The HELPER programme: HEalthy Living and Prevention of Early Relapse three exploratory randomised controlled trials of phase-specific interventions in first-episode psychosis.

Participants

We identified potential participants among outpatients on waiting lists for CBTp from Lancashire Care NHS Foundation Trust’s and Pennine Care NHS Foundation Trust’s early-intervention services. Inclusion criteria were age 18–35 years and first episode of DSM-IV22 schizophreniform disorder, schizophrenia, schizoaffective disorder or delusional disorder, confirmed by semistructured interview. 23 Exclusion criteria were International Classification of Diseases, Tenth Edition (ICD-10)24 organic brain disease; DSM-IV substance abuse or dependence; primary diagnosis of DSM-IV substance-induced psychosis; and insufficient fluency in English to participate in neuropsychological assessment. All participants provided written informed consent and the study was approved by the Bolton NHS Local Research Ethics Committee (reference 08/H1009/76) and was consistent with the UK Research Governance Framework for Health and Social Care. 25

Outcome assessments

At baseline, demographic details (including self-ascribed ethnicity as a potential moderator of CR and CBTp outcome) and medication were recorded. The PSYRATS (Psychotic Symptom Rating Scales26) were the primary outcome measure, completed with other measures at baseline and after 12 and 42 weeks’ follow-up (Figure 1). The PSYRATS validly and reliably assess the severity of delusions and hallucinations in first-episode schizophrenia. 27 Intraclass correlation coefficients between assessors were > 0.99 for subtotals and the total score.

FIGURE 1.

Study design.

Assessments at baseline and 12 and 42 weeks included secondary measures of symptoms and function: the Positive and Negative Syndrome Scale (PANSS28), the Calgary Depression Scale for Schizophrenia (CDSS29), the Rosenberg Self-Esteem Scale30 and the Social and Occupational Functioning Assessment Scale (SOFAS31). A seven-item version of the Insight Scale (IS32), with the hospitalisation item dropped for this community sample, was scored from 0 to 14.

The PSYRATS were also rated at 6-week intervals during the CBTp envelope, at 12–42 weeks. Participants completed these interviews in person, or by telephone if they preferred. As CBTp was sometimes delayed or prolonged, and therefore incomplete by 42 weeks, the PSYRATS were also completed by telephone or in person at 54 weeks.

Cognitive–behavioural therapy for psychosis typically aims to progress from engagement, through examination of the nature of individual problems, to integration of problems into a formulation describing origins, schemata and maintaining mechanisms. At the end of CBTp, therapists assessed this progression using a five-point score. 33

An assessor blind to allocation extracted the number of sessions of CBTp from case records and identified readmission and relapse (defined, as elsewhere,34 as an exacerbation of psychotic symptoms lasting at least 2 weeks, leading to a change in management).

Neuropsychological assessments

Intelligence quotient (IQ) was assessed at baseline with the Wechsler Adult Intelligence Scale (WAIS) block design subtest and the Wechsler Test of Adult Reading (WTAR). 35 Secondary neuropsychological assessments of attention, executive function and memory were completed at baseline and at 12 and 42 weeks using the Wisconsin Card Sort Task (WCST),36 trail making,37 the 0-back version of the n-back task,38 paragraph recall39,40 and the Rey–Osterrieth Complex Figure Test41 (Table 1 contains details and rationales).

Allocation

Within 3 days of initial assessment assessors faxed details of participants to identified administrators who were independent of the trial team and unaware of the hypothesis. They used randomly permuted variable blocks53 to allocate participants to the experimental group or the comparison group, masked from assessors, and then communicated allocation by telephone to the researchers providing the study interventions, before faxing them participants’ details.

Interventions

Cognitive remediation was provided in patients’ homes or service bases using the CIRCUITS software, (Reeder and Wykes, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, UK)54 run from a secure website via the internet or on a DVD. Based on a CR programme validated in schizophrenia patients,42,55 it used a colourful, engaging, virtual town as an environment, guiding participants through a sequence of tasks. This provided a social context for tasks, each requiring a specific mixture of skills (e.g. sustained attention, working memory, registration and recall, planning) and with specific criteria for progression. Early tasks prepared for later ones of increasing difficulty and complexity. Trainers supported participants and could enter their virtual environment with privileges that allowed them to modify the sequence.

Trainers were psychology graduates who had undergone 1 week of specific training (a regular open course provided by Clare Reeder and Til Wykes) to deliver CIRCUITS. CIRCUITS is based on well-established CR principles (Table 2).

The comparison condition was SC with support workers, with the duration of exposure matching that of exposure to the CR trainers. Both conditions provided interpersonal contact, warmth and unconditional positive regard within a professional relationship. Social activity (conversation, neurocognitively undemanding recreations) was the basis of sessions, although when necessary workers supported participants sufficiently to maintain their motivation to attend (i.e. non-directive listening concerning problems and symptoms).

Cognitive–behavioural therapy for psychosis was provided separately by therapists employed specifically for that purpose by the two NHS services who worked in seven different community clinics. All had specific CBTp training and were supervised by experienced senior CBTp therapists according to National Institute for Health and Care Excellence (NICE) standards. 1 CBTp quality was assessed using the Cognitive Therapy Scale for Psychosis (CTSpsy57), independently rated from randomly selected audio-taped sessions, provided that therapists thought it clinically appropriate and patients provided separate written consent. The mean CTSpsy score for seven CBTp interviews was 75.2% [standard deviation (SD) 14.5%].

Analysis

The effect of group on PSYRATS score was estimated using a mixed-effects model estimated using full information maximum likelihood with Stata version 11.2 (StataCorp LP, College Station, TX, USA). Allocation group and a group × time interaction term modelled the intervention’s effect. Time was entered as the square root of weeks from baseline; as PSYRATS scores tended to level off during follow-up this improved the model fit. 58 Demographic and other baseline potential confounders (lack of antipsychotic prescription, previous illicit substance use, diagnosis of schizophrenia) were removed by backward elimination, the criterion for retention being p < 0.20. PANSS, SOFAS and depression and self-esteem scales were analysed in the same way. Insight, in its mixed-effects model, was transformed towards normality by squaring the Insight Scale (IS) score. Insight changed relatively linearly over time and so weeks from baseline was entered as the time variable. A logistic regression against dropout with baseline variables as predictors assessed the pattern of missing data in relation to these variables.

As a categorical outcome, remission was defined as a PSYRATS score of zero. The relative risk of remission after CR was calculated and used to derive the number needed to treat/harm.

For neuropsychological measures, a global score was calculated. Individual baseline scores were transformed to normal distributions with a mean of 100 and SD of 15, with higher scores indicating better performance, before deriving a global score with the same characteristics. The same transformations were applied to 12- and 42-week follow-up scores, giving distributions with various means and SDs. The global score was then entered into a mixed-effects model.

In this relatively able population, WCST categories complete and complex figure copying (rather than recall; see Table 1) scores were too skewed by ceiling effects for this process. Categories complete and copying were modelled by ordinal logistic regression against follow-up scores, adjusting for baseline scores and potential confounders after backward elimination, clustering by clinic. To examine the sensitivity of ordinal logistic and linear regressions to the effect of dropout, all were repeated with cases probability weighted by a function of the risk of attrition, derived from the logistic regression against dropout. That is, the number of each case (n = 1) was multiplied by the reciprocal of the risk of completion of that type of case (calculated from a logistic regression using baseline characteristics to predict which cases would finish the study), so that cases of a type who more often dropped out appeared to be more frequent (e.g. apparent or weighted n = 1.43 if the probability of completion for a case with those demographic and clinical baseline characteristics was 0.7) to make up for the cases who did, in fact, drop out.

Time to relapse and readmission for allocation groups were compared by log-rank tests. Cox regressions adjusted hazard ratios (HRs) for potential demographic and clinical confounders, stratified by clinic.

Sample size

Assuming (after the trial by Eack and colleagues11) a PSYRATS effect size of 0.5, recruiting 64 participants provides > 80% power with a two-tailed alpha of 0.05, correlation of 0.5 between seven successive assessments and 25% dropout.

Experimental and comparison group characteristics

In total, 272 CBTp waiting list patients were screened for eligibility (Figure 2). Of these, 66 consented to participate, of whom four withdrew, leaving 62 to be randomised. One was subsequently excluded as a concealed history of previous psychosis rendered him ineligible and so 61 entered a modified intention-to-treat analysis (Table 3). Seven in the CR group and four in the SC group did not take maintenance antipsychotics (p = 0.51, Fisher’s exact test). No anticholinergics were prescribed and only one in each group took medication with high muscarinic receptor affinity (clozapine). Assessors accidentally discovered two participants’ allocation group during the trial, one from each group. CBTp therapists discovered eight participants’ allocation group (five CR group and three SC group) but their guesses at the allocation of the other participants were no better than chance (p = 0.43, Fisher’s exact test). Logistic regression against completion found no significant predictors of dropout.

FIGURE 2.

Consolidated Standards of Reporting Trials (CONSORT) flow diagram showing participant flow through the trial.

There was no significant difference in exposure to CR trainers/exposure to support workers between the two groups (SC: median 390 minutes, IQR 145–610 minutes, CR: median 375 minutes, IQR 125–600 minutes, Mann–Whitney U-test, p = 0.84; SC: median 8.5 sessions, IQR 3–10 sessions, CR: median 9 sessions, IQR 4–13 sessions, Mann–Whitney U-test, p = 0.20).

Did cognitive remediation improve cognition?

