The study examines the effect of cognitive training on cognitive functioning and everyday competencies of patients with schizophrenia.

120 patients are expected to be included in this randomized controlled trial running at two sites in Denmark starting January 2007 The effect of a 16-week, manualized program of cognitive training integrated in a comprehensive psychosocial treatment (OPUS) for first-episode schizophrenia patients is compared with the effect of standard treatment (OPUS). A six month follow-up assessment is conducted to investigate a possible long-term learning effect of cognitive training.

Blinded assessments include the MATRICS Consensus Cognitive Battery and a co-primary outcome measure of cognitive improvement: A translated version of the UCSD Performance-based Skills Assessment (UPSA) adjusted to a Danish context.

The cognitive training consists of four modules focusing on the domain of attention, executive functioning, learning and memory. Module 1 and 2 are based on computer-assisted training tasks, and the following modules focus on more practical everyday tasks and calendar training. Cognitive training takes place twice a week and every other week the patient and trainer engage in a dialogue on the patient's cognitive difficulties, motivational goals and his/her progress in competence level.

The use of errorless learning principles, scaffolding and attentional externalisation aims at improving the patients' performance on cognitive and everyday tasks by learning to apply compensation techniques as well as limiting dysfunctional uses of available cognitive ressources (i.e. excessive self-focus, rumination).

The study will provide MATRICS Consensus Cognitive Battery results from a relatively large Danish sample of first-episode schizophrenia and contribute with valuable normative data on the UPSA.

It is hypothesized that cognitive training integrated in OPUS treatment enhances both cognitive and everyday competence of patients more than OPUS treatment alone. Expectations are that cognitive training will demonstrate a small to moderate effect on cognitive functioning and a moderate effect on everyday functioning as measured with the UPSA. Moreover, patients allocated to cognitive training are expected to show an improvement in self-esteem.

Neurocognition and competence in schizophrenia A randomized controlled trial of cognitive training (CT) integrated in OPUS treatment versus OPUS treatment-as-usual

The presence of cognitive deficits in schizophrenia It is a well-established fact that cognitive deficits play a major role in functional outcome in schizophrenia (Keefe et al, 2006; Gold, 2004). Up to 85 % patients demonstrate cognitive dysfunctions in the range of 1½-2 SD below norm (Fagerlund, 2004; Corrigan & Penn, 2001; Heinrichs & Zankanis, 1998). Cognitive dysfunctions associated with schizophrenia involve attention, information processing, memory, and executive functions leading to difficulties in learning and poor problem solving abilities.

Evidence for the effect of rehabilitation on everyday competencies In spite of the severely debilitating cognitive dysfunctions, a relatively large number of patients have shown themselves capable of acquiring new abilities (e.g. Kern et al, 2005; Sartory et al, 2005; Ueland & Rund, 2004; Krabbendam & Aleman, 2003, Pilling et al, 2002, Harvey & Sharma, 2001; Tsang & Pearson, 2001, Danion et al, 2001, Bell et al, 2003, Wykes et al, 1999, 2003). Research has shown that training in symptom handling, social proficiency training, and supportive provisions in occupational relations apparently can help patients with schizophrenia to function in their daily lives, learn solving concrete practical and social tasks, and manage relatively independently. These improvements enhance self-esteem.

Effect studies of cognitive rehabilitation A Cochrane review done by Hayes & McGrath (2000) include three small randomized controlled trials and fails to provide conclusive data on evidence for or against cognitive training as a treatment for schizophrenia.

Krabbendam & Aleman's review (2003) of 12 controlled studies finds a weighted mean effect size (Cohen's d) of 0.45 with 95% CI, 0.26-0.64), which indicates a small to moderate effect of cognitive training (CT) for patients with schizophrenia.

In Twamley et al.'s review (2003) of 17 randomised, controlled studies of CT, a weighted mean effect size was calculated across studies. The weighted mean effect sizes (Cohen's d) were 0.32 in improvement in cognitive performance, 0.26 in reduction of symptom severity, and 0.51 in the domain of everyday functioning. The review authors recommend longitudinal designs as well as the inclusion of real-life outcome measures in future trials.

