Caudal anesthesia has been used for postoperative pain management of pediatric patients undergoing infraumbilical procedures. It can be administered quickly and with a high success rate,1 and its usage has become popular. Although an abdominal wall block is also popular due to the widespread use of an ultrasound-guided technique, the caudal block still has some advantages. It does not require additional equipment such as an ultrasound device or additional training to perform the ultrasound-guided technique. A problem of caudal analgesia is the short duration of the analgesic effect; thus, many drugs have been examined for their potential to prolong the duration of the analgesic effect.2–5 Dexamethasone has been reported to prolong the duration of caudal analgesia, either by mixed administration with local anesthesia6,7 or by intravenous systemic administration.8,9 However, there are concerns regarding the safety of caudal steroid administration. Serious complications of caudal or epidural steroids have been reported when a steroid has been administered to patients with chronic lower back pain.10,11 There is still a lack of data demonstrating the safety of potent steroid administration near nerve structures in a growing child.12 There are also safety concerns regarding crystal formation due to mixing dexamethasone with ropivacaine.13 In addition, it is still unknown whether perineurally administered steroids act locally or systemically.14 However, perioperative intravenous dexamethasone has been commonly used for the prevention of postoperative nausea and vomiting (PONV),15,16 and its use was not associated with clinically relevant toxicitity.16
Several studies have demonstrated that intravenous dexamethasone prolonged analgesic duration in patients who received caudal analgesia,8,9 but the increased duration varied among studies. The actual effects and optimal dosages have yet to be fully elucidated. So far, no systematic review has specifically addressed the effects of intravenous steroids in children who received caudal anesthesia.
The aim of this systematic review was to determine the effectiveness of intravenous steroids for the improvement of pain in pediatric patients receiving caudal anesthesia. We also investigated if the use of steroids increases the risk of a postoperative adverse event.
This study is a systematic review and meta-analysis with trial sequential analysis (TSA).
We followed the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement17 and the Cochrane Handbook.18 Our study protocol and analysis methods were prespecified and registered in the University Hospital Medical Information Network (UMIN) Clinical Trials Registry.
The literature databases MEDLINE, Cochrane Register of Controlled Trials (CENTRAL), EMBASE, and Web of Science were searched without language restrictions. The reference lists of the retrieved full articles were also searched. In addition, we conducted a search on the clinicaltrials.gov website and in the University Hospital Medical Information Network Clinical Trials Registry. The PubMed search strategy is provided in the supporting information (Supplemental Digital Content 1, Search Strategy for PubMed, https://links.lww.com/AA/B993). The most recent search was performed on November 20, 2016.
Two authors (H.K. and N.N.) independently examined the titles and abstracts of reports identified by the search strategies described above to exclude clearly irrelevant articles. The complete article was retrieved if eligibility could not be determined from the title or abstract. Potentially relevant studies, chosen by at least 1 author, were retrieved, and full-text versions were evaluated. The articles that met the inclusion criteria were assessed separately by the 2 authors, and any discrepancies were resolved through discussion.
We searched for any randomized controlled trials that tested the efficacy of intravenous steroids added to caudal local anesthesia compared with a placebo in postoperative pain management. We excluded studies in which patients were not administered caudal local anesthesia. We excluded studies that did not investigate postoperative pain and studies in which the subjects were not surgical patients. We excluded nonpediatric patients (>18 years of age). We also excluded data from case reports, comments or letters to the editor, reviews, and animal studies.
Primary and Secondary Outcomes
The 2 primary outcomes from the present meta-analysis were analgesic duration and the need for a rescue analgesic. The secondary outcomes were pain scores during postoperative 12 hours and adverse events including sedation, urinary retention, paresthesia, wound complications, high blood glucose, and prolonged motor block.
A data collection sheet was created and included data on the following: (1) the number of patients in the study, (2) age, (3) the American Society of Anesthesiologists physical status, (4) type of anesthesia, (5) type of surgery, (6) dose of steroid, (7) type of steroid, (8) dose and type of medication used for caudal anesthesia, (9) time to rescue analgesic (analgesic duration), (10) number of cases of rescue analgesic in postanesthesia care unit, (11) number of cases of rescue analgesic within the postoperative 24 hours period after PACU discharge, (12) postoperative pain score, (13) duration of motor blocks, and (14) adverse events. Two authors (H.K. and N.N.) extracted the data independently from the included studies using piloted form and cross-checked the data. If data were reported only in graphs that indicated pain score, we measured the length of the graphs to obtain the score. For data not available in the literature, attempts were made to contact the respective authors.