Cognitive remediation did not predict differences in global cognition scores [intercept (baseline) –1.7, 95% CI –7.7 to 4.4, p = 0.59; gradient (group × time) –0.73, 95% CI –1.84 to 0.38, p = 0.20)]. However, after the intervention the CR group completed significantly more WCST categories: after CR, 74% completed five categories (range 4–5 categories); after SC, 63% completed five categories (range 2–5 categories) [adjusted odds ratio (OR) 2.9, 95% CI 1.3 to 6.9; p = 0.012]. By final follow-up these gains were lost (adjusted OR 0.7, 95% CI 0.2 to 2.5; p = 0.61). The median score for complex figure copying was 35 (IQR 33–36) after CR and 33 (IQR 29–35) after SC (p = 0.11, Mann–Whitney U-test). After removing an outlier with persistent, specific, severe visuospatial impairment (baseline copying score 12; range of other participants’ scores 24–36, mean 32.89, SD 3.89; Dixon’s Q59 0.50, p < 0.01), CR had a trend towards better scores at 12 weeks (adjusted OR 3.83, 95% CI 0.99 to 14.77; p = 0.052) but not at the final assessment (adjusted OR 1.6, 95% CI 0.3 to 7.7; p = 0.56).

Did cognitive remediation improve the efficacy of cognitive–behavioural therapy for psychosis?

Symptoms

Cognitive remediation was not associated with significantly lower PSYRATS scores over the period of study (Figure 3) (intercept –2.8, 95% CI –10.7 to 5.2, p = 0.50; gradient 0.3, 95% CI –0.4 to 1.1, p = 0.39). The effect of CR on the PANSS score was non-significant (intercept 3.8, 95% CI –2.7 to 10.3, p = 0.25; gradient –0.4, 95% CI –1.5 to 0.6, p = 0.44). Eight participants in each allocation group had full remission of symptoms on the PSYRATS (OR 0.96, 95% CI 0.26 to 3.51; p = 1.00, Fisher’s exact test), giving a number needed to harm with CR of 116 (95% CI ranging from a number needed to treat of 3 to a number needed to harm of 12).

FIGURE 3.

Mean AH and delusions subscale scores by visit for each allocation group. AH, Auditory Hallucinations; B/L, baseline.

However, insight changed significantly more positively after CR than after SC (intercept –1.9, 95% CI –27.7 to 23.8, p = 0.88; gradient 0.55, 95% CI 0.08 to 1.01, p = 0.02). At 42 weeks the median insight score was 12 after both CR and SC, but the IQRs after CBTp (CR 9.5–13; SC 6–13) showed that fewer participants in the CR group had substantially impaired insight. There was no significant effect of CR on SOFAS score, depression or self-esteem.

Repeating multiple regressions for clinical and neuropsychological variables, weighting cases according to the probability of dropout, did not alter the significance of the results.

Did cognitive remediation improve the efficiency of cognitive–behavioural therapy for psychosis?

After CR, far fewer sessions of CBTp were required (Figure 4): a median of seven (IQR 2–12) sessions compared with a median of 13 (IQR 6–18) sessions for the SC group (p = 0.039, Mann–Whitney U-test). CR still predicted fewer sessions of CBTp after adjustment for potential confounders (coefficient –1.0, 95% CI –0.3 to –1.7; p = 0.011). Unmasking was not responsible: their allocation of 53 participants remained masked from therapists and the median number of sessions was six (IQR 1–12) after CR and 15 (IQR 7–19) after SC (p = 0.012, Mann–Whitney U-test). Allocation was unmasked for eight participants, for whom the median number of sessions was 14 (IQR 10–18.5) after CR and nine (IQR 3–18) after SC (p = 0.37, Mann–Whitney U-test).

FIGURE 4.

Number of sessions of CBTp by allocation group.

There was no evidence that level of formulation differed between the CR group and the SC group across all participants (adjusted OR 1.1, 95% CI 0.2 to 6.1; p = 0.88) or in masked patients alone (adjusted OR 0.7, 95% CI 0.2 to 2.3; p = 0.50).

Systematic review

Results

Three additional studies were identified. 73–75 We identified 12 studies in total73,75–85 Eleven compared the intervention with routine care and one80 used brief nutritional education as a control. The review by Álvarez-Jiménez and colleagues70 found that weight-management interventions resulted in a mean weight difference of 2.56 kg compared with treatment as usual (TAU). No difference in efficacy was shown between interventions delivered to groups and interventions delivered to individuals or between CBT interventions and nutritional counselling. Similar to the review of Álvarez-Jiménez and colleagues,70 our analysis showed that weight-management interventions are moderately effective, resulting in an average change in BMI of approximately 1 point (–0.708, 95% CI –1.07 to –0.349; Figure 6). We found no difference between studies that were CBT based and studies based on other therapeutic approaches. Unlike the study by Álvarez-Jiménez and colleagues,70 our analysis showed a trend towards greater effectiveness for individual interventions compared with group interventions [standardised mean difference (SMD) –1.04 (95% CI –1.68 to –0.41) vs. SMD –0.376 (95% CI –0.61 to –0.14) respectively]. We also found that studies in which supervised exercise actually formed part of the intervention showed a trend towards greater effectiveness than those that merely advised or encouraged exercise [SMD –1.153 (95% CI –2.57 to 0.27) vs. SMD –0.533 (95% CI –0.73 to –0.32) respectively]. In general, the effect of these interventions attenuated rapidly once the intervention had ceased. None of the reports indicated that the components of the intervention were underpinned by a specific psychological model of behaviour change and only one study had focused specifically on people with early psychosis. 76

FIGURE 6.

Forest plot showing the effectiveness of healthy-living interventions. Std diff., standard difference; Tx, treatment.

Qualitative interviews with service users

Qualitative interviews and focus groups with health professionals

Theoretical framework

The common-sense model (CSM) of self-regulation89 was chosen as the theoretical framework for the healthy-living intervention because it provides guidance on both motivation and the implementation of behaviour change. In the context of rapid weight gain in early psychosis, the CSM would suggest that people’s motivation to lose weight, and the behaviours selected to pursue that goal, depend on their personal beliefs about, or models of, the weight gain – what is causing it, what its consequences will be and how controllable it is. These personal models are influenced by information gained from abstract (knowledge-based) and concrete (experiential) sources. Furthermore, the CSM suggests that people are more likely to change (e.g. to eat more healthily and to exercise more) if they develop belief-congruent,90 ‘if–then’ action plans that they can use to regulate their behaviour. Feedback plays an important role in the CSM, as it suggests that behavioural attempts to regulate weight gain will be appraised and will feed back into both beliefs and subsequent behaviours. Thus, the model suggests combining motivational approaches (including education to bring about belief change) with action-planning interventions and the opportunity to review the effectiveness of changed behaviour. 90 Therefore, the CSM suggested that when planning our intervention we should begin by assessing participants’ own perceptions of their health and weight gain. By doing so we could then ensure that information about healthy living that was provided would be tailored to the needs of each individual. There would also need to be monitoring of the impact of behavioural feedback (e.g. increasing activity levels) on beliefs and of belief change on behaviour. A variety of techniques to modify perceptions would be incorporated to help service users benefit from abstract (knowledge-based) information derived from their own experiences of weight gain. Action planning would be incorporated, with goals set in accordance with motivational beliefs, plans would be developed to regulate behaviour (including using helpful cues and overcoming potential barriers) and explicit monitoring of the effectiveness of behaviour would be carried out.

Cultural adaptations

The area where the study was conducted had a large South Asian population, particularly among the younger members of the community who were likely to access the early-intervention service. In recognition of the importance of culture in influencing lifestyle factors such as diet and exercise, we used the relevant literature91 and consulted with one of the authors (NH) who has expertise in developing and delivering culturally sensitive interventions. Using both sources we identified the key issues that would need to be considered when delivering the intervention to young South Asians. Cultural adaptations to the intervention included recognition of the need to address the importance of understanding the wider family’s view of health education and to involve them when developing action plans. There was a need to translate written materials related to the delivery of the intervention into the predominant language of Urdu, which was still spoken by older family members. Some of the healthy living materials that were developed needed to have culturally specific advice such as information on healthy eating in specific communities. Personnel involved in the delivery of the intervention needed to have a cultural understanding and awareness of the ethnic groups with whom they were working.

Synthesis

The research team held a 2-day workshop meeting in August 2008 to synthesise the data collected and model the draft intervention. The synthesis meeting was conducted in two stages. In the first stage members of the study team presented the findings of each of the preliminary studies, which were then summarised (Table 4). The second stage involved interactive exercises and discussions, focusing on the possible content, duration and delivery of the intervention, drawing on the evidence presented in Table 4. The aim of this stage was to determine the optimum content, delivery and duration of the intervention as well as who should deliver it, what training they would require and any additional materials that might be needed (Table 5).

In summary, the draft healthy-living intervention that we developed was to be delivered in eight individual sessions over a 12-month period by a STR worker specially trained for this purpose. To promote consistency we produced a manual, outlining the content and key activities to be followed in each session. In addition to individual sessions, optional active group sessions were offered to encourage people to make healthy-lifestyle changes. The intervention was culturally adapted to meet the needs of young people from the South Asian community and strategies for behaviour change are underpinned by the principles of the common sense model. 89 Information about healthy living, including a range of healthy recipes, was provided for participants in a booklet.

Stakeholder consultation

The final stage of the development work involved presenting the draft intervention to a range of stakeholders including service users (n = 3), carers (n = 2) and professionals (n = 3) to elicit their views regarding its acceptability and to gather suggestions for the group activities and the duration of the intervention. There was agreement that the intervention was both acceptable and feasible. Two key suggestions emerged: (1) to develop a website that had similar intervention to the healthy-living book and (2) that we should involve service users in the running of the groups.

