Pharmacological Prevention Of Sevoflurane-related Emergence Agitation In Chidren: A Meta-analysis

Objective:To provide a meta-analysis of the studies on the efficacy of various drugs in preventing sevoflurane-related emergence agitation (EA) in children.Method:Articles in Chinese or English published in Pubmed, EMBASE, Springer, Journals@Ovid Full Text, Sciencedirect OnSite Database, Chinese Biomedical Database, Chinese Journals Full-text Database, China Doctor Dissertation Full-text Database, Chinese Conference Dissertation Full-text Database were searched up to February 2010, in which the following search terms were used:" sevoflurane, agitation and children or infant", " sevoflurane, delirium and children or infant", and the studies in section "references" were also searched. Strict inclusive and exclusive criteria were established according to Corchrane Handbook. Randomized clinical trial (RCT) would be assessed for qualities according to the Jadad scale and the concealment of treatment allocation. The trial design, participant, type of intercention and outcome measurement in the obtained articles, where the number of the articles was more than 2 pieces on one specific agent, were independently assessed by two reviewers. The Revman 4.2.2 software provided by Cochrane collaboration was used to review management and analysis. The results were expressed as odds ratio (OR) and confidence internals (CI) for dichotomous outcomes, and forest plot was obtained. I2> 40% and P<0.1 were used to indicate heterogeneity and the funnel plot and fale-safe number to assess bias. The sensitivity analysis were carried out according to characteristics of the studies involved.Results:42 articles and a total of 3029 patients (1519 in the treatment groups and 1510 in the control ones) were identified according to the search criteria and 34 articles were included in the Meta-analysis. The methodological quality of the articles was relatively high. Midazolam were not found to have a protective effect against EA [P=0.49, OR=0.68,95%CI was 0.23-2.01], whereas propofol [OR=0.27 (0.17,0.43)], ketamine [OR=0.13 (0.14,0.37)], fentanyl [OR=0.28 (0.16,0.48)], clondine [OR=0.24 (0.11,0.53)], dexmedetomidine OR=0.19 (0.11,0.33)] were found to have preventive effects. Subgroup analysis according to the preoperative analgesia given did not affect the results. There were not significant heterogeneity within the articles on propofol, ketamin and dexmedetomidine, whereas heterogenetity of studies on midazolam, fentanly and clonidine were significant. Fale-safe number indicated the possibility of published bias of midazolam. The sensitivity analysis of propofol, ketamin, fentanly and clondine showed that the Meta-analysis results of these pharmacological prevention reliable.Conclusions:This Meta-analysis found that propofol, ketamine, fentanyl, clondine and dexmedetomidine had prophylactic effects in preventing EA related sevoflurane. The preventive property of midazolam needs to be addressed in further studies that are well designed.