Assessment of Risk of Bias in Individual Studies
We assessed the risk of bias as described by the Cochrane Handbook for Systematic Reviews of Interventions.19 We assessed the risk of bias in sequence generation, allocation sequence concealment, blinding of patients, blinding of health care providers, blinding of data collectors, blinding of outcome assessors, incomplete outcome data, selective outcome reporting, and other bias. The risk of bias was classified into 3 categories: “low,” “high,” or “unclear.” Two authors (H.K. and N.N.) evaluated the risk of bias in each trial. When there was a difference in the evaluation of bias, the 2 authors discussed the evaluation and reached a consensus. Trials with 1 or more risk of bias domain that was unclear or at high risk of bias were considered to be trials at high risk of bias.
Assessment of Quality of Evidence
We graded the quality of evidence of the main outcomes using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach. Judgments of the quality of evidence were based on the presence or absence of the following variables: risk of bias, inconsistency, indirectness, imprecision of the results, and publication bias. The quality of evidence for the main outcomes was graded as very low, low, moderate, or high. We formulated a summary of findings table using GRADEpro GDT.20
Analgesic duration and the incidence of rescue use were summarized using mean difference (MD) or risk ratio (RR) with a 97.5% confidence interval (CI), respectively. If the 97.5% CI included a value of 0 or 1, respectively, we considered the difference between the steroid and control groups to be not statistically significant. As there were 2 primary outcomes, we used a 97.5% CI to adjust for multiple testing. Pain scores were summarized using the MD with a 95% CI. If the 95% CI included a value of 0, we considered the difference to be not statistically significant. Heterogeneity was quantified with the I2 statistic. We used the random effect model (DerSimonian and Laird21 method) to combine the results. Forest plots were used to graphically represent and evaluate the effects of treatment. Small-study effects were assessed using a funnel plot and an Egger’s regression asymmetry test22 and were considered to be positive if P < .1 in the regression asymmetry test. Sensitivity analyses were performed for the primary outcomes according to the risk of bias (low versus high).
For our primary outcomes, TSA was performed. The addition of each trial in meta-analysis is analogous to interim analysis in clinical trials, and repeated significance testing can cause a false-positive result. TSA is a statistical method developed to avoid spurious significant results in meta-analysis.23 First, the TSA calculates a heterogeneity-adjusted target sample size called the “required information size” (RIS), which is determined based on detecting a predetermined clinically meaningful effect. The RIS is analogous to the target sample size in randomized controlled trials. Second, the TSA monitoring boundary is constructed using an alpha spending function.24,25 Third, a Z statistic is calculated for each trial, and a cumulative Z curve is constructed.25 When the cumulative Z curve crosses the TSA monitoring boundary for benefit or harm, a firm conclusion can be drawn that there is a significant difference before achieving the RIS. When the cumulative Z curve crosses the futility boundary, we can reject the hypothesis and conclude that the difference between the 2 groups is not significant.
TSA monitoring boundaries (ie, monitoring boundaries for meta-analysis) and RIS were quantified, and adjusted CIs were calculated for each outcome. The TSA-adjusted CI is more conservative than the conventional 95% CI. In the TSA, we established the risk of type 1 error to be 2.5% and the risk of type 2 error to be 10%. A difference of 3 hours in the analgesic duration or a RR of 0.75 in the incidence of rescue use was considered minimum clinically meaningful differences. If the TSA-adjusted CI of the MD or RR included a value of 0 or 1, respectively, indicating that the cumulative Z curve did not cross the TSA monitoring boundary, we considered the difference to be not statistically significant. Statistical analyses were performed using the R statistical software package, version 3.3.0 (R Foundation for Statistical Computing, Vienna, Austria). TSA was performed using TSA Viewer, Version 0.9.5.5 β (www.ctu.dk/tsa). There was a problem in the previous version of TSA viewer; we used the updated version that had been corrected.
A total of 1024 publications were identified. Of those, 6 studies were included in this review.8,9,26–29 The PRISMA flow diagram detailing the disposition of retrieved publications is shown in Figure 1. The evaluated trials included data from 424 subjects; 211 of them received intravenous steroids.
The features of the randomized studies included in this meta-analysis are listed in Table 1. The intravenous steroid used in each of the 6 trials was dexamethasone; the patients received intravenous dexamethasone or a placebo between induction of general anesthesia and surgical incision in all trials. The dose of dexamethasone ranged from 0.5 to 1.5 mg/kg.
All 6 trials reported duration of analgesia. The end of the analgesic duration was the point at which rescue acetaminophen was administered for pain; the various types of pain scale were used. The beginning of the duration varied among studies, and these included the time of caudal injection, the end of surgery, and 5 minutes after extubation. One trial did not report the mean and standard deviation27; we contacted the corresponding author and obtained the information. The combined results are shown in Figure 2A. Patients receiving dexamethasone had significantly longer analgesic duration than patients who received a placebo (MD, 244 minutes; 97.5% CI, 188–300; I2, 95.7%).
