Senturk 2002

Methods	Single‐blind (outcome assessor), clinical RCT Sequence generation was random, but not described Follow‐up: 6 months
Participants	Participants: 112 adults at a university setting in Istanbul, Turkey Operation: open thoracotomy for a mix of lung resections 3 groups, size: 28/29/28 Age (group 1, 2, 3): 49 (SD 9), 52 (SD 11), 50 (SD 11) years Men/women: 56/13 (reported at end of study) Comorbidities: not reported
Interventions	Group 1 (preincision): epidural at T7‐8, preincision bupivacaine bolus 10 mL, 7 mL/h infusion (0.1% + 0.1 mg/mL morphine), GA, post‐op 48 h PCEA (0.1% bupivacaine + 0.05 mg/mL morphine, basal rate 5 mL/h, demand 3 mL, lockout 30 min) Group 2 (postsurgery): epidural at T7‐8, GA (fentanyl), postsurgical bupivacaine bolus 10 mL (0.1% + 0.1 mg/mL morphine), post‐op 48 h PCEA (0.1% bupivacaine + 0.05 mg/mL morphine, basal rate 5 mL/h, demand 3 mL, lock time 30 min) Group 3 (control): GA (fentanyl), PCA (morphine, bolus 5 mg, no basal rate, demand 2 mg, lockout 15 min) Adjuvants: none Immediate post‐op pain control: significantly improved
Outcomes	Dichotomous: pain at 6 months, pain affecting daily life at 6 months Continuous: NRS at 6 months Secondary: none
Notes	Regional anaesthesia catheter placement was verified under fluoroscopy. The study author responded and provided additional information regarding randomization allocation concealment, sources of funding and conflicts of interest. Funding sources: "the study was not funded" Conflicts of interest: the authors "have no conflict of interest"
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were "randomly divided into three groups", "using sealed envelopes technique."
Allocation concealment (selection bias)	Low risk	Quote: “randomization was performed at the first presentation of the patient to our department, i.e. 5‐7 days before the operation (just before the anaesthetic evaluation). The result of the randomization was "hidden" by the secretary of the department until the operation date.”
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "patients were not blinded to group, anaesthesia providers aware of allocation at least during treatment."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome assessors "were blinded to the analgesic method." Blinding of only outcome assessors is acceptable.
Incomplete outcome data (attrition bias) All outcomes	High risk	Allocation of excluded participants is not reported, no ITT analysis was considered. Considerable attrition prior to, during and after intervention make bias likely. Adverse effects were not, but attrition was described albeit without group allocation 27 participants were excluded preoperatively, 6 intraoperatively, and 10 postoperatively, without specification of their group allocation. Comorbidities were the preoperative, inoperability the intraoperative and recurrence of pain due to metastasis & reoperation were the postoperative exclusion criteria.
Selective reporting (reporting bias)	Low risk	All expected outcomes included
Null bias	Low risk	Quote: "during movement and cough, Group Pre‐TEA had significantly less pain compared with the other two groups during the entire period. At rest, patients in Group Pre‐TEA reported having significantly lower pain scores during the first 12 h compared with those in Group Post‐TEA and during the first 48 h compared with those in Group IV‐PCA. There were statistically significant differences between Group Post‐TEA and Group IV‐PCA during rest from 8 h after surgery until the end of 48 h, but no difference during cough or movement was recorded"