Many studies of CT have been underpowered (small sample sizes), which limits the conclusions that can be drawn from results of improvement (Ueland & Rund, 2005). The variability of outcome measures used also complicates interpretation of findings, thus, a consensus battery (MATRICS ) of outcome measures has been suggested for future studies. Silverstein & Wilkniss (2004) and Bellack (2004) note that the results of cognitive rehabilitation generally fail to secure the ecological validity, wherefore CT ought to be integrated in a broader psychosocial programme of intervention for schizophrenia. Other shortcomings in earlier studies include short-term programmes lacking sufficient long-sighted follow-up examinations as well as experimental designs lacking consideration of individual abilities and involvement. A recent review of cognitive rehabilitation (Velligan et al, 2006) underscores the importance of addressing motivation as a primary target.

Hogarty et al's (2004) promising results from a randomised, controlled trial (N=121) of computer-assisted CT also highlight the importance of examining to what extent the control condition (psychosocial treatment) improve cognitive functioning. Broadly viewed, there is sufficient evidence that the combination of CT and psychosocial treatment, rather than individually appearing initiatives, have the greatest potential for promoting treatment outcome (Bell et al, 2003).

There is a need for a well-organised and sufficiently large randomised clinical trial of the effect of integrated CT and psychosocial rehabilitation. In order to examine to what extent CT improves the psychosocial rehabilitation and facilitates new learning, one must identify and compare improvements on the level of neuropsychological performance as well as the level of everyday competencies (Reeder et al, 2006; Buchanan et al, 2005; McKibbin et al, 2004). The present trial employs a prospective design of 16 weeks with a following follow-up 10 months after inclusion, and seeks in the CT to secure an individual adaptation and motivation trough continuous evaluations and level adjustments between patient and trainer.

Objectives

1. Primary hypothesis The effect of a 16-weeks programme of computer-based CT integrated in OPUS treatment on the patients' competencies in daily life is superior to standard OPUS treatment. Primary response variable: Total score on UPSA-B (Patterson et al, 2001 Danish version).
2. Secondary hypothesis The effect of a 16-weeks programme CT integrated in OPUS treatment on patients' cognitive functioning is superior to standard OPUS treatment. Response variables: Scores on MATRICS Consensus Cognitive Battery (MCCB), Trailmaking B and WCST-64 (computerized version).
3. Tertiary hypothesis The effect of a 16-weeks programme of computer-based CT integrated in OPUS treatment is superior to standard OPUS treatment in terms of patients' association with the labour market, general health and self-esteem. Response variables: Occupational status and general functioning measures and Rosenberg's Self-Esteem Scale (Rosenberg, 1989).

Furthermore, we want to examine the association between first episode patients' cognitive function level and competencies in daily life, can be divided in primary and secondary objectives:

Blinding The instructor for CT will make a call with participants' data to CTU and thereafter inform the patients which intervention programme he or she has been allocated to. The study is not blinded in regard to patients, cognitive trainers and therapists. The blinding applies raters engaged with the outcome evaluation. In the follow-up the patient is instructed in advance not to reveal what type of treatment they have received. By registering which type of intervention the independent assessors of outcome believe the patient to be receiving it will be evaluated if blinding succeeded. The randomised intervention allocation is concealed until the statistical analyses of resulting data have been completed.

Participants Patients are recruited from OPUS which is implemented as a standard treatment programme for young adults with firstepisode psychosis in Copenhagen and Århus. The staff in OPUS recruit patients to the project. The independent assessors interview referred patients and evaluate the following in- and exclusion criteria.

Inclusion criteria 1) between 18 and 35 years of age, 2) first episode psychosis within F2 spectrum or schizotypal disorder in ICD 10 - post-acute phase of illness, 3) sufficient comprehension of Danish, 4) written informed consent.

Exclusion criteria

1) rejection of participation, 2) organic disorder, 3) noticeable misuse of psychoactive drugs

Control intervention group All patients receive the treatment usually provided in OPUS.

Standard treatment in OPUS consists of affiliation with a primary contact person, involvement of family, possibility of psychoeducation and social skills training described in details at http://www.psykiatri.aaa.dk/cent\_enh/opus.htm and www.opus-kbh.dk. Depending on individual needs patients are offered to take part in group therapy either social skills training (SST) or cognitive-behavioral therapy (CBT). Participation in SST or CBT might be beneficial for both cognition and everyday skills. To ensure comparability between the patients allocated to experimental intervention (CT) and the patients allocated to standard OPUS treatment, we stratify patients according to group therapy (SST/CBT) yes/no.