The TSA revealed that the accrued information size (n = 424) reached only 17.0% of the estimated RIS (n = 2499). The cumulative Z score crossed the trial sequential monitoring boundary for benefit (Figure 2B). TSA-adjusted CI was 43–444.
Need for Rescue Acetaminophen
Four trials reported the number of patients who received postoperative rescue acetaminophen.8,9,27,28 The combined result is shown in Figure 3A. There was no significant difference (RR, 0.53; 97.5% CI, 0.06–3.30), and considerable heterogeneity was noted with an I2 value of 98.7%. The RIS was 2090, and the accrued information size (n = 260) reached only 12.4% of the estimated RIS. The cumulative Z score did not cross the trial sequential monitoring boundary for benefit (Figure 3B). The TSA-adjusted CI was 0.06–5.37.
The pain score was reported in 3 trials. The pain scales used varied among the trials, and pain was measured at various time points. Hong et al8 used reported better pain control in the intervention group at 1 hour after surgery, but no difference at 3 hours. Abd-Elshafy et al9 demonstrated that pain was less in the intervention group at 4 and 8 hours after surgery and that at 12 hours, pain scores were not significantly different. In the trial by Modi et al,29 pain control was significantly better at 8 hours after surgery in the intervention group than the control, but not at 1, 2, 4, or 12 hours after surgery.
Hong et al8 and Modi et al29 used a Faces, Legs, Activity, Cry, Consolability score to measure pain. They both recorded pain scores at postoperative 1 and 2 hours. The combined results of the pain score are detailed in the supporting information (Supplemental Digital Content 2, Data, https://links.lww.com/AA/B994). There was no significant difference in pain score between groups at postoperative 1 and 2 hours. Abd-Elshafy et al9 reported the median pain score with an objective pain scale. They did not report standard deviation or interquartile range, and we could not combine their results.
Adverse Events and Other Outcomes
Adverse events were not recorded in 3 trials. In 2 trials, there was no significant difference of adverse events between the groups.8,29 In 1 trial, blood glucose levels were measured postoperatively and showed no complications such as increased blood glucose, delayed wound healing, or wound infection.9 The motor block was assessed in 1 trial, and no difference was reported between groups.9 The number of cases of rescue analgesic in PACU was measured in only 1 trial, and significantly more patients in the control group required rescue than in the dexamethasone group (38.5% vs 7.9%; P < .01).8
Sensitivity analysis was not performed for the primary outcomes, as only one of the 6 trials was at low risk of bias. All trials reported a significantly longer analgesic duration in the intervention group compared to the placebo.
We could not perform an asymmetry test for the funnel plot because only 6 trials were included.
The Risk of Bias of the Included Trials
The risk of bias in the included trials is summarized in Table 2. Only 1 trial was considered to be at low risk of bias; the rest were considered to be at high risk of bias.
Quality of the Evidence
The quality of the evidence of the effect of intravenous dexamethasone on analgesic duration as compared with a placebo was graded as “very low” (Table 3). It was downgraded due to limited study design, inconsistency, and possible publication bias. We did not detect imprecision or indirectness. The quality of the evidence for the reduction in the number of patients requiring postoperative rescue analgesia was graded as very low. It was downgraded due to limited study design, inconsistency, and publication bias.
Our meta-analysis showed that intravenous dexamethasone significantly prolonged the analgesic duration of caudal anesthesia in children (GRADE: very low). However, our meta-analysis failed to demonstrate the reduction in the number of patients requiring rescue analgesics (GRADE: very low). The GRADE of both primary outcomes was very low; thus, our meta-analysis should be considered hypothesis generative. High-quality trials are needed to reach a firm conclusion.
The MD of analgesic duration was approximately 4 hours longer in the dexamethasone group than in the control group, which we consider clinically significant. Administration of a systemic corticosteroid has been reported to reduce tissue bradykinin30 and neuropeptides around injury sites,31 but the mechanism by which a corticosteroid exerts an analgesic effect in surgical patients is still not understood. Clinically, intravenous steroids have been investigated for postoperative pain management and have proven to be effective in adults32,33 and in pediatric tonsillectomy patients.34 Intravenous dexamethasone has also been shown to increase the analgesic duration of peripheral nerve blocks by several hours.35,36 Shirazi et al37 demonstrated that intravenous dexamethasone increased the analgesic duration of penile block in pediatric urologic patients by 5 hours. The baseline analgesic duration of a peripheral nerve block is longer than that of a caudal block, and the increase in duration using additional intravenous dexamethasone appears longer in a peripheral nerve block than in a caudal block. Although the result of the TSA was positive, the GRADE was evaluated as very low because 5 of the 6 trials were at high risk of bias, considerable heterogeneity was noted, and the possibility of publication bias could not be denied. This meta-analysis should be considered hypothesis generative, and further studies with low risk of bias may give us a better estimate of how long the analgesic duration is increased by intravenous dexamethasone.
