Liu 2015.

Methods	Assessor‐blinded, randomized clinical trial Sequence generation not described Follow‐up for 3 months
Participants	Participants: 120 adults in a university setting in China Operation: open thoracotomy 2 groups, size: 60/60 Age (± SD), group 1, 2: 61 (10), 58 (10) Men/women, group 1, 2: 33/27, 36/24 Exclusion criteria: paralysis, known allergy to LAs, active bacterial infection, clinically severe liver or kidney diseases, neurologic dysfunction, chronic use of systemic lidocaine, NSAIDs or opioids, insulin‐dependent diabetes mellitus and para‐aminobenzoic acid
Interventions	Group 1 (ropivacaine wound infusion): the moment participants entered the operating room, standard monitoring was performed by 5‐lead electrocardiography, pulse oximetry, and non‐invasive arterial pressure measurement. GA was induced with midazolam at 0.05 mg/kg, propofol at 1.5 mg/kg to 2.5 mg/kg and fentanyl at 3 µg/kg. When loss of consciousness was confirmed, a bolus of 0.8 mg/kg rocuronium was intravenously injected for tracheal intubation. Anaesthesia was maintained with continuous infusion of propofol and a bolus of fentanyl at 1 µg/kg/h to 2 µg/kg/h in order to keep the bispectral index monitor (BIS, Aspect 1000, Aspect Medical System Inc., Natick, MA, USA) between 40 and 60. Neuromuscular blockade was conducted by continuous infusion of cis‐atracurium at 0.06‐0.07 mg/kg/h. Participants in both groups were accessible to rescue analgesia via pethidine, if needed, during the postoperative period. The catheter was positioned in the SC tissues above the fascia along the inferior edge of the rib along the incision. The catheter consisted of a multi‐orifice tube that was connected to an elastomeric infusion pump (Beijing tech‐bio‐med medical equipment Corporation, China) for postoperative continuous SC infusion with an anaesthetic at the end of surgery. After skin closure, the infusion pump containing 0.5% ropivacaine (Naropin®‐produced by AstraZeneca) was connected, and the wound was infused at 2 mL/h. Group 2 (control): same intervention induction procedure as above. No catheter was inserted. Sufentanil was injected intravenously via an analgesia pump after surgery, followed by intravenous PCA with sufentanil at 2 mL/h Adjuvants: fentanyl Immediate post‐op pain control: no difference
Outcomes	Dichotomous: pain vs no pain Continuous: none Secondary: the level of sedation, severity of pain at rest and movement, the amount of opioid analgesics administered, and participants’ satisfaction with their postoperative pain management
Notes	We were unable to obtain additional information about randomization and blinding methods from the study author. Funding sources: "this work was supported by Natural Science Foundation of Jinling Hospital." Conflicts of interest: the study authors have no conflicts of interest to disclose.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomization technique not described
Allocation concealment (selection bias)	Unclear risk	Allocation of concealment not described
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding of participants and personnel not described
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "postoperative evaluations were performed by an observer blind to this study."
Incomplete outcome data (attrition bias) All outcomes	High risk	There was a substantial degree of attrition.
Selective reporting (reporting bias)	Low risk	ITT principle was used and no subgroup analysis was performed
Null bias	High risk	Quote "There were no statistical differences in the VAS scores... between the two groups"