Patients randomized to the control group will receive treatment as usual as described above.

Experimental intervention A pilot study in OPUS Århus has led to the completion of a manual for a 16-week CT programme for patients with schizophrenia (Christensen & Olsen, 2006).

One hour twice a week in 16 weeks patients engage in computer-assisted CT plus one competence dialog every other week. Training consists of four modules: the first three modules cover the areas of attention, memory and executive functions, and the last module focus on the cognitive area and related tasks that the individual patient need or prefer to work more on: Thus, the content of module 4 is based on both the patient's and the trainer's judgement.

Training contains exercises of simple attention, attention span and vigilance, planning, problem solving, stimulus-bind and perseveration tendency, interaction-based training of working memory, and verbal and visual long-term memory. The CT approach is oriented towards teaching of compensatory strategies, but simple repetition of non-social cognitive tasks on a gradually increasing level of difficulty is also part of the CT program. By 90 % correct task-completion, the level is increased.

Module 1 and 2 are based on computertasks (COGNIsoft), whereas the following modules 3 and 4 include practical everyday tasks. In their review Twamley, Jeste and Bellack (2003) stress the importance of directly targeting everyday functioning of the patient. Thus, calendartraining is a vital part of this intervention: a way of addressing common difficulties of memory and planning ability. Calendartraining and competence dialogues both support environmental adaptation and transference of learning.

Training relies on errorless learning principles, scaffolding, repetition, and strategy-learning as recommended by Wykes & van der Gaag (1999). Positive reinforcement, modelling, and verbal instructions are widely used. Trainers describe the structure and explain the purpose of training exercises at the beginning of every session to provide a consistent learning environment.

Assessments All assessments except Present State Examination-interview (PSE, clinical version) and premorbid IQ (DART: Danish version of National Adult Reading Test, Nelson, 1991) are conducted at baseline, 4 months and 10 months after inclusion, irrespective of whether patients were following the full training programme (intention-to-treat principle).

• Positive and negative symptom evaluated with Positive and Negative Symptom Scale (PANSS).
• Assessment of health and occupational and general function.
• Neuropsychological testing of functional level in 7 cognitive domains. According to the NIMH-initiative the MATRICS Cognitive Consensus Battery (MCCB, Nuechterlein et al, 2004) includes seven separate cognitive areas. See table 1.
• Current level of competencies in 2 domains of daily life assessed using UCSD Performance-Based Skills Assessment (UPSA-B, Patterson et al, 2001, Danish version)
• Rosenberg's Self-Esteem Scale (Rosenberg, 1989 (1965), Danish version)
• Type and dose of antipsychotic medication (chlorpromazine equivalents).

Table 1. Cognitive test battery.

Domain of cognition: Tests:

1. Speed of information processing BACS Symbol CodingCategory FluencyTrailmaking A
2. Attention CPT-IP Identical Pairs
3. Working memory Spatial SpanLetter-Number SpanTrailmaking B
4. Verbal learning and memory Hopkins Verbal Learning Test-Revised
5. Visuel learning and memory Brief Visuospatial Memory Test-Revised
6. Problem-solving NAB MazesWisconsin Card Sorting Test, computerized version,64 cards
7. Social cognition MSCEIT (Managing Emotions)

Ethical considerations and informed consent The project has been approved by the Scientific Ethical Committees for Copenhagen and Frederiksberg and assigned case number (KF) 01 300017. The guidelines of The Danish Data Inspection and "clinicaltrials.gov" are followed. The interventions and methods of investigation involve no known physical or mental risks. Participation is voluntary and written informed consent is obtained. All patients are informed both verbally and in written form that they can withdraw from the trial at any time, without it having any consequences for their continued treatment.

Inclusion Criteria:

• Age between 18-35 years,
• First-episode schizophrenia og schizotypal disorder in ICD-10,
• Post-acute phase of illness,
• Sufficient comprehension of Danish,
• Written informed consent

Exclusion Criteria:

• Rejection of participation,
• Organic disorder,
• Misuse of psychoactive drugs