Following the stakeholder consultation we started to develop materials to support the delivery of our intervention. These included a training manual for the STR workers who would deliver the intervention, a healthy-living booklet for participants in the study and a website (see www.helper-interact.co.uk/: website active at time of writing). We also recruited four service users who were interested in running groups for the study and organised a workshop to train both the STR workers to deliver the individual intervention and the service users to support the STR workers when delivering group activities.

Aims and objectives

Our overall aim was to determine the feasibility, acceptability and effectiveness of our developed health living intervention utilising an exploratory RCT.

Methods

Results

A total of 971 participants were caseload screened (for flow diagram see figure 1 in Lovell et al. 92). Potentially eligible participants providing consent to be contacted were approached by the researcher (n = 148), of whom 105 individuals providing consent and meeting eligibility criteria were randomised. Of the 54 individuals randomised to the intervention (51 to TAU), five withdrew. The follow-up completion rate at 12 months was 88.6%.

Results found no significant effects between the intervention and TAU group. However, the effect of the intervention was larger (effect size 0.54, not significant) in 15 (28%) intervention and 10 (20%) TAU participants who were taking olanzapine or clozapine at randomisation.

Discussion

We found that our healthy-living intervention had a small but non-significant reduction in BMI compared with TAU. We found that in those participants who were taking olanzapine or clozapine at randomisation, there was a larger but non-significant effect. We successfully recruited to target and achieved an 89% retention rate at 12-month follow-up rate. Of the intervention participants, 78% completed 6–8 sessions. Overall costs of health and social care services used were lower but not statistically significant in the intervention group, compared with the TAU group.

Qualitative acceptability interviews

Introduction

In this stage of the research we aimed to access the psychological factors associated with substance use in first-episode psychosis, specifically the factors that enable or present barriers to stopping using substances. We investigated reasons for substance use; why people stop or do not stop using substances; and why some people start to use substances again following a period of abstinence. The exploratory nature of this work made a qualitative design ideal. We chose the grounded theory approach as our aim was to generate or discover a theory114 of substance use in first-episode psychosis. Our ultimate aim was to use the theory generated from this qualitative enquiry to aid in the development of a phase-specific intervention for substance use in early psychosis.

Methods

Participants were recruited from an early-intervention service based in the north-west of England between January and September 2008. Purposive sampling was used to achieve variation in age, gender and ethnic background to ensure that the sample would reflect the whole service population. Inclusion criteria were:

• acceptance into the early-intervention service for treatment of psychosis (i.e. not including patients from the ‘high-risk’ group)

• a period of 3 months when the client used substances on at least 2 days of each week for half of the weeks in the 3-month period. This 3-month period will be either at present or in the past, as far back as 6 months before the first contact with the early-intervention service

• working knowledge of the English language

• aged > 16 years.

Participant eligibility was checked using a substance use checklist that detailed use of substances (amount and frequency of use) over the preceding 3 months and the substance use modules of the Structured Clinical Interview for DSM-IV Axis 1 Disorders (SCID). 23

Interviews took place once eligibility was confirmed and informed consent obtained. Interviews were conducted at a place chosen by the participant, typically their own home. The main interview was topic guided (see Appendix 6) and lasted between 60 and 90 minutes. The study followed an iterative approach with the topic guide developing as the study progressed to reflect the developing inductive theories. This practice is in line with a grounded theory approach. 115 All interviews were digitally recorded and transcribed verbatim.

Analysis

The qualitative data were analysed by a team comprising two academic researchers, three clinical academics with psychological perspectives, one specialist nurse therapist and a non-clinical academic with expertise in qualitative methodology. One member of this analysis team had previous experience of using mental health services and of substance misuse.

Interview transcripts were read and preliminary coding was conducted concurrently with data collection. To ensure transparency and reliability, all transcripts were read and analysed by the interviewer and coded by at least two members of the team. The analysis process was managed using NVivo software (version 9; QSR International, Warrington, UK). Team members discussed coding and their interpretation of the transcripts in detail to refine codes and identify key themes emerging from the data. The original transcripts were then reread, coded and indexed. The themes highlight that which is shared by interviewees, but we have also acknowledged data that are at odds with these themes. 116 The team approach to analysis allowed inconsistencies between the data and themes to be debated, refined and reflected in the final presentation of the main themes.

Results

The therapeutic model

The heuristic model underpinning the intervention (Figure 7) was also developed for use in the MIDAS trial. The model combines aspects of Marlatt and Gordon’s social cognitive theory of problematic substance use119 with elements of Blanchard and colleagues’ affect regulation model. 120 The model, shown here as applicable to people with co-occurring psychosis and cannabis use, proposes that certain situations and cues trigger cannabis use-related thoughts, which, in the absence of alternative strategies and in the context of low self-efficacy for resisting use and positive expectancies from use, make the person vulnerable and more likely to use cannabis. This interaction between situations and cognitive/emotional reactions becomes the basis of a repeated cycle that maintains the cannabis use.

FIGURE 7.

The therapeutic model.

Within this context, poor problem-solving abilities and limitations in obtaining pleasure in ways other than through cannabis use increase the likelihood of the learned expectations of the positive benefits of repeated cannabis use being reinforced. We have found in clinical practice that the contributing role of cannabis use to psychosis may not be as significant to the individual as to other concerned people, that is family or mental health services. Rather, cannabis use is likely to be associated with more positive than negative expectancies, leading to a situation in which cannabis use is maintained. Such perceived positive benefits may include, for example, mood enhancement, relief from boredom, anxiety reduction, opportunities for social interaction with other cannabis users, ‘time out’ from problems, the blocking out of unpleasant thoughts, general feelings of well-being, release from unpleasant self-consciousness and performance enhancement through physiological effects on arousal or other mechanisms.

Using the model as a template facilitates the therapist in completing a detailed functional analysis of the links between the person’s psychosis, associated difficulties and continued cannabis use. To achieve this, the therapist elicits information at distinct stages of the therapeutic intervention. The model assists the therapist to elicit self-motivational statements from the individual regarding the negative aspects of continued substance use and to then feed this back through a shared understanding.

Methods

Results

Participants

A total of 325 patients were identified as being potentially eligible. Of these, 138 declined to be screened for eligibility and 43 did not meet eligibility criteria; 144 consented to take part, of whom 34 withdrew consent before randomisation; 110 completed all baseline assessments and were randomised. Of these, 38 were allocated to brief therapy, 37 to long therapy and 35 to TAU. Data on the primary outcome were collected for 83 (75.5%) participants at 4.5 months, 79 (71.8%) participants at 9 months and 76 (69.1%) participants at 18 months. Full data on relapse and hospitalisation were available for 108 participants (98.2%). Figure 8 shows the participant flow through the trial.

FIGURE 8.

Consolidated Standards of Reporting Trials (CONSORT) flow diagram showing participant flow through the trial. DNA, did not attend.

Tables 7 and 8 present baseline demographic, psychiatric and substance use information for participants in each trial arm. Participants were mostly male and aged in their mid-20s. The majority were unemployed. They had a history of psychosis of < 18 months and had been using cannabis for around 10 years, having begun using cannabis in their early teenage years. The majority (n = 100, 90.9%) met criteria for cannabis use dependence. Participants had been using cannabis on 59 out of the 90 previous days on average (SD 27.5 days), consuming an average of 107 g of cannabis in the same time period. The average amount of cannabis per cannabis-using day was 1.7 g (SD 1.5 g). The majority of participants were smoking mostly skunk (78.2%), with 20% smoking mostly resin and the remainder (1.8%) smoking equal amounts of each. More than four out of five participants were contemplating changing their cannabis use at entry to the study according to the RTCQ, with 53.2% identified as being at the ‘contemplation’ stage and 31.2% at the ‘action’ stage.

TABLE 7 - Baseline demographic, psychiatric and substance use variables in the three treatment arms

NOS, not otherwise specified.

Variable	TAU (n = 35)	Brief therapy (n = 38)	Long therapy (n = 37)
Age (years), mean (SD)	23.4 (3.8)	24.9 (5.6)	24.1 (5.4)
Gender: male, n (%)	30 (85.7)	34 (89.5)	34 (91.9)
Living arrangements, n (%)
Alone/house-share/hostel	15 (42.9)	14 (36.8)	15 (40.5)
With partner or family	20 (57.1)	24 (63.2)	22 (59.5)
Ethnicity, n (%)
White	33 (94.3)	35 (92.1)	34 (91.9)
Black and minority ethnic	2 (5.7)	3 (7.9)	3 (8.1)
Attended higher education, n (%)	21 (60.0)	16 (42.1)	19 (51.4)
Employment, n (%)
Unemployed/retired	27 (77.1)	30 (78.9)	32 (86.5)
Employed/self-employed	4 (11.4)	2 (5.3)	2 (5.4)
Student	4 (11.4)	6 (15.8)	3 (8.1)
History of psychosis (months), median (range)	17.2 (2.3–57.0)	13.4 (1.4–59.6)	17.5 (1.8–62.8)
DUP, n (%)
< 4 months	17 (54.8)	13 (37.1)	10 (30.3)
> 4 months	14 (45.2)	22 (62.9)	23 (69.7)
Adherent to medication (DAI), n (%)	25 (71.4)	30 (78.9)	30 (81.1)
Diagnosis (SCID), n (%)
Schizophrenia	18 (51.4)	20 (52.6)	16 (43.2)
Schizophreniform disorder	3 (8.6)	3 (7.9)	3 (8.1)
Schizoaffective disorder	3 (8.6)	5 (13.2)	5 (13.5)
Delusional disorder	3 (8.6)	2 (5.3)	4 (10.8)
Substance-induced psychosis	3 (8.6)	2 (5.3)	1 (2.7)
Psychotic disorder NOS	5 (14.3)	6 (15.8)	8 (21.6)
PANSS score, mean (SD)
Positive	14.9 (3.1)	15.4 (4.5)	14.8 (4.7)
Negative	14.1 (5.4)	15.1 (3.8)	13.0 (4.9)
General	32.7 (6.8)	35.6 (7.1)	33.8 (7.2)
GAF score, mean (SD)	37.9 (9.0)	35.1 (7.2)	39.0 (10.5)
CDSS score, mean (SD)	5.7 (5.5)	7.7 (4.6)	7.3 (3.9)
Anxiety (BAI), mean (SD)	14.8 (10.9)	17.1 (11.7)	20.3 (12.8)
Relapsed (9 months pre baseline), n (%)	18 (51.4)	16 (42.1)	17 (45.9)
Admitted (9 months pre baseline), n (%)	7 (20.0)	6 (15.8)	10 (27.0)
History of cannabis use (years), mean (SD)	9.0 (4.3)	10.3 (5.3)	10.4 (5.5)
Substance use disorder (SCID), n (%)
Cannabis abuse	2 (5.7)	5 (13.2)	3 (8.1)
Cannabis dependence	33 (94.3)	33 (86.8)	34 (91.9)
Alcohol abuse	3 (8.6)	2 (5.3)	2 (5.4)
Alcohol dependence	1 (2.9)	2 (5.3)	0 (0)

