Bell 2001.

Methods	Double‐blinded (participants, outcome assessors), placebo‐controlled, clinical RCT Sequence generation randomized but not described Follow‐up: 6 months
Participants	Participants: 8 adults in a university setting in Bergen, Norway Operation: bilateral reduction mammoplasty 2 groups, size: 8/8 Age: 28.5 years (range 18‐34) Men/women: 0/8 Remarks: body sides, not participants randomized
Interventions	Breast group 1 (preop infiltration): GA (fentanyl), preincision: infiltration with lidocaine (0.5%, 100 mL with epinephrine 5 µg/mL), post‐op as needed ketobemidone (oral, 5 mg) and paracetamol (1000 mg 3 x daily) Breast group 2 (placebo): GA (fentanyl), preincision: infiltration with normal saline (100 mL with epinephrine 5 µg/mL), post‐op as needed ketobemidone (orally, 5 mg) and paracetamol (1000 mg 3 x daily) Adjuvants: none Immediate post‐op pain control: significantly improved in treated breasts
Outcomes	Dichotomous: pain at 6 months Continuous: none reported Secondary: thermal thresholds were reported as tables, touch allodynia, or hyperalgesia
Notes	Some details, reported as graphs, are difficult to compare and extract. We acknowledge the study author's response regarding sources of funding and conflict of interest statement. Funding sources: the author informed us that this was an investigator‐initiated study, supported by an unrestricted grant from Astra Zeneca initially to study the effects of ropivacaine. When the study authors could not obtain approval to study this drug, the company maintained their support. The study author wrote that "the results were analysed with the help of a statistician at Astra Zeneca... we were allowed to keep the equipment... and that Astra financed my travel to a conference..." Conflicts of interest: the author had "no conflict of interest... and did not receive any [other] salary or economic compensation from Astra Zeneca."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: “patients’ breasts were randomized to test and control groups”, but the method was not described in detail.
Allocation concealment (selection bias)	Unclear risk	Efforts to conceal allocation were not described. Bias is rather unlikely, because body sides, not participants were randomized.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "the procedure was performed double blind", however blinding of participants and personnel not explicitly described
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "the procedure was performed double blind", however outcome assessor blinding not explicitly described
Incomplete outcome data (attrition bias) All outcomes	Low risk	Withdrawals and attrition reported as none, except one participant excluded for drug spillage. With only one withdrawal, body parts randomized not participants, even though no ITT analysis was performed, bias seems unlikely.
Selective reporting (reporting bias)	Unclear risk	Quote: "some details, reported as graphs, are difficult to compare and extract"
Null bias	Low risk	Quote: "the sum of VAS scores for pain intensity was significantly lower in the lidocaine group than in the placebo group for the entire registration period of 10 h after wound closure"