Giorgio 2011.

Study characteristics
Methods	Randomised clinical trial
Participants	Country: Italy Number randomised: 285 Postrandomisation dropouts: 0 (0%) Revised sample size: 285 Average age: 70 years Females: 78 (27.4%) Cirrhosis: 285 (100%) Very early HCC: 71 (24.9%) Portal hypertension: not stated Viral aetiology: 285 (100%) Immunotherapy/antiviral adjuvant therapy: not stated Average follow‐up period in months (for all groups): mean: 37 months Criteria for early or very early HCC and other inclusion criteria: Single nodule, ≤ 3 cm
Interventions	Participants were randomly assigned to 2 groups: Group 1: PEI (n = 143). Further details: PEI using 4 to 20 mL of 95% ethanol depending upon tumour volume. Group 2: radiofrequency ablation (n = 142). Further details: radiofrequency ablation generator details not stated.
Outcomes	The outcomes reported were: mortality, adverse events.
Notes	Although mortality was reported, this was a severely biased estimate, as 14 people who could not undergo radiofrequency ablation were excluded. We therefore did not use the survival information. Authors provided additional information in February 2017.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "random number generator"
Allocation concealment (selection bias)	Low risk	Quote: "The person randomising the patient were unaware of what the next treatment allocation was. It was used a centralised randomisation service to ensuring allocation concealment. So it was not possible for the investigators to know the allocation sequence in advance" (author replies)
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "The patients and healthcare providers were not blinded due to the nature of the treatments used in to the study (PEI versus RFA)" (author replies)
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "The outcome assessors were blinded as they did not know the patient was referring to the results" (author replies)
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: there were no postrandomisation dropouts.
Selective reporting (reporting bias)	Low risk	Comment: important clinical outcomes were reported.
For‐profit bias	Low risk	Quote: "The study was not funded. It was self‐financed by the hospital" (author replies)
Other bias	Low risk	Comment: no other bias noted.