TABLE 8 - Frequency and amount of substance use at baseline (previous 90 days)

Max., maximum; min., minimum.

Variable	TAU (n = 35)	Brief therapy (n = 38)	Long therapy (n = 37)	Comparison
	Median (min.–max.)	Median (min.–max.)	Median (min.–max.)	χ2, p-value
Proportion of days abstinent from cannabis (%)	42.2 (0–88.9)	21.1 (0–92.2)	26.7 (0–6.7)	1.4, 0.50
Total cannabis use (g)	87.8 (7.5–322.5)	108.3 (1.2–640.0)	42.8 (1.0–377.4)	4.4, 0.11
Average amount of cannabis used per cannabis-using day (g)	1.3 (0.3–9.4)	1.5 (0.1–7.1)	1.2 (0.1–4.2)	5.3, 0.069
Proportion of days abstinent from all substances (%)	31.1 (0–84.4)	13.9 (0–82.2)	22.2 (0–81.1)	1.4, 0.49

Discussion

We have conducted the largest trial to date of integrated MiCBT in a first-episode cannabis-using population. Our aim was to evaluate the effectiveness of MiCBT and compare the efficacy of a brief intervention (4.5 months) with that of a longer one (9 months), hypothesising that the MiCBT intervention would be efficacious in reducing substance use in people with early psychosis and that the brief version would be as efficacious as the longer version.

The outcomes of the trial are not yet available and we have therefore outlined the trial, described the sample at baseline and reported recruitment and retention rates. Indications are that despite the challenges associated with conducting trials with this client group the study has been successful in recruiting and retaining participants over a long follow-up period. We worked closely with the early-intervention teams to set up the study and identify potentially eligible participants and employed assertive engagement strategies once participants were recruited, seeing them in their own homes, arranging appointments flexibly but with persistence and conducting repeat visits for missed therapy and research appointments when necessary.

The refusal rate for participants identified as potentially eligible to take part was higher than those in recently conducted RCTs that had shorter intervention and follow-up periods109–111 but was in line with rates obtained from previously conducted studies. 107 We note that this was a non-treatment-seeking sample and all service users identified by their case managers as cannabis users were considered for inclusion. We do not have information on the service users who refused to participate or to be screened for eligibility and therefore cannot compare them with those included in the study. This may reduce the generalisability of our findings. It may be that those who declined the invitation to participate were more likely to be at a pre-contemplation stage of change. We found that > 80% of those randomised were contemplating changing their cannabis use at baseline. It is logical to assume that those who were contemplating change were more likely to agree to take part than those who were not.

Our assertive engagement strategies resulted in good retention for follow-up assessment. Almost 70% of the sample was retained at 18 months’ follow-up. Many of those who were lost to follow-up had moved away without leaving a forwarding address and were, therefore, not contactable. Take up of therapy was satisfactory, with 80% attending at least two sessions. Brief therapy was better attended than long therapy with twice the number of brief therapy participants attending 75% of the sessions offered. The median number of therapy sessions attended was similar in the two therapy arms, indicating that a brief form of the intervention may be more acceptable than a longer form. Qualitative interviews with therapy participants will seek to establish why brief therapy was better received and why some people refused therapy when offered it.

The sample is comparable with those in the four earlier RCTs of interventions with young cannabis-using recent-onset psychosis patients,106–109 three of which had not reported their results at the time that the ReCAP trial commenced,107–109 and to early-intervention service users generally: the majority were male, in their 20s and unemployed and had been using cannabis since their mid-teens. We recruited from services in various geographical locations encompassing both urban and rural populations and we are, therefore, confident that the sample is representative of our target population and that the results will be generalisable. We note that frequency of cannabis use, measured by percentage days abstinent, was higher in this trial than the previously reported trials with regard to percentage days abstinent. The amount of cannabis consumed is difficult to compare between the trials because of methodological differences, but ReCAP participants appeared to be smoking approximately 50% more cannabis than CapOpus participants. 108 Rates of abuse/dependence on other substances were much lower in the ReCAP trial than in the CapOpus trial (< 10% vs. approximately 50%), which may indicate that our sample was to some extent a ‘purer’ sample of cannabis users and more in line with the study by Bonsack and colleagues,107 which excluded those who were dependent on other substances and purposely recruited ‘heavier’ cannabis users. Rates of abuse and dependence on other substances were not reported in the other two trials and comparisons cannot be made.

The three previously conducted trials of integrated MiCBT in early psychosis106,108,109 found that the intervention did not confer an advantage over the control conditions in terms of reducing cannabis use or improving clinical outcomes. The results of the ReCAP trial will, therefore, be an important addition to the evidence base; with the longest intervention of all of the trials and the longest follow up, the results will be crucial in determining whether MiCBT can reduce cannabis use or a new therapeutic approach to cannabis use in psychosis will be required. It is perhaps noteworthy in this context that none of the recent trials, including the ReCAP trial, has involved families, possibly because of the constraints on recruitment of selecting only those who have regular contact with relatives. However, two studies that have found benefits from psychological therapy for people with psychosis and substance use have included a family-based component. 134,135

The strengths of this study include the high-quality randomisation process, the single-blind design, the range of substance use, symptom and functional outcomes and the 18-month follow-up period. The majority of participants engaged with the intervention and we were able to follow up a high proportion of participants. However, a significant minority (one in five) did not engage well with the intervention despite our assertive approach and future research should seek to understand influences on engagement in therapy as well as treatment preferences overall. Limitations include the potential bias resulting from early-intervention service users under-reporting substance use to their case managers and therefore not being considered eligible for referral; likewise, some potential participants may not have perceived their cannabis use to be problematic and declined to take part. Nonetheless, as the largest RCT conducted to date of an intervention to reduce cannabis use in first-episode psychosis and the only trial to compare brief therapy with longer therapy, we expect the results of this trial to be of considerable interest to academics and clinicians concerned with the development of an evidence base for the treatment of this client group.

First qualitative interview schedule prompt sheet

1. Can you tell me when your mental health problems first started?

   Probe

   • What do you remember about the onset of your problems?

   • What symptoms were you experiencing?

   • What did you think was happening to you at the time?

   • How did you react initially to experiencing symptoms?

   • What do you think caused the problems that you have experienced?

   • How is life different now to before your mental health problems started?

1. What sort of things have you noticed make your mental health problems either better or worse?

   Probe

   • Do you think that there is anything that you or anyone else can do to help control your symptoms?

   • Can you identify any particular situations/triggers that make your symptoms worse?

   • How do you cope with these situations?

1. What if any medication are you taking at the moment?

   Probe

   • Can you tell me the name and dosage of your medication?

   • How long have you been taking this medication?

   • Do you manage to take your medication exactly as it is prescribed?

   • How do you make sure that you remember to take it?

   • Do you feel that taking the medication helps you in any way?

   • Is there anything that you don’t like about taking medication?

1. If not taking medication

   Probe

   • Why do you not take medication? Is it because you have never been asked to or is it a decision on your part?

   • How would you feel about taking medication to help your mental health problems if your doctor asked you to?

1. What information have you been given regarding medication?

   Probe

   • What information have they been given?

   • Who gave them this information/how did they obtain it?

   • Do they know about any potential side effects?

   • Do they experience any unpleasant side effects, how much do these bother them and how do they cope with them?

1. What sort of things do you think are important for maintaining a healthy lifestyle?

   Probe

   • What do you do at the moment to stay fit and healthy?

   • Have there been any changes to your lifestyle since you first experienced mental health problems?

1. Has your lifestyle changed in any way since you first developed mental health problems?

   • Tell me a bit about the types of things you like to eat

     • Do you eat regular meals at set times or are you more likely to graze on snacks between meals?

     • Can you tell me how many portions of fruit or vegetables you eat in a typical day?

     • How often do you eat fried food?

     • What type of bread do you prefer?

     • How much coke or how many other soft drinks do you have in an average day?

   • Do you smoke cigarettes?

     • If yes, how many in a typical day?

     • How long have you smoked for?

   • Do you drink alcohol?

     • If yes, how much do you drink on an average day?

   • Do you take any drugs that haven’t been prescribed by a doctor, such as cannabis?

