Yon 2014.

Methods	Randomized, controlled trial. Double‐blinded. The aim of this study was to assess the effect of intraoperative systemic lidocaine infusion in patients who underwent subtotal gastrectomy. The study was conducted in Korea between May 2012 and March 2013. (ACTRN12612000545864)
Participants	Number assessed for eligibility: 40 Number randomized: 36→ 17:19 Number analysed: 36→ 17:19 Inclusion criteria Adult patients (age 18 to 80 years) with subtotal gastrectomy Exclusion criteria Weight less than 45 kg or more than 100 kg; severe underlying respiratory, renal or hepatic disease; history of allergies to local anaesthetics, evidence of previous opioid medication or psychiatric medical history Baseline details Experimental group (n = 17) Age (years) (median): 59.00, IQR (57.00 – 66.50) M = %, F = % 1 patient missing male/female (n) (10:6) Mean weight (kg): 63.56, SD = 11.36 ASA I/II/III (n): 2:13:2 Mean duration of anaesthesia (min): 316.00, SD = 43.58 Mean duration of surgery (min): 271.47, SD = 33.11 Main surgical procedures (n): subtotal gastrectomy (17) Control group (n = 19) Age (years) (median): 66.00, IQR (59.00 – 72.00) M = 63.2%, F = 36.8% Mean weight (kg): 61.56, SD = 7.83 ASA I/II/III (n): 1:17:1 Mean duration of anaesthesia (min): 331.74, SD = 56.45 Mean duration of surgery (min): 291.32, SD = 50.47 Main surgical procedures (n): subtotal gastrectomy (19)
Interventions	Experimental group (17 patients) Patients assigned to the lidocaine group received an intravenous bolus infusion of 1.5 mg/kg of lidocaine followed by a continuous infusion of 2 mg/kg/hr (preoperatively and throughout surgery). Control group (19 patients) Patients in the placebo group received the same amount of normal saline.
Outcomes	The primary endpoint of the study was pain score. Dichotomous Nausea (overall) Vomiting (overall) Adverse events (shivering, tinnitus) Continuous Pain score at rest at 2, 4, 8, 12, 24, 48 hrs (VAS 0 to 100 mm, mean + SE, data presented graphically) Length of hospital stay (days, median + IQR) Patient satisfaction at 48 hrs (0 to 10, median + IQR) Total fentanyl consumption before discharge (µg, mean + SD) Frequency of pushing PCA button (number, median + IQR) Resume regular diet (days, median + IQR) CRP (mg/l, mean + SE) Fentanyl consumption (µg/hr, to 2, 4, 8, 12, 24, 48 hrs, mean + SE)
Notes	Small trial sample size (< 200 patients) Power analysis performed (VAS pain score 4 hrs, n = 16 per group) Medication The patients were not premedicated. No additional analgesics were injected during surgery. To control postoperative pain, intravenous fentanyl was administered with the use of a PCA system. The mode of PCA was a 0.3 µg/kg bolus with a lockout interval of 15 minutes, continuous infusion and 0.2 µg/kg/hr (total regimen of 100 ml) of fentanyl. In the case of persistent pain exceeding a visual analogue scale (VAS) pain score of 30 mm, an additional 50 µg of fentanyl (rescue) was intravenously injected by an investigator until the pain was relieved to a level falling below a VAS pain score of 30 mm. Anaesthesia The anaesthesia regime was standardized in both groups. Funding This work was supported by the 2012 Inje University research grant.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “random assignment was based on a random table generated using PASS software version 11 (NCSS). We used block randomization with a block size of 4 and equal allocation to prevent imbalances in treatment assignments. The randomization sequence was generated by a statistician who was not otherwise involved with the study.”
Allocation concealment (selection bias)	Unclear risk	Quote: “the details of the series were unknown to the investigators, and the group assignments were kept in sealed envelopes, each bearing only the case number on the outside.” Not explicitly mentioned that the envelopes were opaque and sequentially numbered.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: “to keep the anesthesiologist blind to the patients’ assigned group, lidocaine or placebo were prepared in a syringe and a bottle labelled only with a case number. The preparations of bolus and continuous infusions were arranged by an additional investigator (H.S.Y.) who read the card.” Quote: “all parties involved, including the patients, the surgeon, the anesthesiologists and the investigator (J.H.Y.) collecting the data were unaware of the study drugs or the patients’ group assignment.”
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: “The VAS scores were collected by 1 blinded investigator (J.H.Y.) with more than 2 years of experience interviewing patients about postoperative pain.” Quote: “all parties involved, including the patients, the surgeon, the anesthesiologists and the investigator (J.H.Y.) collecting the data were unaware of the study drugs or the patients’ group assignment.”
Incomplete outcome data (attrition bias) All outcomes	High risk	Dropout rate (experimental/control): 0%:0% Quote: “we used an intention‐to‐treat strategy — that is, all participants were included in the analysis irrespective of whether they had completed the study. Missing data were completed using a last observation carried forward analysis.” Quote: “data were incomplete for 3 patients. One patient in the lidocaine group and 1 patient in the placebo group were treated with other painkillers for PONV that was unresponsive to antiemetic treatment and likely induced by fentanyl infusion. One patient in the placebo group received meperidine owing to postoperative shivering. Despite incomplete data, these 3 patients were included in our analysis according to the intention‐to‐treat principle.” Reasons for missing data are likely to be related to true outcome.
Selective reporting (reporting bias)	Low risk	The study protocol is available (ACTRN12612000545864). It has been prospectively registered (22 May 2012, study start: May 2012). The primary and two secondary outcomes have been reported in the prespecified way. All other secondary outcomes were not prespecified but were not judged as selective reporting.
Other bias	Low risk	The study appears to be free of other sources of bias.