Underhill 1990.

Methods	Study design: randomized controlled trial Study duration: recruitment period between April and October 1988 Setting: hospital setting: accident and emergency departments of two teaching hospitals Country: UK Number of individuals randomized: 60 Number of individuals receiving the intervention: Gastric lavage: 14 Acitvated charcoal: 20 Ipecac: 21 Number of individuals receiving the control: 5 Number of individuals lost to follow‐up: 0 Sample size calculated:
Participants	Sex: 16 male and 44 female Age: mean (range): 25.7 (16‐62) years Country (if different from study authors'): NA Type, dose and timing of poisoning: participants presenting within 4 h after an overdose (mean delay: 123 min, range 30‐240 min) of at least 5 g paracetamol. 48 took paracetamol without another drug; 21 took paracetamol with alcohol Inclusion criteria: > 16 years, presenting < 4 h after intake, ingested > 5 g paracetamol Exclusion criteria: depressed conscious level, conditions that might preclude use of any one of the treatment methods
Interventions	Intervention arm 1: Type: gastric lavage Timing: NA Dose: NA Frequency: NA Integrity: NA Intervention arm 2: Type: activated charcoal (Carbomix) Timing: no information Dose: AC:Drug ratio of 10:1 Frequency: 1× Integrity: 16 participants managed to take the recommended dose. 4 participants vomited and 1 refused to take more than one mouthful Intervention arm 3: Type: ipecac Timing: no information Dose: 30 mL Frequency: 1×, repeated if no vomiting after 30 min Integrity: mean time to emesis was 20 min (range 5‐50), 2 participants did not vomit until 50 min and 2 did not vomit at all Control arm: Type: no intervention Timing: NA Dose: NA Frequency: NA Integrity: NA
Outcomes	Type (unit): Plasma paracetamol levels Adverse events Timing: prior to treatment and 60 min, 90 min and 150 min after the first sample (early)
Funding	No information
Notes	No treatment group was stopped for ethical reasons when the serum paracetamol levels increased between the first and last samples in 4 out of 5 participants. Data for gastric lavage was not extracted, because not within scope of this review.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No information on randomization process
Allocation concealment (selection bias)	Unclear risk	No information on allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Participants were not blinded (not possible due to difference in interventions); might influence outcomes
Blinding of outcome assessment (detection bias) All outcomes	Low risk	No information on blinding, but should not affect outcome measurements
Incomplete outcome data (attrition bias) All outcomes	Low risk	No indication of incomplete outcomes
Selective reporting (reporting bias)	High risk	No clinical outcomes reported, adverse events incompletely reported
Other bias	Low risk	No other risk of bias detected