Schizophrenia is a mental health problem usually starting in the late teens/early twenties, and often lasting many years. The standard medication ('antipsychotics') for this problem is usually helpful, and if taken continually can keep people well, reducing the likelihood of further episodes. However, in up to one in three people with schizophrenia, the illness does not show much improvement with antipsychotic medication. For some of these 'resistant' illnesses, one particular antipsychotic, clozapine, can work well, but one disadvantage is the risk of a severe blood side effect which means that regular blood testing is necessary. If the response to clozapine treatment is disappointing, there is some evidence that adding another antipsychotic can sometimes produce more improvement. However, it seems that the added antipsychotic may need to be taken by the person for at least 10 weeks in order to work well. The investigators plan to test carefully the possible benefits and problems when the antipsychotic amisulpride or a dummy tablet ('placebo') is added to clozapine for 12 weeks in people whose schizophrenia illness has not been helped much by any antipsychotic medication on its own, and who are now taking clozapine, but again with not much improvement. The investigators have chosen amisulpride because its pharmacological action may be complementary to that of clozapine, and also it is less likely than some other antipsychotics to compound some of the characteristic side effects of clozapine, such as sedation, weight gain and other metabolic problems.

Adjunctive amisulpride or placebo will be randomly assigned. The investigators expect that adding amisulpride will be more likely to cause an improvement than adding placebo. But the investigators should learn more about the risks and side effects of combining these two medications. Also, the investigators should gain a greater understanding of the possible benefits of adding another antipsychotic to clozapine in relation to particular problem symptoms, and a person's ability to live and work in the community.

This 12-week, placebo-controlled RCT will be conducted in secondary care, specifically mental health services, at UK centres. The health technology to be assessed is the augmentation of clozapine treatment with another second-generation antipsychotic, amisulpride, which will be compared with placebo: 400mg amisulpride or 1 matching placebo capsule for the first 4 weeks, then the option of titrating up to 800mg amisulpride or 2 matching placebo capsules for the remaining 8 weeks. The study will be double-blind, with medication supplied as identical capsules containing either 400mg amisulpride or placebo. The optimum dose of clozapine at entry and subsequent augmentation will be achieved through a flexible dosing regimen whereby treating psychiatrists will be able to flexibly alter dose regimens to maximise clinical risk-benefit ratios; there will be opportunities for clinical titration of clozapine dose at two and six weeks. Any direct pharmacokinetic effect on clozapine levels will be assessed by pre- and post-augmentation plasma levels of clozapine, samples being taken at baseline and at the end of the 12 weeks. Recommended pharmacovigilance procedures will be followed. Clinicians will be asked not to prescribe any additional medication during the course of the study, and will be reminded of the drugs with potential adverse interactions, as mentioned in the SPCs for clozapine and amisulpride. Medication adherence will be assessed by 'pill count' and clozapine/norclozapine plasma level ratio.

Therapeutic improvement will be assessed in terms of overall symptom severity, but also using broader, clinically-relevant outcome measures of social and occupational function and target symptoms and/or behaviours as well as overall health status and utility. Side effects will be systematically assessed. The costs and outcomes for a cost effectiveness acceptability and net benefit analysis will also be measured. The primary economic measure will be the incremental cost effectiveness ratio of clozapine augmentation, estimated as the net cost of clozapine augmentation divided by net QALY of clozapine augmentation.

Twenty-four months will be allowed for recruitment of participants, plus 3 months for the final follow-up assessments.

Inclusion Criteria:

1. A criterion level of persistent symptom severity despite an adequate trial of clozapine monotherapy in terms of dosage, duration and adherence (as used by Honer et al 2006):

   • Treatment for at least 12 weeks at a stable dose of 400 mg or more of clozapine a day, unless the size of the dose was limited by side effects
   • A total score of 80 or greater at baseline on the Positive and Negative Syndrome Scale (PANSS: Kay et al 1987, 1988); the range of possible scores is 30 to 210, with higher scores indicating more severe symptoms.
   • A Clinical Global Impressions (CGI: Guy 1976) score of 4 or greater (range of possible scores, 1=not mentally ill to 7=extremely ill)
   • A Social and Occupational Functioning Assessment Scale (SOFAS: Goldman et al 1992, DSM-IV 1994) score of 40 or less; range of possible scores, 1 to 100, with lower scores indicating impaired functioning.

2. Age 18-65 years, inclusive

3. Clinically stable for the last 3 months with a consistent clozapine regimen.

4. Competent and willing to provide written, informed consent.

Exclusion Criteria:

1. Clinically-significant alcohol/substance use in the previous three months
2. Developmental disability
3. Indication for current treatment with clozapine was intolerance/movement disorder
4. A previous trial of clozapine augmentation with amisulpride.
5. Existing relevant physical health problems: such as cardiovascular disease, previous problems with prolactin, and impaired liver/ renal function.
6. Any woman who is pregnant or planning a pregnancy, and any woman of child bearing potential unless using adequate contraception.