Can 2013.

Methods	Double‐blind, clinical RCT Randomization using "the envelope method" but no report on sequence generation technique. Follow‐up: 6 months
Participants	Participants: 60 adult participants from university‐affiliated hospital in Turkey Operation: thoracotomy, elective 3 groups, size: 20/20/20 Age (± SD), group 1, 2, 3: 52.20 (± 17.05), 45.00 (± 17.46), 50.9 (± 16.12) Men/women, group 1, 2, 3: 15/5, 15/5, 15/5 Comorbidities: no concomitant disease
Interventions	Group 1 (control): preoperative and intraoperative analgesia with 0.25 µg/kg/h to 0.60 µg/kg/h remifentanil infusion. No epidural analgesic medication before or during operation through epidural catheter Group 2 (incision‐sensitized): preoperative analgesia with 0.25 µg/kg/h to 0.60 µg/kg/h remifentanil infusion. 10 min after surgical incision, epidural admin 10 mL to 15 mL 0.1% levobupivacaine and remifentanil infusion then remifentanil continued for 20 more min for a total of 30 min then 10 mL 0.1% levobupivacaine epidural every 45 min Group 3 (pre‐emptive analgesia group): preop analgesia: 0.1% levobupivacaine 10 mL to 15 mL at 2nd dermatome superior and inferior to incision dermatome (between T4 to T10) through epidural catheter prior to induction. Intraop analgesia: 10 mL 0.1% levobupivacaine epidural injection every 45 min. In all groups epidural catheters were placed preoperatively at 6th‐7th or 7th‐8th thoracic intervals. All received the same GA regimen. Postoperatively all received morphine (3 mg) + fentanyl (50 µg) in 15 mL isotonic solution via epidural route at skin closure and every 12 h for 48 h Adjuvants: none Immediate post‐op pain control: not significantly improved
Outcomes	Dichotomous: pain/no pain at 3 and 6 months Continuous: VAS score 3 and 6 months Other reported: participant satisfaction levels at discharge and at month 6
Notes	Presence of chronic pain defined as VAS score > 3. Epidural catheters were placed in all participants, and after placement a 3 mL test dose of 2% lidocaine with 1/200,000 adrenalin was injected. Thus, all participants did receive small amount of lidocaine via epidural catheter. We acknowledge the study author's response on allocation concealment, blinding, source of funding and whether there was any conflict of interest. Funding sources: response from study author, "the authors declare... [their] university... funded this study" Conflicts of interest: "the authors declare... that they have no conflict of interest to the publication of this article..."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomized envelopes drawn "when patient come to operation room a staff get an envelope and open it", from study author
Allocation concealment (selection bias)	Low risk	On questioning, study author responded "Envelopes are opaque and include equal groups symbols. When patient come to operation room a staff get an envelope and open it."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "double blind" study. When questioned, study author responded "The personal collecting the pain data was not involved in the previous study phases"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "the outcome assessor collecting pain levels postoperatively and at 1, 3, 6 months was blinded" says the study author
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "2 patients from control group and 1 patient from preemptive analgesia group died and 1 patient from preemptive analgesia and other one patient from incision sensitized group wound infection were excluded" stated author. "New participants that were compliant with the inclusion criteria were enrolled."
Selective reporting (reporting bias)	Low risk	No protocol available but all specified outcomes were reported on.
Null bias	High risk	Table 3 demonstrates no significant difference in VAS scores between the 3 groups at hours 1, 4, 24 or 48 after surgery