2. Suggested design for a randomised trial.

Methods	Allocation: randomised, clearly described. Blinding: double, described and tested. Duration: 52 weeks.
Participants	Diagnosis: have psychoses. N=300*. Age: any. Sex: both. History: persistently aggressive.
Interventions	1. Haloperidol: dose flexible within recommended limits. N=150. 2. Other antipsychotic drug: dose flexible within recommended limits. N=150.
Outcomes	Outcomes grouped into short term (3 months or less), medium term (> 3 and up to 6 months or less) and long‐term (> 6 months). Aggression: events. Adverse effects. Specific behaviours ‐ self harm including suicide, injury to others, aggression. Global outcomes ‐ overall improvement, use of additional medication, use of restraints/seclusion. Service outcomes ‐ duration of hospital stay, re‐admission. Mental state ‐ no clinically important change in general mental state. Leaving the study early ‐ why. Satisfaction with and acceptance of treatment. Quality of life ‐ no clinically important change in quality of life. Economic outcomes.
Notes	* Powered to be able to identify a difference of ˜ 20% between groups for primary outcomes with adequate degree of certainty.

N: number of participants.