Weber 2007.

Methods	Single‐blinded (outcome observer) clinical RCT Sequence generation via computer‐generated randomization list Follow‐up: 6 months
Participants	Participants children and adolescents ≥ 10 years at a university hospital in Vienna, Austria Operation: pectus excavatum repair (minimally invasive using a thorascope for creation of retrosternal tunnel) 2 groups, size: 20/20 Age (± SD), group 1, 2: 16.7 (± 5.2), 14.8 (± 4.2) Men/women, group 1, 2: 17/3, 15/5 Comorbidities: except for 1 participant in TEA group, all procedures were primary operations. Vertebral index (vertebral diameter x 100/sagittal diameter + vertebral diameter), group 1, 2 (± SD) = 32.05 (± 36.2), 31.85 (± 4.15)
Interventions	Group 1 (PCA): post‐op IV PCA 0.02 mg/kg morphine bolus, lockout 6 min, max 6 bolus/h, no continuous rate. Postoperatively, both groups 1 mg/kg diclofenac IV every 8 h scheduled until POD 4, rescue pain medication with IV paracetamol 15 mg/kg, followed by 1.5 mg piritramide IV bolus as needed Group 2 (TEA): catheter placed once in operating room by median approach at T6/7 or T7/8 corresponding with likely insertion site of steel bar. After induction, bolus of 0.2 mg/kg ropivacaine 0.2% with 2 µg/mL fentanyl, then continuous rate of 0.2 mL/h same mixture throughout surgery, continued until POD 4 (96 h). Post‐op scheduled 1 mg/kg diclofenac IV every 8 h until POD 4 rescue pain medication with IV paracetamol 15 mg/kg, followed by epidural bolus of 0.1 mL/kg ropivacaine 0.2% with 2 µg/mL fentanyl as needed. Both groups received standardized GA with propofol, fentanyl, rocuronium. 15 min before end, IV paracetamol bolus Adjuvants: none Immediate post‐op pain control: significantly improved
Outcomes	Dichotomous: pain/no pain at 3 and 6 months Continuous: VAS pain score 3 and 6 months Secondary: satisfaction with type of anaesthesia at 3 and 6 months Adverse events reported: sedation, nausea, pruritis
Notes	Presence of pain defined by VAS ≥ 3. We acknowledge the study author for providing response regarding VAS cutoff for presence of pain, allocation concealment, blinding and source of funding. Funding sources: "AstraZeneca, Bristol Myers‐Squibb, and Smiths Medical Austria supported the study with an unrestricted grant". We contacted the study author on their specific involvement, who responded, "Funding by the three companies included just paying for the insurance (approximately one third by each company). None of the companies were involved in conducting the study or writing the manuscript." Conflicts of interest: no direct conflicts of interest statement given
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "computer generated randomization list"
Allocation concealment (selection bias)	Low risk	Study author specified "Group allocation was concealed in an opaque envelope"
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants, surgeons and providers were not blinded. The study author clarified that "the PCA pump and the TEA continuous infusion (depending on the study group) were hidden from the persons assessing the VAS scores".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The study author stated "For postoperative data collection, the PCA pump and the TEA continuous infusion (depending on the study group) were hidden from the persons assessing the VAS scores. The persons who made the follow up questioning [at 3 and 6 months] were unaware to which group the patients were assigned".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Study author specified "All 40 patients were available at three and 6 months for follow‐up".
Selective reporting (reporting bias)	Low risk	Primary outcomes fully reported on
Null bias	Low risk	Quote: "Patients treated with a thoracic epidural catheter after pectus excavatum repair reported lower postoperative pain scores... than did patients treated with intravenous PCA containing morphine. Postoperative pain scores in the intravenous PCA group were higher despite higher intraoperative fentanyl use in the intravenous PCA group"