     • If yes, what do you take and how much in a typical day/week?

1. Do you consider that you have any problems with your weight?

   Probe

   • What do you consider these problems to be?

   • Have they ever tried to lose/gain weight?

   • How did they do this and how successful was it?

1. Would you like to make any changes to your current lifestyle in order to make it healthier?

   Probe

   • What changes if any would you like to make?

1. What help or advice (if any) would you like to have in order for you to make healthy changes to your lifestyle?

   Probe

   • Who do you think is the best person to give this advice, e.g. CPN, GP, practice nurse?

   • Where would they like to receive advice, e.g. hospital, GP surgery?

1. What type of healthy living intervention would you like to receive?

   • If free access to a gym or other fitness facility was made available do you think that you would use it?

   • Would you prefer to go with a group of people or alone?

   • Where do you think the best place to deliver the intervention would be (probe – mental health setting, community setting, other)?

   • What time of day would you prefer to attend a healthy living group (probe – afternoon, morning, early evening; fixed time, flexible times; how long, frequency of sessions; rationale for choice)?

   • What you think would be the best way to run the intervention (probe – explicitly here, i.e. games, live demonstrations, e.g. healthy eating, workbooks, small group work, discussion, bringing in some experts, e.g. dieticians, hands on, e.g. practical such as an actual cooking class; open or closed group)?

   • Who do you think are the best people to deliver the intervention (probe – MHW, users, experts in health promotion, etc.)?

   • What do you think are the things that will make the intervention successful in helping people to develop a healthier lifestyle (probe – free access to exercise facilities, etc.) and what things might get in the way of it being successful [probe – reasons and solutions, i.e. what do you think might help us overcome these (barriers) or how can we make sure we do this (facilitators)]?

1. Is there anything else that you would like to tell me about the service you are receiving? Or anything about the type of healthy living intervention that you would like to receive?

Thank you for giving up your time to see me today. It has been very helpful talking to you.

Interview/focus group schedule for early intervention services/primary care staff

1. Introduction and presentation/information sheet on synthesis of our findings from phase 1 (i.e. content elements).

2. What are your views of your roles and responsibilities in the physical health care of people with psychosis (probe – reasons, i.e. policy, professional, etc.)?

3. What training have you received in providing health education advice to service users about healthy living (probe – adequacy of training; mandatory or voluntary)?

4. What training have you received in assessing and managing physical health problems (probe – adequacy of training; mandatory or voluntary)? Do you routinely assess for unwanted effects of medication (probe – how do they do this, what side effects do they assess for)?

5. Would you be prepared to run healthy living groups within the early intervention service (probe – exactly who, resource implications, training, etc.)?

6. If healthy living groups were to be delivered by the early intervention service, what would help them to run smoothly (probe – venue, group leader, time of day)?

7. Do you think there might be any obstacles to the healthy living groups running smoothly in the early intervention service (probe – pressure of work, service user apathy, not a priority, etc.)?

Introduction

• Explanation of ethics, consent and confidentiality of interview and analysis.

• Explanation of objective of acceptability substudy.

• Structure and duration of the interview.

• Any questions?

Clarify – talking about intervention rather than the research.

Context

• Tell me about the healthy living intervention?

• Can you tell me about your experiences while taking part in the healthy living intervention? Probe – Individual and groups or individual only? If individual only why not groups?

Process

• What were your initial expectations of the healthy living intervention?

• What were the specific advantages and disadvantages of the healthy living intervention?

  Probe

  • Time commitments, timing of appointments, travel, etc.

  • (What specific thing achieved change?)

• What specifically did you like/dislike about the healthy living programme?

  Probe

  • What did you think about the goal-setting and action plans?

  • How realistic were the goals you set?

  • How well did you feel you achieved your goals.

  • If you did the programme again, would you do anything differently?

  • (What specific element/thing achieved change?)

• Were you provided with information about the website?

  Probe

  • Did you use it? If not, why not?

  • How many times did you go on to it?

  • Did you find it useful?

  • How did you find getting access to the internet?

  • Did it help with the intervention?

• Were you provided with information about a booklet to help with the intervention?

  Probe

  • Did you use it?

  • How many times did refer to it?

  • Did you find it useful?

  • Did it help you achieve your goals?

• Which would/did you prefer, the booklet or the website? Why?

• STR worker – how did you get on with your STR worker? Good points? Things that could be improved?

  Probe specific intervention factors (i.e. goal-setting, action plans, implementation of action plan, interventions and non-specific factors such as warmth, empathy, listening)

  • Some people have said that it was useful for them that their STR worker knew about their past – what you think about this? Did you find this useful? Why?

  • Some people have said that they found it useful that their STR worker knew about their mental health whereas others did not – how do you feel about this?

• What was your experience of your family/carer being involved in your treatment?

  Probe

  • Were family involved? If so, what were the advantages and disadvantages of your family being involved? How helpful was it?

• Did the STR worker offer to see your family or other people involved in your care, for example close friends? If so, what was the outcome? How did you feel about them being involved?

Outcome

• Do you feel that you have benefited from being on the programme?

  Probe

  • Any specific benefits for healthy living or other benefits for functioning (social, occupational/school, private, leisure, family relationships), mood/depression, anxiety, stress, the way you feel about yourself, your appearance?

  • (What specific thing achieved change?)

• Would you recommend this programme to other people with similar problems to your own?

  Probe

  • Yes/no answers on reasons for their response.

  • (What specific thing would you recommend to someone about the programme or is it the programme as a whole?)

• Is there anything that you would add to or change about the intervention?

  Probe

  • For example, some people have said that a weekly weigh-in would be useful whereas others have said that they would not find this useful? What do you think?

• Is there anything else that you would like to say about the healthy living intervention?

Ending

• Thank participant for their time and information.

• Next steps: would they like a copy of the transcript to be sent to them for validity checking?

• A one-page copy of results will be sent to all participants.

Introduction

Cognitive Remediation (CR) is a method for improving neuropsychological function in schizophrenia. In essence the intervention is a form of ‘brain training’, consisting of regular practice on a number of mental puzzles. In clinical trials, CR has been shown to improve:

1. executive function (Kurtz et al. , 2001, 2004; Penades et al. , 2006; Reeder et al. , 2004; Twamley et al. , 2003; Wykes et al. , 1999, 2007);

2. attention (e.g. Kurtz et al. , 2001, 2004; Medalia et al. , 1998; Silverstein et al. , 2005, 2007; Twamley et al. , 2003);

3. long term memory (Kurtz et al. , 2001, 2004; Penades et al. , 2006; Twamley et al. , 2003; Wykes et al. , 1999);

4. insight into cognitive problems (Granholm et al. , 2005); and

5. positive symptoms (Reeder et al. , 2004).

Whilst previous meta-analyses have reached conflicting conclusions about the effect size of the intervention (Pilling et al. , 2002), more recent trials have produced increasingly positive results, as the intervention and its associated outcome measures have become more sophisticated and specific. Furthermore, there is evidence for combining CR with other interventions in trials of rehabilitation for the severe and long term mental illness (Bell et al. , 2007; Mueser & McGurk, 2007; Silverstein et al. , 2005, 2007).

We hypothesise that CR will enhance the efficacy of the cognitive behavioural therapy (CBT) by improving cognitive functioning, especially attention and executive functions, (such as set shifting). If so, we expect that those who receive CR will received greater benefit from CBT, by virtue of being able to acquire more complex skills and reaching a better understanding of their problems. We therefore predict that they will score lower on the Psychotic SYmptom RATing Scales (PSYRATS; Haddock et al., 1999), which measures the psychotic symptoms targeted by CBT. We also predict they will get better more quickly and score more highly on a scale that describes the quality of engagement in CBT, designed by Professor Gillian Haddock (personal communication). The scale is already being used in an MRC-funded trial of cognitive behavioural therapy and motivational interviewing for drug misuse conducted at the University of Manchester. Other secondary measures will include: other symptoms; neuropsychological function; cognitive insight; attitudes to illness; vocational achievement; and self-concept.

Methods

Analysis

The main outcome will be the PSYRATS. Curves will be fitted to the longitudinal data using mixed effects models estimated by maximum likelihood methods. Allocation group (to CR or befriending) and potential confounders will be independent variables and the data will be clustered by cognitive–behavioural therapist.

The pattern of any missing data (e.g. due to drop-out) will be analysed using logistic regression to test whether the assumptions underlying this type of modelling (i.e. data ‘missing-at-random’) are met.

The principal secondary outcomes will be number of sessions and CBT complexity score. Others, adjusted for multiple testing, will include neuropsychological measures, global illness severity & dysfunction, attitudes to illness, cognitions and self esteem. In the event of positive results the effect of including process measures related to neuropsychological variables and attitudes will be examined.

Power

Based on a t-test of final interview scores, with an alpha value of 0.05 and power of 0.80, 64 participants allows detection of a difference between groups of effect size d of 0.50 (PS v 2.1.31, Dupont & Plummer 1997; using methodology from Dupont & Plummer 1990). This difference represents 8.1 points in PSYRATS totals (based on data from a first-episode sample in Lewis et al. , 2002).

The same data-set shows that there is substantial independence between measurements made at successive points in follow-up six weeks or more apart. Thus one may make the approximation that data from the n = 6 follow-up interviews (at six week intervals; see Figure 1) will each contribute independently to estimating outcome, which in turn would reduce the standard deviation for the control mean across the repeated measures by 1/√n. This approximation reduces the minimum detectable difference of approximately 3.4 points in PSYRATS total (about 25% of the controls’ overall mean score for the repeated measures).

