ATLAS: Antipsychotic Treatment of very LAte-onset Schizophrenia-like psychosis

1. Brief Psychiatric Rating Scale (BPRS), a widely used clinician-rated 24-item instrument for assessing positive, negative and affective symptoms in patients with psychotic disorders (Ventura et al 1993). The BPRS (Appendix D) consists of 18 symptom constructs and takes 20-30 minutes for the interview and scoring. Each item is assessed by the rater on a 7-point scale ranging from 1 (not present) to 7 (extremely severe). The total score is obtained by summing the scores from the 18 items. Scores thus range from 18 to 126, with higher scores indicating greater levels of psychopathology. The BPRS will be administered at baseline, at week 4, then between weeks 10-12 and between weeks 34-36. Changes in BPRS score between baseline and the week 10-12 assessment and between the week 10-12 and week 34-36 assessments are the trial’s co-primary outcomes.
2. The proportion of patients withdrawn to open treatment with amisulpride by their physicians between Weeks 13 and 36 (Stage 2) because of perceived lack of efficacy, which will be compared in participants randomised to continue amisulpride (arm A) and switch to placebo (Arm B). The difference between groups in the percentage of patients stopping trial treatment because of physician concerns about non-efficacy was used as an outcome measure in the CATIE-AD trial (Schneider et al 2003).

1. Extrapyramidal side-effects (EPSE) measured with the Simpson-Angus Scale (SAS). The modified SAS being used in ATLAS measures nine extrapyramidal signs: gait, arm dropping, shoulder shaking, elbow rigidity, wrist rigidity, leg pendulousness, glabellar tap, tremor, and salivation, all of which are assessed by direct examination. The head dropping item is omitted because of difficulties with this assessment in home visits. Each item is rated on a scale of 0-4, with higher scores indicating greater severity of EPSE. The scale range of the modified SAS is thus from 0-36. A standardised description is given of each item to optimise reliability. The SAS has been widely used to measure EPSE within clinical trials and will be administered at baseline, 4, 10-12 and 34-36 weeks. The change in SAS between Baseline and Week 10-12 and between Week 10-12 and Week 34-36 will be compared between groups to assess EPSE.
2. Compliance expressed as treatment discontinuation rates and as percentage of prescribed trial medication taken between Weeks 1 and 12 and between Weeks 13 and 36 will be compared in patients allocated to receive amisulpride and those allocated placebo.
3. Quality of life measured with the self-rated, short, 26-item, WHO Quality of Life Scale (WHOQOL-BREF) at baseline, 10-12 and 34-36 weeks. The WHOQOL-BREF includes two items about an individual’s overall perception of their quality of life and health and questions assessing four domains: physical, psychological, social and environmental well-being, with higher scores denoting a better quality of life. This instrument has been previously used in studies of older people with schizophrenia and psychosis patients in residential care settings (Picardi et al 2006).
4. Cost-effectiveness calculated by attaching nationally applicable unit cost measures to health and social service use and medication data collected for each patient with a modified version of the Client Service Receipt Inventory (CSRI) at 10-12 and 34-36 and the EQ-5D (EuroQol) at baseline, 10-12 and 34-36 weeks. We will also collect data on informal carer inputs, and attach imputed values.
The cost-effectiveness of oral amisulpride versus placebo over 12 weeks and the cost-effectiveness of treatment continuation for a further 24 weeks versus discontinuing amisulpride treatment at 12 weeks will be investigated. The differences between patient groups in the costs for each Stage (covering health and social care service use, medication and carer support) will be calculated, and an incremental cost-effectiveness ratio estimated using QALYs computed from EQ-5D. We will also examine cost-effectiveness by plotting a cost-effectiveness acceptability curve (CEAC) generated from the net benefit approach using bootstrap regression for a range of hypothesised values for willingness to pay for incremental improvements in psychiatric symptoms measured on the BPRS. Each cost-effectiveness analysis will be conducted from the perspective of (a) the NHS and social services, and (b) society as a whole, the main difference being the exclusion of formal and informal carer costs and out-of-pocket patient or carer payments