Fassoulaki 2005.

Methods	Double‐blind (participant, outcome assessor), placebo‐controlled, randomized clinical trial Sequence generation by computer‐generated random number tables Follow‐up: 6 months
Participants	Participants: 50 adults in a university setting in Athens, Greece Operation: breast surgery (modified radical mastectomy and lumpectomy plus axillary dissection) for breast cancer 2 groups, size: 25/25 Age (group 1, 2): 49 years (SD ± 8.4), 48 (SD ± 8.1) Men/women: 0/50
Interventions	Group 1 (multimodal): GA, brachial plexus irrigation with ropivacaine (0.75%, 10 mL), intercostal ropivacaine (0.75%, 3 mL) at intercostal spaces 3‐5, post‐op for 3 d topical (wound, sternum, axilla) EMLA cream (20 g, 2.5% lidocaine/prilocaine), codeine, paracetamol Group 2 (control): GA, brachial plexus irrigation with normal saline, sham intercostal block at intercostal spaces 3‐5, post‐op for 3 d topical (wound and axilla) placebo cream, codeine, paracetamol Adjuvants: Group 1: gabapentin (400 mg, orally every 6 h starting the night before surgery) for 8 d, Group 2: placebo as above Immediate post‐op pain control: significantly improved
Outcomes	Dichotomous: pain, analgesic consumption at 6 months Continuous: none reported Adverse effects, withdrawal and attrition were reported with group allocation.
Notes	We contacted the study author and we acknowledge the response, providing details on source of funding and conflict of interest. Funding sources: study author responded "the study was funded from Departmental sources only." Conflicts of interest: the study author responded "none of the authors has conflict of interest relevant to the study."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "fifty envelopes, 25 containing odd and 25 containing even numbers, obtained from a computer‐generated table, were prepared and sealed...," this is an adequate description of an acceptable randomization technique. Bias is unlikely.
Allocation concealment (selection bias)	Low risk	Quote: "an independent anesthesiologist, who did not participate in the study or data collection, read the number contained in the envelope and made group assignments." Bias is unlikely.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "except for the independent anesthesiologist, [not involved in the study] no other physician or nursing staff member was aware of the interventions administered to each patient." "Regarding EMLA cream and possible interference with blinding, EMLA or placebo was applied in the morning after pain assessment"... "pain was assessed by an anesthesiologist blinded to group assignment." "Placebo capsules were identical in appearance with the gabapentin capsules. The same number of capsules was packaged in group‐specific bottles and coded as bottle A and bottle B for the control and treatment groups, respectively. A white odourless cream was the control treatment corresponding to the EMLA cream. Similarly, cream for each group was kept in boxes labelled as A and B for the control and treatment groups, respectively."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "except for the independent anesthesiologist, (not involved in the study) no other physician or nursing staff member was aware of the interventions administered to each patient." "Pain was assessed by an anesthesiologist blinded to group assignment."
Incomplete outcome data (attrition bias) All outcomes	High risk	Study authors provide a good account of attrition, including group allocation, but considered no ITT analysis: dropouts, participants lost to follow‐up, failures, etc were all excluded.
Selective reporting (reporting bias)	Low risk	Primary outcomes fully reported on
Null bias	Low risk	Quote: "the treatment group consumed less paracetamol in the PACU... and fewer Lonalgal® tablets... than the controls, exhibited lower visual analog scale scores at rest in the PACU... and on postoperative Days 1, 3, and 5"