Unpublished data from a recent trial in 67 schizophrenia-spectrum psychosis sufferers of cognitive remediation combined with social cognition training compared to control supportive therapy (Eack and colleagues, in press) found benefits of d 0.48 for a composite neuropsychological measure similar to the tasks proposed here and d 0.51 for symptoms.

References

Addington D, Addington J, Atkinson M. (1996) A psychometric comparison of the Calgary Depression Scale for Schizophrenia and the Hamilton Depression Rating Scale. Schizophr Res. 19, (2–3) 205–12.

Bell MD, Greig TC, Zito B, Wexler BW (2007) An RCT of neurocognitive enhancement therapy with supported employment: employment outcomes at 24 months. Schizophrenia Bulletin, 33, (2) 420.

Couture SM, Roberts DL, Penn DL, Cather C, Otto MW & Goff D. (2006). Do baseline client characteristics predict the therapeutic alliance in the treatment of schizophrenia? J Nerv Ment Dis. , 194, (1) 10–4.

Denver Recovery Markers Inventory, (2006) Mental Health Centre of Denver, 4141 E Dickenson Place, Denver, CO 80222.

Devane SM, Haddock G, Lancashire S, Baguley I, Butterworth T, Tarrier N, James A & Molyneux P. The clinical skills of community psychiatric nurses working with patients with severe and enduring mental health problems: an empirical analysis (1998) J Adv Nursing, 27, 253–60.

Drake RJ, Corcoran R, Kaiser SL, Smallman R, Aslam M, Lewis SW (2007). A pilot study of CR for poor insight in non-affective psychosis: a minimized, controlled trial. Schizophrenia Bulletin, 33, (2) 428.

Dupont WD & Plummer WD: (1990) Power and Sample Size Calculations: A Review and Computer Program. Controlled Clinical Trials; 11:116–28.

Eack SM, Hogarty GE, Cooley SS, DiBarry AL, Hogerty SS, Greenwald DP, Montrose DM & Keshavan MS. Cognitive Enhancement Therapy for Early Course Schizophrenia: Effects of a Two-Year Randomised Controlled Trial. Schizophrenia Bulletin (in press).

Granholm E, McQuaid JR, McClure FS, Auslander LA, Perivoliotis D, Pedrelli P, Patterson T & Jeste DV. (2005) A randomised, controlled trial of cognitive behavioral social skills training for middle-aged and older outpatients with chronic schizophrenia. Am J Psych, 162, 520–9.

Haddock G, McCarron J, Tarrier N & Faragher EB. (1999). Scales to measure dimensions of hallucinations and delusions: the psychotic symptom rating scales (PSYRATS). Psychol Med. , 29, (4) 879–89.

Kay SR, Fiszbein A & Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. (1987). Schizophr Bull., 13, (2) 261–76.

Kurtz MM, Moberg PJ, Gur RC & Gur RE (2004) Approaches to cognitive remediation of neuropsychological deficits in schizophrenia: a review and meta-analysis. Neuropsychology Review, 11, 197–210.

Kurtz MM, Moberg PJ, Mozley LH, Swanson CL, Gur RC & Gur RE (2001) Effectiveness of an attention- and memory-training program on neuropsychological deficits in schizophrenia. Neurorehabilitation & Neurorepair, 15, 75–80.

Lewis S, Tarrier N, Haddock G, Bentall R, Kinderman P, Kingdon D, Siddle R, Drake R, Everitt J, Leadley K, Benn A, Grazebrook K, Haley C, Akhtar S, Davies L, Palmer S, Faragher B, Dunn G. (2002). Randomised controlled trial of cognitive-behavioural therapy in early schizophrenia: acute-phase outcomes. Br J Psychiatry Suppl., 43:s91–7.

Lobban F, Barrowclough C, Jones S. (2005). Assessing cognitive representations of mental health problems. I. The illness perception questionnaire for schizophrenia. Br J Clin Psychol. , 44, (Pt 2):147–62.

Medalia A, Aluma M, Tryon W & Merriam AE (1998) Effectiveness of attention training in schizophrenia. Schizophrenia Bulletin, 24, 147–52.

Mueser KT & McGurk S (2007) 2-3 year outcomes of a randomised controlled trial of cognitive training and supported employment for people with severe mental illness. Schizophrenia Bulletin, 33, (2) 448.

Penades R, Catalan R, Salamero M, Boget T, Puig O, Guarch J & Gasto C. (2006). Cognitive remediation therapy for outpatients with chronic schizophrenia: a controlled and randomised study. Schizophr Res. , 87, (1–3) 323–31.

Pilling S, Bebbington P, Kuipers E, Garety P, Geddes J, Martindale B, Orbach G and Morgan C (2002) Psychological treatments in schizophrenia: II. Meta-analysis of randomised controlled trials of social skills training and cognitive remediation. Psychological Medicine, 32, 783–91.

Reeder C, Newton E, Frangou S & Wykes T (2004). Which executive skills should we target to affect social functioning and symptom change? A study of a cognitive remediation therapy program. Schizophrenia Bulletin, 30, 87–100.

Silverstein SM, Berten S, Savitz A, Spaulding W & Menditto A (2007) Attention shaping procedures as a motivation-based intervention to improve attentiveness and skill acquisition in chronic schizophrenia patients. Schizophrenia Bulletin, 33, (2) 460.

Silverstein SM, Hatashita-Wong M, Solak BA, Uhlhaas P, Landa Y, Wilkniss SM, Goicochea C, Carpiniello K, Schenkel LS, Savitz A & Smith TE (2005). Effectiveness of a two-phase cognitive rehabilitation intervention for severely impaired schizophrenia patients. Psychol Med. , 35, (6) 829–37.

Twamley EW, Jeste DV & Bellack AS (2003) A review of cognitive training in schizophrenia. Schizophrenia Bulletin, 29, 359–82.

Wykes T, Reeder C, Corner J, Williams C & Everitt B. (1999). The effects of neurocognitive remediation on executive processing in patients with schizophrenia. Schizophr Bull. , 25, (2) 291–307.

Wykes T, Reeder C, Landau S, Everitt B, Knapp M, Patel A & Romeo R. (2007) Cognitive remediation therapy in schizophrenia: Randomised controlled trial. Br J Psychiatry. , 190, 421–7.

1.1 Programme title

The HELPER programme (HEalthy Living and Prevention of Early Relapse).

1.2 Stream title

Stream 3: A Healthy Living Intervention to encourage activity, improve diet, and control weight gain in people with a first episode of psychosis within the past three years.

1.3 Trial acronym

INTERACT Trial (INTERvention to encourage ACTivity, improve diet, and control weight gain).

2.1 Background

The life expectancy of adults with schizophrenia is reduced by about 15 years when compared with the general population (Hennekens et al. , 2005), and while some of this premature mortality is accounted for by suicide, approximately 62% of all deaths are attributable to natural causes (Harris & Barraclough, 1998). Adults with a diagnosis of psychosis are twice as likely as members of the general population to suffer from ischaemic heart disease, stroke, hypertension, epilepsy or diabetes by the age of 55 years (Hippisley-Cox & Pringle, 2005).

This poor physical health may be explained both by the side effects of antipsychotic medication (Marder et al. , 2004) and by the unhealthy lifestyles that many individuals with psychosis lead (Brown, Inskip & Barraclough, 2000). People with psychosis have been shown to take less exercise (McCreadie, 2003), eat poorer diets (McCreadie et al. , 1998) and to be significantly more likely to smoke (Brown et al. , 1999) than members of the general population. They are more likely to be obese (Homel et al. , 2002) and show a higher prevalence of metabolic syndrome (Sacks, 2004) and diabetes mellitus (Bushe & Holt, 2004). Patients taking second-generation antipsychotics, which are frequently prescribed to control symptoms of psychosis, are at particular risk of weight gain (Green et al. , 2000). The risk of antipsychotic induced weight gain may be even higher among young people experiencing a first episode of psychosis (Zipursky et al. , 2005). Recent guidance from the Department of Health (2006) has therefore highlighted the need to develop and evaluate health education programmes for people who have recently had a first episode of psychosis.

2.2 Development of the intervention

In deciding on the form of our intervention, which will be a pragmatic Phase II exploratory trial (Craig et al. , 2008) we carried out four pieces of work (collectively termed Phase 1 work), and then synthesised the findings. Firstly, we reviewed studies already in the literature, focussing on controlled interventions. Secondly, we carried out two qualitative studies, one with service users and one with case managers, to discover their views on the content, format, setting and mode of delivery of the proposed intervention. Thirdly, we reviewed the key features of Leventhal’s self-regulatory model in order to provide a theoretical underpinning for our intervention. Finally, we considered how to make our intervention culturally sensitive for delivery in a multi-cultural environment.

A systematic review and meta-analysis of ten non-pharmacological interventions to manage antipsychotic induced weight gain (Álvarez-Jiménez et al. , 2008) concluded that healthy living interventions for this population are effective, although the effect size was rather small, with an average reduction of about 4% of body weight. Only two of the interventions reviewed by Álvarez-Jiménez et al. were for first-episode psychosis, many of the studies had small samples, and some failed to describe their interventions adequately. Our own systematic review (HELPER team, 2008) of 12 trials, focussed on reduction in BMI, which we would argue is a better outcome measure than weight as there is less heterogeneity. We concluded that while some interventions have been effective, more work is needed to determine the ideal content of interventions. Analysis of group versus individual interventions suggested an advantage for individualised interventions, and evidence is emerging that interventions with a supervised exercise component may be superior to those in which exercise is merely prescribed or is absent. There is little evidence on long-term effects of the interventions, but what evidence there is suggests that the difference between treated and control participants tails off after about 2 months, suggesting a need for booster sessions and longer term follow up. A final observation is that none of the interventions in the literature to date was explicitly informed by any theoretical framework or model of behaviour change. We concluded from our literature review that there is a need for a phase-specific intervention to reduce BMI; that at least some components of the intervention should be individualised; that the intervention should include supervised exercise; that there should be post-intervention booster sessions; and that there should be long term follow up.

