Methods	Allocation: random, computer-generated randomisation. Blindness: single, rater-blinded. Duration: 10 weeks. Design: parallel. Location: multicentre. Setting: in- and outpatient (initially inpatient).
Participants	Diagnosis: (DSM-IV) schizophrenia, resistance to previous treatment. N = 20. Sex: 10 M, 9 F (of intent-to-treat population). Age: 24-55 years (mean clozapine = 35.7 years, mean risperidone = 36.8 years) (of intent-to-treat population). History: duration ill mean clozapine = 12.6 years, mean risperidone = 13.1 years (of intent-to-treat population), age at onset not reported. Setting: in- and outpatient (initially inpatient).
Interventions	Clozapine: flexible dose. Allowed dose range: 25-600 mg/day. Mean dose: 385 mg/day. N= 11. Risperidone: flexible dose. Allowed dose range: 2-10 mg/day. Mean dose: 7.8 mg/day. N = 9
Outcomes	Leaving the study early: any reason, adverse events, inefficacy. Mental State: PANSS total score, PANSS positive subscore, PANSS negative subscore. General functioning: GAF, social functioning scale, patient global impression scale. Satisfaction with treatment: Drug attitude inventory. Adverse effects: death (natural causes, suicide), EPS (use of antiparkinson medication), sedation, laboratory (white blood cell count)
Notes
Risk of bias
Bias	Authors’ judgement	Support for judgement
Adequate sequence generation?	Low risk	Random, computer-generated randomisation.
Allocation concealment?	Unclear risk	No further details.
Blinding? subjective outcomes	Unclear risk	Single, rater-blind. Whether blinding was successful has not been examined, but the compounds differ quite substantially in side effects. This can be a problem for blinding
Blinding? objective outcomes	Low risk	Objective outcomes such as laboratory measures or death are unlikely to have been much affected by problems of blinding
Incomplete outcome data addressed? All outcomes	High risk	The overall attrition was high (35%). The LOCF method was used to account for people leaving the study early. It assumes that a participant who discontinued the study would not have had a change of his condition if he had remained in the study. This assumption can obviously be wrong
Free of selective reporting?	Low risk	No evidence for selective reporting.
Free of other bias?	High risk	The number of participants included was rather low. Quote: “…readers have to bear in mind, that the statistical power of this study may hide some clinically relevant differences in drug efficacy.”