O'Neill 2012.

Methods	Single‐blind (outcome assessor), RCT Sequence generation by computer‐generated random numbers Follow‐up: 3 months
Participants	Participants: 67 women aged 18‐50 years, gestational age 37‐42 at hospital setting in Lisbon, Portugal Operation: elective caesarean section delivery (with Pfannenstiel incision) Groups, size: 29/29 Age (years ± SD; group 1, group 2): 33 ± 5, 33 ± 5 Men/women (group 1, 2): 0/29, 0/29 Primary caesarean delivery (n, group 1/2): 25/24
Interventions	Group 1 (continuous wound infusion group): anaesthesia was performed through SAB with hyperbaric bupivacaine and sufentanil with single‐shot SA. Intra‐op: catheter placed in wound below fascia after peritoneum closed, 10 mL ropivacaine 10 mg/mL injected during wound closure, then continuous infusion ropivacaine 2 mg/mL at 5 mL/h for 48 h Group 2 (epidural morphine): anaesthesia initiated with combined spinal‐epidural technique to site epidural catheter, single‐shot SA. Intra‐op: upon partial recovery from motor blockade (Bromage score 2), initiated 2 mg/10 mL bolus epidural morphine every 12 h (x 4 times). Neither group received any preanaesthetic medication. Both received standardized post‐op analgesia with paracetamol 1 g every 6 h x 48 h, breakthrough pain (VAS > 3) with IM diclofenac 75 mg every 6 h as needed, ondansetron 4 mg IV for nausea or vomiting as needed Adjuvants: none Immediate post‐op pain control: significantly improved
Outcomes	Continuous: presence or absence of "residual pain related to the scar or pain that the patient related to caesarean delivery" at 3 months Dichotomus: none Other reported: neurologic sequelae (paraesthesia, tactile hyperaesthesia), surgical wound healing impairment, surgical wound infection, impact on care provided to newborn/relationship, satisfaction score all at 3 months Adverse events: nausea, vomiting and anti‐emetic therapy requirements, incidence of pruritus, urinary retention, sedation, incidence of neurologic alterations (paraesthesia, tactile hyperaesthesia, headache)
Notes	Because no events were detected in either arm, we could not include the study in the meta‐analysis. Funding sources: "Dr Patricia O'Neill received speaker fees from Baxter Healthscore in 2010. B. Brain and Baxter were contacted simultaneously by authors to provide devices to perform the study. B Braun declined and Baxter showed interest and provided the devices for the study. Dr O'Neill helped design the study, conduct the study, analyse the data and write the manuscript and was paid by the company providing the devices for the study, to speak, after the study was finished being conducted but the results were not yet published. All four other authors reported no conflict of interest." Conflicts of interest: "we do not see a conflict of interest for the authors and no risk of bias of undue sponsor influence."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "computer‐generated random number list"
Allocation concealment (selection bias)	Low risk	Quote: " list concealed in an opaque envelope". Randomization was done after consent and prior to initiation of anaesthesia.
Blinding of participants and personnel (performance bias) All outcomes	High risk	The intraoperative and postoperative anaesthesia managers were not blinded, nor were the surgeons, This is acceptable for inclusion.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Three months after discharge, patients were interviewed by telephone by an investigator blinded to group assignment".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Per protocol analysis done, no ITT analysis. Number of participants in each group who were excluded is given, as well as the reasons for exclusion (e.g. accidental removal of catheter, did not receive allocated intervention, etc). Low overall attrition, fairly balanced numbers between groups.
Selective reporting (reporting bias)	Low risk	Primary outcomes listed in manuscript completely reported on. No protocol available for review.
Null bias	Low risk	Pain scores (quote:) "at rest at 2, 6, and 48 hours were lower in the continuous wound infusion group than in the epidural morphine group... (pain scores) evaluated at mobilization were higher in the epidural morphine group at 2 and 6 hours"