According to Blackburn (1995), health benefits can be seen with a body weight reduction of 5% or more, so this should be the minimum goal for any intervention. For a man aged 16–24 of UK average height (177 cm) (Health Survey for England 2006, Latest Trends) with a BMI in the overweight but not obese range of 27 (weight 84.6 kg), a loss of 5% of body weight (to 80.36 kg) would equate to a reduction in BMI of 1.35 points. We therefore decided that, in order for our intervention to have clinically meaningful effects, we should aim for a reduction in BMI of at least 1.5 points.

The most important features to emerge from our interviews with service users were that there is a general willingness to undertake a healthy living programme, that the intervention must be accessible in terms of both travel and cost, but that the actual place of the intervention was of less importance. Some service users would prefer a group intervention, with social activities, while others would prefer individual sessions. Mainly for reasons of time constraints, case managers were not enthusiastic about delivering the intervention themselves, and felt that whoever delivered the intervention would need additional training. There was agreement that the intervention needed to target both diet and exercise and that it needed to be enjoyable and client-centred. In order to further tailor the intervention to the views of service-users, we have decided to use people with lived experience of early psychosis as facilitators for some of the group activities.

Leventhal’s self-regulatory model (Leventhal, Nerenz & Steele, 1984) states that when people encounter a threat to their health (such as the threat posed by weight gain in first-episode psychosis) their behavioural response to that threat is generated by their personal perceptions or model of the threat. These personal perceptions are formed from both concrete perceptual experiences (such as previous experience of the health threat) and from abstract information (such as information from health care professionals and health education) and, if maladaptive, are more amenable to change if information comes from a both concrete and abstract sources. Threats to health may be actual (such as a current illness) or future risks (such as the threat posed by weight gain).The self-regulatory model suggests that it is important to take into account people’s personal models when prescribing (or developing collaboratively with patients) treatment programmes aimed at reducing health threats. Empirical work over the years since the model has developed has shown that health threats are conceptualised along a number of underlying dimensions, which can now be measured (Moss-Morris et al. , 2002), and used to inform individualised treatment programmes. Furthermore, recent work has started to adapt these established measures to enable the assessment of risk perceptions of potential threats to health (Cameron, 2008; Figueiras & Alves, 2007), and the intention to adopt preventive behaviours. The model prescribes the setting of goals, the use of well-specified action plans, and emphasises the roles of appraisal and feedback in modifying perceptions and behaviours. All of these elements are to be incorporated into our intervention.

Our intervention will take account of ethnic differences in diabetes and cardiovascular disease risk and of cultural factors with the potential to impact on both exercise behaviour and dietary change. There is evidence that BMI cut-offs for overweight, obesity and health risk are different in South Asian and other populations. Specifically, members of the South Asian population begin to be at risk of diabetes and cardiovascular disease at a lower BMI (≥ 24) than do members of the general population in the UK (Huxley et al. , 2008; Obesity in Asia Collaboration, 2007) and we will take account of this when determining our entry criteria and when evaluating progress during the intervention. Exercise prescriptions will be agreed with the participants, and participants will be invited to involve family members in their exercise programme should they so wish. Dietary recommendations will take account of cultural and ethnic differences in diet; dietary recommendations will be based on food groups and will be made in consultation with dieticians experienced in working with culturally diverse populations. We will endeavour to make our intervention materials relevant, comprehensible and acceptable to participants from different backgrounds. After the intervention, we will seek the views of participants on its acceptability.

2.3 Summary

To summarise, our Phase 1 work has led us to design a twelve-month healthy living intervention for patients who have experienced their first episode of psychosis in the past 3 years, who are attending an early intervention service, and who have a BMI of ≥ 25 (24 for people of South Asian origin). The intervention will focus on increasing activity and improving diet, and will:

• aim to reduce BMI by 1.5 points at six months after the start of the intervention and to maintain this loss at the one year follow up;

• consist of a mixed programme of individual and group sessions delivered over the first six months (the intervention proper);

• provide a booster session in the second six months to reduce the risk of relapse;

• contain an exercise component;

• be delivered by a support time recovery worker (STR worker) who will utilise community resources and other professionals already working with this population, and will be assisted by service-users or ex service-users acting as group facilitators;

• tailor treatment to patients own personal perceptions of weight gain and the risks associated with it;

• be adapted to take account of cultural factors.

5.1 Inclusion criteria

The following people will be eligible for the trial:

1. aged 16 to 35 years

2. with a diagnosis of schizophrenia, schizophreniform disorder, schizoaffective disorder, delusional disorder, brief reactive psychosis, or psychosis not otherwise specified, determined using a checklist of criteria and review of case notes (OPCRIT http://sgdp.iop.kcl.ac.uk/opcrit/)

3. first episode of psychosis occurred within the three years preceding the trial

4. current user of an early intervention service

5. have stable accommodation (i.e. not street homeless or roofless)

6. able to give informed consent

7. have a BMI of ≥ 25, or a BMI of ≥ 24 for service users from the South Asian community (Huxley et al. , 2008; Obesity in Asia Collaboration, 2007)

5.2 Exclusion criteria

People will be excluded from the INTERACT trial if they:

1. have a diagnosis of substance dependence or abuse as determined from a review of case notes. Sub-diagnostic levels of substance use or abuse will not exclude people from the trial.

2. have a significant history of organic factors implicated in the aetiology of psychotic symptoms.

8.1 The active intervention

The INTERACT intervention will comprise of 8 individual sessions over a 12 month period with an emphasis on facilitating participatory exercise and dietary change through the development and implementation of patient led action plans. These sessions will be delivered by STR workers who will be trained in the delivery of the intervention. The STR workers may also be able to access other workers (e.g. occupational therapists) as a resource for the intervention. To facilitate implementation of exercise and dietary change a range of optional active group sessions (e.g. football, walking, netball groups, and a cooking group) will be offered by the STR worker for those who prefer a group based activity. The STR workers will be assisted in the running of the groups by service user or ex-service user facilitators. To optimise engagement, choice and self management a booklet will be given to all participants providing educational advice, action plans, goals, details of the group sessions, healthy eating recipes on a budget etc. A strong emphasis will be placed on maximising carer/family engagement.

8.2 Standard Care

The EIS works individually with each service user and his or her family to address problems/needs that are identified during his or her assessment/reassessments; all service users have enhanced care coordination and all should have a specific care plan personalized to their needs. Each service user would be seen on a regular basis; this may be weekly, fortnightly or more or less frequently by his or her case manager depending on the stage of the intervention and the need.

The case managers are supported by support time recovery workers who are commissioned and supervised by the case managers to do specific work, often for time limited periods. The service users are also offered cognitive behaviour based interventions by case managers. Case managers may consult with and obtain input from psychological therapists/psychologists who may themselves also be involved in delivering formal CBT.

9.1 Timing and administration

There will be three assessment points: Baseline assessment prior to randomisation into the trial; at 6 months; and final assessment at the end of one year. The primary outcome point will be at one year.

Assessments will be administered by the trial research assistant(s). While the majority of measures are self-report measures, in accordance with a blinding protocol every effort will be made to keep research assistants blind to randomisation codes so that post-intervention and long-term follow up assessments can be carried out blind to treatment allocation. Any unblindings will be recorded, and we will the degree of blindness maintained at the end of the trial (see Section 7 above).

9.2 Measures

The primary outcome will be change in BMI at one year post-baseline. This will be determined by measuring the participants’ height at baseline assessment and weighing the participant using the same set of scales at each assessment. The formula weight in kg/height in metres² will then be applied to determine BMI. The intervention will be deemed successful if a patient has reduced his or her BMI by at least 1.5 points at one year follow up.

Secondary outcomes will be: weight loss at 6 months (ie end of the intervention proper); change in waist circumference, level of daily activity as measured by a standard activity level measure which has been validated for use in this population (short form IPAQ, Craig et al. , 2003; Faulkner et al. , 2006); adherence to the agreed exercise programme; relapse of psychosis, as determined from a review of case notes; health status and quality of life as measured by the EQ-5D (Rabin & de Charro, 2001) and the MOS SF-36 (Ware & Sherbourne, 1992); food frequency diary, with cultural adaptations (Crozier et al. , 2008).

Measures of predicted process variables and mediators of change will be; a measure of representations of weight gain and intentions to take preventive/remedial action (Cameron, 2008; Figuieras & Alves, 2006); a measure of perceptions of antipsychotic medication (Fialko et al. , 2008);. Depression will be measured using the Calgary depression scale (Addington et al. , 1996) as a potential predictor of response to the intervention; reduction in depression may also be an outcome of interest. We will also measure adherence to medication (Brief Adherence Rating Scale, BARS; Byerly et al. , 2008) and keyworker rating of adherence to medication on four point scale (Fialko et al. , 2008).

9.3 Schedule of assessments and measures

9.4 Training and supervision of research assistants

Our research assistant(s) will be trained to administer the outcome measures in a standardized way. They will receive regular debriefing and supervision which will focus both on the integrity of the assessment procedures and on any health and safety or other issues arising from the assessments.

9.5 Risk assessments and health and safety

Risk assessments will be carried out prior to each assessment visit, in accordance with the procedures of the Early Intervention Services from which the patients are recruited. No RA will visit any patient in the absence of an up-to-date risk assessment. In order to maintain blindness, RAs will not have direct access to risk assessments.

