TREC‐Rio‐I.
| Methods | Allocation: randomised. Blindness: none. Duration: 14 days. Setting: 3 inner‐city emergency rooms of middle‐income country | |
| Participants | Diagnosis: psychosis (219), substance abuse (51), others (30). N=301. Sex: women 155, men 146. Age: mean ˜ 38 years (SD ˜ 11). History: agitation on presentation to emergency room, first psychiatric attendance (26), markedly severely agitated or worse (192) | |
| Interventions | 1. Haloperidol IM: dose up to 10 mg stat + promethazine IM: dose up to 50 mg stat. N=150. 2. Midazolam IM: dose up to 15 mg stat. N=151 | |
| Outcomes | Tranquil or asleep*. Specific behaviours: other episodes of aggression. Global state: use of additional medication, use of restraints/seclusion, needing extra visits from the doctor, refusing oral medication. Adverse effects: serious adverse effects. Service outcomes: no hospital discharge. Leaving the study early | |
| Notes | *Primary outcome chosen by emergency room staff (tranquil or asleep by 20 minutes) | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "Randomization by excel generated numbers, block sizes were applied to a table of random numbers, sealed packs". Review author judgement: Adequate random sequence generation |
| Allocation concealment (selection bias) | Low risk | Quote: "Table of allocation sequence independent of block size produced to ensure correct drug was consecutively numbered and then sealed". Review author judgement: Adequate concealment |
| Blinding of participants and personnel (performance bias) All outcomes | Unclear risk | No participant blinding |
| Blinding of outcome assessment (detection bias) All outcomes | Unclear risk | No outcome blinding, outcomes selected to be robust to detection bias, and additional blinded student concurrently accurately timed their opinion of when tranquillisation occurred; this blind and accurate rating concurred with that of the unblinded staff |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Participants randomised were 301, and the follow‐up was explained well |
| Selective reporting (reporting bias) | Unclear risk | No evidence or information |
| Other bias | Unclear risk | None known |