Chiu 2008.

Methods	Triple‐blind (participant, provider, outcome assessor) placebo‐controlled, clinical RCT Sequence generation method not described Follow‐up: 3 months
Participants	Participants: 40 adults at a teaching hospital in New Taipei City, Taiwan Operation: minimally invasive cardiac surgery (coronary artery bypass performed through left thoracotomy via 4th or 5th intercostal space without cardiopulmonary bypass, valvular surgery through a right lateral thoracotomy via 4th intercostal space with cardiopulmonary bypass) 2 groups, size: 19/19 (actually completed) Age (± SD), group 1, 2: 57.4 (± 15.2), 59.7 (± 13.8) Men/women (group 1, 2): 12/7, 13/6 Remarks: 40 participants were randomized, but 2 were excluded, 1 per group, because of protocol violation Surgery type: coronary artery bypass/valve surgery (group 1, 2): 5/14, 6/13
Interventions	Group 1 (placebo): 10 mL saline infused via catheter at end of operation, continuous infusion saline 2 mL/h x 48 h Group 2 (thoracotomy wound infusion): 10 mL 0.15% bupivacaine infused at end of operation then continuous infusion 2 mL/h x 48 h Both groups had same GA regimen with etomidate, fentanyl, rocuronium and sevoflurane and multi‐orifice catheter placed at a SC layer during wound closure. Post‐op breakthrough analgesia for both groups with IV PCA (morphine 0.5 mg/mL, fentanyl 5 µg/mL, tenoxicam 0.8 mg/mL) basal infusion rate 0.1 mL/h, bolus 1 mL, lockout 15 min. After 72 h, oral or parenteral NSAIDs or opioids were used. Adjuvants: none Immediate post‐op pain control: significantly improved, significantly reduced analgesic consumption
Outcomes	Dichotomous: none Continuous: VAS Other reported: IV PCA consumption in first 72 h post‐op
Notes	Funding sources: source of funding not reported. Conflicts of interest: no conflict of interest statement given
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "patients were randomly assigned" but no description of method of randomization or at what time point it was done.
Allocation concealment (selection bias)	Unclear risk	Allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "the nurse connecting the infusion bag to the catheter, the surgeons, the patient...were all blinded to the nature of the infusion".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The nurse evaluating the pain score was blinded to the nature of the infusion. Does not explicitly say, but likely the individual evaluating pain score at 90 days after was also blinded.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	1 participant in each group was excluded as a result of "protocol violation (limited consciousness)". No ITT analysis was done. Did not report on the number of individuals assessed at 3‐month follow‐up time point (or if any lost to follow‐up)
Selective reporting (reporting bias)	Low risk	No protocol available but primary outcomes specified in paper were fully reported on
Null bias	Low risk	Quote: "not only did the bupivacaine wound infusion reduce pain during the first 48‐hour infusion period, but it also provided reduced pain at 24 hours after cessation of the infusion"