3.1 Study selection and characteristics. After initial searching, a total of 2960 researches were included, and 2545 studies were included after removing duplicates. After reading the title and abstract, 2502 articles that did not meet the inclusion criteria were eliminated. After full text reading and screening of the remaining 43 articles, 13 articles[10–22] were finally included (including 1131 patients, 486 in the lidocaine group, and 489 in the control group) (Fig. 1). The characteristics of the included studies are shown in Table 1. The included randomized controlled trials were published between 2010 and 2022, with a mean age of 20 to 65 years and ASA grade I-III. Types of surgery were thyroid surgery, laparoscopic gastrointestinal surgery, laparoscopic gynecological surgery, video-assisted thoracoscopic surgery, and breast surgery. 13 included studies compared intravenous lidocaine with placebo (Normal saline).
3.2 Bias risk assessment. Most of the studies had a low risk of bias (Fig. 2). Only one study[10] had allocation concealment, blind method, and the integrity of the result data were not mentioned. In one study[17] did not blinding of outcome assessment. Neither researcher(L.F.Z. and G.L.) disagreed with the assessment of the risk of bias in either study.
3.3 Quality of Recovery. Six RCTs[11–13, 16, 17, 21] were included in the Meta-analysis of the quality of postoperative recovery. Meta-analysis showed that compared with the control group, perioperative systemic intravenous infusion of lidocaine significantly increased postoperative total QoR-40 score, which was conducive to postoperative recovery of patients(WMD: 4.98; 95%CI: 1.78 ~ 8.19; P = 0.002; I2 = 63%)༈Figure 3༉. The leave-one-out sensitivity analysis indicated that the study conducted by Kendall et al[12] may be the source of heterogeneity༈Figure 13༉. The trim and fill method was used to analyze the results of publication bias. The results showed that the WMD and 95%CI calculated by the fixed effect model were 0.402 (0.209–0.596) and 1.469 (1.233–1.814) before and after trim and fill (P < 0.001). The WMD and 95%CI calculated by random effect model were 0.920 (0.041–1.799) and 2.509 (1.041–6.047) before and after trim and fill (P values 0.04). The results showed that there was no significant change in the results (the WMD was a significant difference before and after trim and fill), publication bias had no significant effect on the conclusions of this study.(Fig. 14)
3.4 Postoperative analgesia
3.4.1 Postoperative VAS scores. Three RCTs[10, 15, 18] evaluated the pain at 12 and 24 hours after surgery, respectively. Meta-analysis results showed that, in the lidocaine intravenous infusion group, VAS scores at 12 hours (WMD: -0.31; 95% CI: 0.69 ~ 0.07; P = 0.11; I2 = 64%; Egger's publication bias P = 0.985) and 24 hours (WMD: -0.18; 95% CI: 0.43 ~ 0.04; P = 0.02; I2 = 34%; Egger's publication bias P = 0.204) after surgery had no significant difference(Fig. 4). According to the leave-one-out sensitivity analysis, the study conducted by Chen et al[10] may be the source of heterogeneity.
3.4.2 Postoperative NRS scores at 0.5 hours. NRS pain scores in the rest state and movement state were assessed at 0.5 hours after surgery by Four RCTs[14, 19, 20, 22]. In the lidocaine intravenous infusion group, NRS scores in both rest (WMD:-0.70; 95%CI: -1.05~-0.35; P = 0.0001; I2 = 0%; Egger's publication bias P = 0.491)and movement state (WMD:-0.76; 95%CI:-1.18~-0.34; P = 0.0004; I2 = 9%; Egger's publication bias P = 0.945) decreased significantly at 0.5 hours after surgery, with no significant heterogeneity(Fig. 5A).
3.4.3 Postoperative NRS scores at 24 hours and 48 hours. Four RCTs[14, 19, 20, 22] and three RCTs[19, 20, 22] evaluated postoperative NRS scores at 24 hours, and 48 hours respectively. Meta-analysis results showed that, perioperative systemic intravenous infusion of lidocaine had no significant effect on NRS scores at rest(24h: WMD:-0.04; 95%CI:-0.38 ~ 0.29; P = 0.81; I2 = 54%;Egger‘s publication bias P = 0.415) (48h: WMD: -0.08༛95%CI: -0.19 ~ 0.02; P = 0.13; I2 = 24%༛Egger‘s publication bias P = 0.355) and movement state(24h: WMD: -0.13; 95%CI:-0.28 ~ 0.02; P = 0.08; I2 = 0%; Egger‘s publication bias P = 0.232)(48h: WMD:-0.03; 95%CI:-0.09 ~ 0.04; P = 0.40; I2 = 0%; Egger‘s publication bias P = 0.792 ) at 24 hours and 48 hours after surgery.(Figs. 5B, 5C). The leave-one-out sensitivity analysis at 24 hours postoperative indicated that the study conducted by Lv et al[14] may be the source of heterogeneity.
3.5 Postoperative nausea and vomiting. Three RCTs[10, 11, 14] were included to study the incidence of postoperative nausea and vomiting after intravenous lidocaine or placebo. Meta-analysis showed that compared with the control group, the incidence of postoperative nausea was significantly reduced in the lidocaine group(RR: 0.37; 95% CI: 0.21 ~ 0.65; P = 0.0006; I2 = 0%). However, there was no significant difference between the lidocaine group and the control group in the incidence of postoperative vomiting (RR: 0.43; 95% CI: 0.17 ~ 1.08; P = 0.07; I2 = 0%)(Fig. 6).
3.6 Bowel function recovery. Five RCTs[10, 15–18] reported time to recovery of postoperative bowel function (time to first flatus or defecation). Meta-analysis results showed that the postoperative bowel function recovery time in the lidocaine group was significantly shorter than that in the control group (WMD: -3.85; 95% CI: 5.94 ~ 1.75; P = 0.0003; I2 = 93%) (Fig. 7). The leave-one-out sensitivity analysis showed that the studies of Chen et al.[10] and Wang Lei et al. [16] may be the source of heterogeneity.
3.7 Length of hospital stay. Three RCTs[10, 14, 17] were included to report length of hospital stay. Meta-analysis showed that there was no significant difference between two groups in length of hospital stay (WMD: -0.20;95%CI: -0.73 ~ 0.33; P = 0.46; I2 = 47%) .