Bollag 2012.

Methods	Triple‐blinded (participant, provider, outcome assessor) RCT Sequence generation with computer‐generated list of random numbers Follow‐up: 12 months
Participants	Participants: 90 healthy non‐labouring pregnant women from Maternity Hospital in Sao Paulo, Brazil Operation: caesarean delivery, scheduled (under SA with Pfannenstiel incision) Three groups, size: 30/25/26 Age (± SD), group 1, 2, 3: 30.5 (± 6.7), 31.8 (± 4.5), 29.5 (± 6.7) Only female participants Comorbidities: previous caesarean delivery (%), group 1, 2, 3: 46/48/35. Gestational age in weeks, mean (± SD), group 1, 2, 3: 38 (± 1), 38 (± 1), 38 (± 1.5)
Interventions	Group 1 (placebo/control): TAP block with 20.5 mL 0.9% NaCL per side. Group 2 (bupivacaine TAP): TAP block with 20 mL bupivacaine 0.375% + 0.5 mL NaCl 0.9% per side. Group 3 (bupivacaine + clonidine group): TAP block with 20 mL bupivacaine 0.375% + 75 µg (0.5 mL) clonidine per side All TAP blocks were performed in PACU within 1 h post‐op All groups: spinal anaesthetic with 12 mg hyperbaric bupivacaine, 25 µg fentanyl, 100 µg morphine. IV ketoralac at skin closure. Post‐op analgesia: in PACU, IV morphine as needed; in postpartum unit paracetamol (1 g every 6 h standing) and diclofenac (75 mg every 8 h standing), with tramadol 50 mg as needed Adjuvants: clonidine (group 3 only) Immediate post‐op pain control: significantly reduced morphine use in TAP groups compared to placebo in PACU but no change in resting pain scores. Effective regional anaesthesia: reported. "Block success and dermatomal extent of the sensory analgesia were assessed bilaterally by pinprick after recovery from the spinal anaesthetic".
Outcomes	Dichotomous: pain/no pain at 3, 6, 12 months Continuous: short‐form McGill Pain questionnaire at 3, 6 and 12 months
Notes	We contacted the study author who provided dichotomous pain data for 3, 6, and 12 months' follow‐up. Funding sources: no financial support was received for the study. Conflicts of interest: "the authors declare no conflict of interest."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "a computer‐generated list of random numbers was used (www.randomizer.org) for group allocation of the participants".
Allocation concealment (selection bias)	Low risk	Quote: "each woman was assigned a study number upon enrolment and received a TAP block with the corresponding numbered syringe. The allocation sequence was concealed from investigators and patients". While it does not state method with which allocation was concealed, it states it was concealed thus little risk of bias.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "an investigator with no clinical involvement in the trial prepared the solutions following exact preparation guidelines. All syringes were labelled with the amount and concentrations of all possible contents, as well as a study number. Both operator [who performed TAP block] and patient were blinded to the study group."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "hyperalgesia was evaluated by the same research investigator (who was not involved in placement or evaluation of the TAP blocks in the PACU)". "At 3, 6, and 12 months, telephone interviews were performed to assess development of chronic postoperative pain using the Short‐Form McGill Pain Questionnaire 2 (SF‐MPQ‐2)". While it does not explicitly state chronic pain assessment was performed by a blinded investigator, based on the other descriptions of how participants were assigned to groups and blinding was maintained, it seems very unlikely the telephone interviewers knew which group they were assigned to.
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "five women from [group 2] and 4 women from [group 3] were excluded from the study because of block failure (absence of sensory block on the abdomen assessed by pinprick after recovery from the spinal anesthetic)". No ITT analysis was performed, only per‐protocol. Flow diagram depicts loss of follow‐up for each group at 3‐, 6‐, 12‐month periods, with 2 participants in the control, 6 participants in [group 2] and 5 participants in [group 3] lost at 12 months, and fewer in each group at 3 and 6 months. SF‐36 survey reports "return rate" at each time point in terms of percent but does not provide raw numbers. Discordance between flow diagram and numbers included in analysis in neuropathic pain descriptors (table 4)
Selective reporting (reporting bias)	Low risk	Protocol reviewed and primary outcomes fully reported on
Null bias	High risk	Quote: "the incidence of wound hyperalgesia and the WHI were similar among groups at 24 hours (Fig. 2). At 48 hours, the incidence of wound hyperalgesia was not different among groups".