased. Patients are offered participation in free excursions, cinema visits etc. in the company of one of the addiction consultants. Also, sandwiches are served in conjunction with group sessions.

   An important part of "Cannabis and Psychosis" is to understand and to help the patients understand the mechanisms that prevent them from abstaining from cannabis. The overall target is harm reduction, a method which has proved effective in several studies [Ziedonis, 2004b; Rachbeisel et al., 1999]. The treatment is based on the patients' own goals for tackling cannabis abuse.

   The treatment is structured around the circle of change [Prochaska, 1991; Prochaska and Diclemente, 1992]. This describes changes in behaviour as a process that runs through phases of pre-contemplation, contemplation, preparation, action, and maintenance. Relapse is considered to be an integrated part of the process, after which the phases must be repeated.

   Pedagogy of the group intervention takes into account that psychotic patients most often have subnormal cognitive functioning. It is therefore important that the structure of each session is predictable and over-learning is encompassed. This is done by using the same agenda at each session:

   • Round - what has happened the past week.
   • Repetition from last session.
   • Homework assignment for this session.
   • Talk/psychoeducation (new topic in each session).
   • Discussion of participants' experience with the topic.
   • New homework assignment.
   • Finishing round - evaluation.

   The active participation of the patients is facilitated through discussions and role-playing and homework between each session. To increase motivation, the teaching is made relevant in relation to the life of each patient and their individual goals for the treatment. The overall principles are described in the manual, but each session leaves room for adaptation to the patients' wishes and stage in the circle of change.

   Therapist roles are directive and direct, but non-confronting non-critical. Emphasis is on empathy and positive reinforcement, problem solving and generalization to the patients' daily life. The atmosphere is sought to be relaxed and with room for humour. For the concluding part of the group intervention, a patient with former cannabis abuse is involved as a role model.

   Methods of treatment are all part of the cognitive therapeutical framework, using psychoeducation, cognitive behavioural therapy, and social skills training. The aim is to ensure that the patients gain insight into inappropriate patterns of thoughts and actions and develop coping strategies. This is facilitated through: Exploration of the patient's reasons for using cannabis and the connection to symptomatology. Mapping of advantages/disadvantages of cannabis use and cessation of use. Warning signs of craving/relapse. Problem-solving skills and coping skills for symptoms. Use of behavioural chain analysis (Mørch and Rosenberg, 2005). Work with negative automatic thoughts and alternative thoughts. Social skills training (e.g. to refuse drugs, solve conflicts, and engage in new contacts). Facilitating daily and recreational activities. Relapse prevention and developing a crisis plan.

   Structure of the modules in the group-intervention:

   The structure of the intervention is based on the stages in Prochaska and Diclemente's model of change (circle of change). The model is based on the view that the patient often fluctuates between the different stages and undergoes relapses before being able to reach the final stage, which marks lasting change.

   In order to support the group and patient in the process of changing lifestyle, the therapists must be able to identify the current stage of the group/patient. Thus, the intervention can be aimed specifically, either by being oriented towards motivation enhancing or advisory techniques, or by changing focus, e.g. from exploring reasons for change to making plans for change. The purpose is to motivate the patient to move on through the stages; therefore, the model is also called "the motivational cycle" or "circle of change". In this way, each session in the group can be adapted to the participants' motivational level and stage of change.

   Sessions for the pre-contemplation stage:

   Psychoeducation about:

   Knowledge of psychotic symptoms. Knowledge of the psychosis-inducing effect of cannabis. Knowledge of risks and harms caused by cannabis use. Knowledge of abstinence symptoms following cannabis use (Budney et al., 1999). Knowledge of the effect of decreased use (harm reduction).

   Mapping the individual pattern of use, with focus on disadvantages of cannabis use and the triggers for craving and use. This is done by using registration charts.

   Contemplation stage:

   Exploration of pros and cons of cannabis use and cessation of use. This is used with the aim of replacing some of the advantages of cannabis use with other interventions.

   Exploration of connections between cannabis abuse and symptomology. Ambivalence and resistance is addressed from an accepting viewpoint.

   Preparation stage:

   Individual goal-setting with focus on disadvantages of cannabis use. Exploration of the connection between cannabis use and symptomatology. Development of symptom-coping skills and alternative coping strategies. This is facilitated through exploration of pros/cons, behavioural chain analysis and exploration of negative automatic thoughts and actions in relation to the cannabis abuse.

   Social skills training to enhance ability to refuse drugs, solve conflicts, and initiate contacts to non-users. Introduction to the problem-solving model (Mørch and Rosenberg, 2005).

   Action stage:

   Problem solving, where coping strategies are tested. Mapping of warning signs for relapse and triggers for cannabis use. Individual crisis plans in relation to relapse and worsening of symptoms.

