Fang 2012.

Methods	Allocation: randomised. Blindness: open. Duration: 120 hours (as for session I).
Participants	Diagnosis: DSM‐IV diagnosis of schizophrenia (N = 192), schizophreniform (N = 11) or schizoaffective (N = 0) disorders, no information* (N = 2); PANSS‐EC ≥14; PANSS ≥60. N = 208 screened, 205 randomised. Age: range 18‐45 years. Sex: 99 males, 106 females. History: acute agitation. Excluded: "women pregnant, breastfeeding or planning to become pregnant during the study", "psychotic agitation caused by delirium, epilepsy, mental retardation, or affective disorder", "intoxication or symptoms of withdrawal from alcohol or other psychoactive substances", "serious medical illness", "known hypersensitivity to any of the study medications or no response to risperidone or haloperidol during previous treatment", "treatment with a depot antipsychotic with one cycle of screening", "use of disallowed medication". Setting: multi‐centre (6), China.
Interventions	1. Haloperidol: flexible dose IM 10 mg/day to 20 mg/day (mean prescribed daily dose 12.2 ± 3.7 mg), N = 101. 2. Risperidone: flexible dose oral 2 mL/day to 6 mL/day (mean prescribed daily dose 3.4 ± 0.7 mg), N = 104 + clonazepam: flexible dose oral 0 mg/day to 8 mg/day (mean prescribed daily dose 2.9 ± 1.5 mg), N = 99.
Outcomes	Not asleep. Mental State: PANSS total. Agitation: PANSS‐EC. Adverse events: insomnia, tachycardia, EPS, akathisia. Unable to use: Adverse effects: SAS (mean, SE/SD not reported). Adverse effects: BAS (mean, SE/SD not reported).
Notes	*Author assume a typing error in the demographic‐clinical "table 1". An attempt to contact the authors was made (2nd July 2016).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"A telephone‐based central service was used to randomly assign eligible patients" (p.108).
Allocation concealment (selection bias)	Unclear risk	Not stated.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open trial.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Open trial.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No evidence of incomplete outcome data.
Selective reporting (reporting bias)	High risk	SAS and BAS scores at 2, 4, 24, 72, and 120 hours are not reported. Only side effects occurred in ≥ 5% of people are reported.
Other bias	High risk	Sponsored by drug company (Xian‐Janssen Pharmaceutical Ltd).