12.1 Data entry and checking

Data spreadsheets will be set up by IT staff at the Mental Health Research Network. Data will be entered into data spreadsheets (one for each time point) by research assistants as they collect it. Data entry will be centralised using PsyGrid software so that RAs from each of the three HELPER trials can enter data online onto a central server. The primary outcome measure (weight and height before and after the intervention) and secondary outcome measures (weight at one-year follow up, activity levels, measure of quality of diet) will be double entered by two research assistants. Other data will be checked and cleaned by the research assistant before analysis.

12.2 Data protection and confidentiality

Paper copies of questionnaires will be stored in locked filing cabinets at the University of Manchester. Details of patient randomisation will be stored separately from all other data and password protected.

Monitoring arrangements

12.3 Analysis

15.1 Ethical approval

Ethical approval will be sought from the appropriate NRES committee.

15.2 Insurance and indemnity

A procedure to ensure the health and safety of trial workers will be appended to the protocol.

15.3 Trial management and oversight

Members of the HELPER INTERACT research team will meet monthly, and will report to the HELPER Programme Management group at their three-monthly meetings.

There will be two independent committees overseeing the work of the entire HELPER programme – a Trial Steering Committee (TSC) and an independent Data Monitoring and Ethics Committee (DMEC).

The DMEC will consist of members external to the study team including a statistician, a clinician and an expert in health services trials. A DMEC report template will be devised for reporting purposes and agreed by the DMEC committee prior to the commencement of the study.

Planned trial interventions

The integrated MiCBT will match the focus of therapy to the patient’s stage of change regarding cannabis use.

For all clients initially substance use will be approached within the context of the client’s lifestyle and mental health problems, and the intervention will be described as an approach to identifying helpful lifestyle changes (too much of an emphasis on cannabis use in the initial stages of the intervention may be detrimental to the motivational processes in clients not ready for change). With the client’s permission, later sessions will focus on examining the role of cannabis and how it relates to problems and desired goals. Using the motivational interviewing style, the therapist will seek to understand the client’s frame of reference and build up a shared model of how cannabis relates to their mental health and other life concerns. Once client commitment to change is secured, the therapist will assist the client to make a plan for cannabis cessation or reduction. Cognitive behavioural techniques will be used to assist with the implementation of change. These will include identifying and problem solving for high risk situations; CBT for coping with symptoms associated with substance use; and relapse prevention. Whilst the intervention will use a range of strategies to develop and consolidate the client’s motivation to reduce cannabis, we do not expect all clients to move through stages at the same pace and some may be unwilling to change their substance use. As noted below, these clients will be supported in selecting a goal that is of high priority for them whilst the therapist will seek further opportunities to review substance use within this context. The care co-ordinator will be invited to 3 meetings (beginning, middle, end of therapy) to ensure good communication between the intervention and the EIS.

From our previous studies, we anticipate that there will be considerable variation in the number of sessions participants opt to attend; that some patients will require assertive outreach (unlimited repeat appointments; flexibility in timing and location of sessions; tracking of change in residence) in order to maintain engagement; and that engagement may wax and wane over the treatment period. The number of sessions attended will be controlled for statistically in the analyses.

Mindful of the client’s stage of change regarding cannabis, for both Long and Brief conditions the therapist will attempt to progress through the following steps:

Building engagement – developing a relationship:

1. Discussion with client about current concerns and how they have arrived at current situation particularly looking at mental health issues. Development of a wants list and problem list looking at obstacles to progress and identification of how their life values fit within these.

2. Exploration of their explanatory model of psychosis (onset, symptoms and consequences). Feedback information from assessments to further explore their understanding.

3. Exploration of how cannabis use fits into life generally and then elicitation of understanding of how cannabis use fits into their explanatory model of psychosis.

4. Provision of information about cannabis and mental health, feedback of information from substance use assessments and where appropriate further monitoring of links between cannabis and mental health.

5. Development of a shared understanding/formulation outlining links between psychosis, difficulties in life and cannabis use.

6. Based on formulation, discussion of possible change options, including change in cannabis use, to address difficulties identified by client.

7. Development of a change plan. Whilst abstinence will be the primary target of intervention, it will be the client’s choice whether they wish to plan for abstinence, reduction including harm reduction, or no change in substance use. If the latter option is selected then the therapist will help the client to focus on a goal that is of importance to the client and where possible the therapist will use this alternative goal as a context in which to continue to review the role of cannabis use and to facilitate motivation for change.

8. Development of a maintenance and relapse prevention plan.

Differentiating the Brief and Long Interventions: Clients in the Brief intervention condition will be offered up to 12 sessions of MiCBT over 4.5 months; clients in the Long term intervention condition will be offered up to 24 sessions over 9 months. Both interventions will attempt to progress through the 8 steps delineated above. However, the Long intervention will potentially allow:

• More time to work through all the above steps. Those who have low motivation may be unlikely to proceed through all the steps in 12 sessions.

• More time to develop the change plan and particularly the use of CBT within the plan.

• Where clients are unwilling to change their substance use, more time to work on alternative goals and to develop motivation for changing substance use within these different contexts.

The MiCBT treatments will follow written protocols developed in our previous studies (Barrowcloughet al. , 2008), and further modified for this study by making it phase-specific for the ‘critical period’ of early psychosis. Therapists will undergo initial training and then supervision throughout the trial, and were participants give permission, sessions will be audio taped for the purpose of rating treatment fidelity.

Standard care from the Early Intervention Services involved in the study is compliant with the Policy Implementation Guide and includes intensive case management, crisis response, behavioural family therapy, and cognitive therapy for persistent symptoms.

Procedures for recruitment

Staff will be informed about the study prior to its commencement via presentations to teams and information sheets etc.

Elaboration regarding criterion a

We acknowledge the complexity of diagnostic issues with recent onset psychosis. Clients with a diagnosis of Brief Reactive Psychosis will be included at the screening stage since often this develops into full Psychosis. Information from all such cases and any others where there is diagnostic uncertainty will be reviewed by the trial psychiatrists who will decide whether diagnostic criteria are met and those included will be further reviewed diagnostically at the end of the study.

Context of the research study

The proposed study is part of an NIHR funded programme of research ‘The HELPER programme: Healthy Living and Prevention of Relapse’. The programme focuses on developing interventions to target known risk factors for relapse and poor health, in young people with early psychosis. The interventions are linked by a common theoretical framework, the Self-regulatory Model of Illness Management and designed for optimum delivery by the national network of Early Psychosis Teams. The Self Regulation Model (SRM) (Leventhalet al. , 1984) proposes that people are problem solvers whose coping strategies are attempts to close the gap between their current health (as they see it) and the best health state they believe they can achieve. As part of the research program we intend to collect data to determine how far the SRM can be applied to people with psychosis. Hence the inclusion in this study of measures common to the programme.

Symptoms, functioning and service engagement:

Symptomatology: PANSS (Kay et al. , 1989) to assess severity of symptoms; PSYRATS (Haddocket al. , 1999) to assess distress regarding symptoms.

Relapse: Two methods of assessing the frequency and duration of relapse will be used: 1) number and duration of hospital admissions identified from hospital record systems, 2) number and duration of exacerbations of symptoms lasting longer than 2 weeks and requiring change in patient management (operationally defined, Barrowcloughet al. , 1999).

Global Assessment of Functioning (GAF-D: APA, 1994): Measure of global level of functional disability, 0–100 with anchor points. Widely used and good reliability.

Service Engagement Scale (Taitet al. , 2002): A 14 item measure of client engagement with services, with subscales of availability, collaboration, help seeking and treatment adherence. Disengagement with services is a widely acknowledged clinical problem with people with dual diagnosis. For 0, 12, 24 months.

Calgary Depression Scale (Addington and Addington, 1997): This is a 9-item scale designed to assess depressed mood in schizophrenia.

Deliberate self harm assessment: This is a structured assessment of whether an episode of near-fatal DSH has occurred in the previous time period (2 minutes), as used in the CUtLASS trial. (Both substance use and psychosis are risk factors for DSH).

Attitudes to Medication Questionnaire (Hayward, 1995).

Brief Illness Perception Questionnaire (Lobbanet al. , 2005).

Schedule for the Assessment of Insight (David, 1990).

Health status: EQ-5D and Sf6D 9 (Brooks, 1996; Jenkinson & Layte, 1997). This measure is included for two purposes: Firstly, to explore the potential impact of the interventions on the health staus and quality adjusted life years of patients. Secondly, to collect data to inform an economic analysis of the interventions included in the overall HELPER programme. For the full programme, we will examine how baseline characteristics of service users affect the cost effectiveness of the interventions and economic models (decision analytic models) will be developed to explore the potential impact of each of the interventions.

Use of health and social care services: Data will be collected from case notes and assessment about the range and frequency of health and social care service use. The data will be collected using the data collection forms developed for the MIDAS trial of combined motivational and cognitive behaviour therapy for people with schizophrenia and substance abuse. Again, this measure is included for two purposes: First to explore the potnetial impact of the intervention on the health status and quality adjusted life years of patients. Secondly, to collect data to inform an economic analysis of the interventions included in the overall HELPER programme.

Substance use outcomes:

Readiness to change: The Readiness to Change questionnaire (Rollnicket al. , 1992).

Consequences of use: The Inventory of Drug Use Consequences (Blanchardet al. , 2003) a 12 item self report questionnaire assessing consequences of use.

Cannabis Experiences Questionnaire (CEQ): A measuring the subjective experiences of cannabis (Barkuset al. , 2006).

RESUS-D: Reasons for drug use questionnaire (Gregget al. , 2009).