   Maintenance stage:

   Development of new skills and habits in daily life. Support (in cooperation with the case manager) to maintain an active life with focus on activities of daily life (ADL), exercise and work/education/recreational activities. Widening of personal network with non-cannabis users. Training of general coping skills such as social skills training, handling of unpleasant emotions, stress management, and relaxation techniques. Strategies for relapse prevention and coping strategies for craving are tested in daily life. Involvement of patient with former cannabis abuse as a consultant and role model.

   Relapse:

   De-dramatizing and normalization with focus on learning from relapse. Support in order to re-engage in the stages of change, with focus on the patient's own coping strategies.

   Programme adherence:

   To ensure programme adherence, the number of contacts with the patient, the patient's family, and the case manager are registered during the six-month intervention period. Forms for registration of programme adherence are implemented in individual and group sessions. Registration forms are also used to ensure that the planned content in the group intervention is sufficiently implemented. These forms are anonymized and reviewed by the research assistant.

   The addiction consultants are supervised by an external supervisor with expertise in cannabis abuse, motivational interviewing, and cognitive behavioural therapy.

2. The control intervention: non-specialized individual treatment:

The non-specialized treatment programme is carried out by staff in OPUS (a treatment for young people with psychosis in Copenhagen), in Assertive Community Treatment or in Community Mental Health Centres. Thus, the control intervention is identical to the treatment that is ordinarily offered to this patient-group. The frequency of contact with the patients may vary. There is no standardized manual for this treatment. The treatment approach is supportive and not condemnatory. It is important to advise the patient about alternative coping strategies and to encourage every small reduction in abuse.

RESEARCH PLAN:

A research assistant is responsible for conducting interviews at the time of inclusion in the trial, at six months, and again ten months later. Thereby, the first follow-up interview is held when the specialized CapOpus treatment is concluding. The second follow-up interview is held four months after the specialized treatment has ended.

The assessment instruments used as effect measures are all validated psychometric scales, which the research assistant is certified to use.

Referred patients will be assessed by a research assistant to ensure that they meet the criteria for inclusion.

Randomization and blinding:

Randomization into one of the two treatment arms is performed after the baseline assessment. Firstly, randomization is stratified by severity of cannabis abuse as measured by TLFB, by creating two sub-strata; one of patients having used cannabis less than 15 days in the preceding 30 day period, and one of patients having used cannabis 15 days or more. This is done to avoid the risk of heavy-users of cannabis being overrepresented in either of the treatment groups. Similarly, randomization will be stratified based on the referrer - OPUS, CMHC, or ACT, as it is unlikely that patients from these three referrers are comparable with regard to psychotic symptoms and level of functioning.

The Copenhagen Trial Unit (CTU) will perform the centralized randomization, and only the CTU will know the block size used to randomize. A research assistant provides information on the patients to the CTU after conducting the baseline assessment. The CTU then performs the randomization procedure based on computer-generated allocation sequences, and communicates the allocated treatment to the CapOpus consultants responsible for the CapOpus experimental intervention. If a patient is allocated to CapOpus intervention, the consultants contact the patient. Regardless of allocation, the consultants also contact the patient's case manager, so that the treatment as usual intervention can be commenced. This procedure ensures that the research assistant, who will be doing the subsequent patient assessments, is blinded to the allocated intervention. Blinding is maintained until the end of the trial. Patients are instructed not to reveal details that may cause the research assistant to deduce which treatment they are receiving. At each of the two follow-up assessment points, the research assistant will register a guess as to which treatment the patient is receiving. When the trial is finished, inter-rater reliability between the actual treatment and the research assistant's guess will be estimated using Cohen's kappa coefficient to assess whether blinding has been successful.

Effect measurements:

The following effect measurements are used at baseline, after six months and after ten months:

1. Severity of abuse. Assessment of number of days with cannabis abuse during the last month with Time Line Follow-Back (Sobell and Sobell, 1992). This effect measurement is the best for measuring the effect the trial aims to influence; namely whether number of days with cannabis use can be reduced. Information supplied by patients on cannabis use will be validated with the results of blood analysis. Blood tests are taken at the six-month and ten-month follow-up interviews.

   If the comparison of results of blood analysis and the patient's self-report in Time Line Follow-Back indicates that the patient's reports are reliable, Time Line Follow-Back is used to measure effect.

2. Influence and severity of other substance use, including prescribed medication, and establishment of the severity of consequences of cannabis use, are assessed using sections 11 (use of alcohol) and 12 (use of psychoactive substances other than alcohol) of the SCAN interview (Schedules for Clinical Assessment in Neuropsychiatry).[Wing et al., 1990]

3. Psychosis symptoms are assessed by the use of The positive and negative syndrome scale (PANSS) for schizophrenia. [Kay et al., 1987] A psychiatrist, who is blinded to patients' allocations, rates samples of videotaped interviews to measure the reliability of the rating of these interviews

4. Cognitive function: Danish Adult Reading Test (DART) is used as an estimate of prepsychotic IQ. Speed of information processing is assessed with BACS' symbol coding and with Trailmaking A. Attention/vigilance is assessed by CBT test. Working memory is assessed with Trailmaking B. Memory and verbal learning is assessed with Hopkins Verbal Learning Test. Executive functioning is assessed with NAB Mazes.

   Social functioning (major life areas; community, social, and civic life) is assessed with WHODASII.

5. Quality of life is assessed with MANSA and EQ-5D.

6. User satisfaction is assessed with Client Satisfaction Questionnaire [Larsen et al., 1979]

7. Expenses for the experimental intervention (CapOpus) and control intervention is measured by examining number of outpatient treatment and bed days in both treatment groups.

Follow-up is planned to be undertaken at six months, because patients in specialized CapOpus treatment have just finished the intervention at that time. Follow-up at ten months is chosen, because this allows for a four-month follow-up period after the end of specialized treatment.

Sample size and power calculation:

Our power sample size calculation is based on a t-test with five fewer days of cannabis use in one of the treatments, and with a standard deviation of five. This means that 22 patients should be allocated to each of the two treatments. In the OPUS-study, those receiving standard treatment and who had a substance abuse at baseline had a drop-out rate of 37% after one year. Generalizing this to the present study, and ignoring the likelihood of a smaller drop-out in the intervention arm, we require at least 35 patients to be allocated to each intervention. However, we aim to include between 60 and 70 patients in each group in order to be able to measure differences in secondary outcomes. In the group intervention, the maximum inclusion is 32 patients annually (4x8); therefore, it is necessary for the project to continue for several years in order to recruit the necessary number of patients.

The necessary number of patients can be ensured by launching the project as a cooperation between the three OPUS teams in Copenhagen, Assertive Community Treatment, Community Mental Health Centres, and treatment facilities for cannabis abusers in Copenhagen and Frederiksberg.

Statistics:

The main null-hypothesis to be tested is that there is no difference between the two treatment arms with regard to decrease in cannabis use. All randomized patients will be analyzed, including those who stop receiving treatment, according to the intent-to-treat principle. Continuous outcome measures, including the primary outcome of reduction of days with cannabis abuse, will be analyzed using analysis of variance (ANOVA) mixed models with repeated measurements. This will include interaction analyses of time and type of treatment, in order to evaluate the effect over time. Binary outcome measures will be analyzed using multivariate logistic regression models. The subgroup of patients who complete the entire experimental treatment will be compared with all those randomized into the trial. Sensitivity analyses will be carried out to evaluate the effect of treatment in patients that are inaccessible for follow-up. This will be conducted by multiple imputation methodology.

Publication:

The results of the trial will be published in national and international journals. Authorship is determined in an agreement of cooperation for the entire project. The publications will be publicized according to Consort and Vancouver guidelines of publication of randomized trials.

Ethical considerations:

The participants are invited by letter to an assessment by the research assistant, who presents the trial to the patients, both orally and in writing. The research interviews can take place at Bispebjerg Hospital, in OPUS, in the patient's home, or wherever it may be possible. In the oral presentation, it is stated explicitly that participation is voluntary and not dangerous, and that the patient can withdraw informed consent at any time without any consequences for the treatment. Oral and written informed consent are obtained. The trial is registered with the ethics committee, the data surveillance agency and 'clinicaltrials.gov'.

Inclusion Criteria:

• The patient must fulfil research criteria for F2 in ICD-10 (schizophrenia and schizophrenia-like conditions) and diagnosis of F12 (mental illness or disturbances caused by cannabis)
• The patient must understand Danish language to the extent that assessment and treatment can be conducted without an interpreter.
• The patient must give informed consent to participate in the trial. In addition, the patient must consent to participate in the specialized treatment (CapOpus) and consent to continuation or initiation of treatment for the psychiatric condition.
• Patients in OPUS, Assertive Community Treatment, Community Mental Health Centres, psychiatric wards, and others who meet the criteria can be included in the trial. (OPUS is a treatment for young people with first-episode psychosis in Copenhagen - www.opus-kbh.dk)
• Cannabis abuse must be the dominant form of abuse. Other substance abuse can be present sporadically.
• Patients must be 18 to 35 years of age and have legal residence in or near the municipality of Copenhagen or Frederiksberg

Exclusion Criteria:

• Patients who meet the criteria of alcohol-dependence syndrome (F10.2), opioid dependence syndrome (F11.2) or cocaine dependence syndrome (F14.2)
• Patients who do not give informed consent