Considerable heterogeneity existed among our selected studies and was attributable to various clinical factors. The analgesic durations of the control group varied among the trials. That variability may be due to different surgical procedures, different local anesthetics, or different concentrations and amounts of local anesthetics. Intraoperative pain medication varied among trials. In a trial by Murni Sari Ahmad et al,27 all patients received suppository acetaminophen after caudal block, but patients in other studies did not. The analgesic duration ended when the rescue analgesic was administered in all included trials, but the decision to administer the rescue analgesic was based on different pain scales among the trials. Each of these factors may have contributed to the high heterogeneity. We could not explore the cause of heterogeneity by conducting a subgroup or meta-regression analysis because the total number of included trials was not enough to conduct these additional analyses.18,38 Despite the heterogeneity observed, all of the clinical trials included demonstrated significant improvement in postoperative pain; heterogeneity was due to the variability in increase of the analgesic duration.
Our meta-analysis failed to show the effectiveness of intravenous dexamethasone for reducing the number of patients who required rescue analgesics within 24 hours after surgery despite the significant reduction in analgesic duration in the dexamethasone group. There are 2 possible explanations for the conflicting results. First, our meta-analysis did not have adequate power to detect the effect of dexamethasone on reduction of rescue use. The point estimate of the RR for the incidence of rescue use was as low as 0.53, but it did not reach a statistically significant level because the 97.5% CI was very wide. The result of the TSA also suggested that this meta-analysis did not have an adequate sample size to detect the difference. The effect of dexamethasone on reducing rescue use might be confirmed after accumulating additional evidence. Second, the RR of the incidence of rescue use may be an inappropriate outcome to evaluate the effect of drugs on improving pain. As administration of rescue analgesics is a time-to-event outcome, it is more appropriate to compare this outcome using a hazard ratio. For example, in patients undergoing procedures with severe pain, even effective intervention may not reduce the number of patients who require rescue analgesics, but it may prolong the interval before the rescue becomes necessary (ie, analgesic duration). Consequently, the results of the analgesic duration and that of the incidence of rescue use do not agree. Use of a hazard ratio could solve the conflict, although no trials included in our meta-analysis reported this outcome using a hazard ratio.
There was also considerable heterogeneity among the studies in rescue use. In 2 of the included trials, the number of patients requiring rescue analgesics was significantly reduced.8,28 However, in the trial by Abd-Elshafy et al,9 that reduction was very small and only 1 patient in the intervention group did not require rescue analgesics,9 while the difference of the mean of analgesic time was much greater than that of other trials. In their study, the number of rescue analgesic doses and total amount of acetaminophen administered were significantly larger in the control group than in the intervention group. The reason for this difference may be because the authors recruited patients undergoing lower limb orthopedic surgery, which may be more painful and require more intense postoperative analgesia than inguinal or urologic procedures. The procedure might have been so painful that almost all patients in both groups required rescue analgesia by 24 hours, while in other trials, pain was not as severe and not all patients in the control group were administered rescue analgesia. Thus, the number of patients who required rescue analgesics by 24 hours postoperatively may not be a good indicator of analgesic effect in those patients. When patients were expected to have sustained pain postoperatively, adding an intravenous dexamethasone to caudal analgesia may be insufficient for pain control.
The difference in the pain score was observed at different time points among the trials. We could include only 2 studies in a meta-analysis of the difference of the pain score, and the results showed that there was no statistical difference between the dexamethasone and control groups due to a very wide 95% CI. Hong et al8 reported that dexamethasone significantly improved the pain score, whereas Modi et al29 did not demonstrate better pain control in the intervention group soon after the surgery, possibly due to the higher concentration of ropivacaine used for caudal analgesia.
The doses of dexamethasone given in these studies were at least 0.5 mg/kg, which is much larger than that recommended for prevention of nausea and vomiting. In adults, a meta-analysis has revealed that intravenous dexamethasone was effective in reducing postoperative pain at doses >0.1 mg/kg,32 which is similar to the dose administered for PONV prevention,15 and that the effect of dexamethasone on postoperative opioid consumption was not dose dependent.33 It is possible that pediatric patients also do not require a high dose of dexamethasone to increase the analgesic duration, but we could not draw any conclusions due to the lack of evidence. Although a randomized controlled trial investigating the effects of low-dose dexamethasone on the analgesic duration of caudal analgesia is registered at ClinicalTrial.gov (registration number: NCT02436265), we could not include the trial because the results have not been reported.
When administering dexamethasones to surgical patients, the risk of wound infection or delayed healing is a concern. Although there was no significant difference in the incidence of the adverse events between the groups, all but 1 trial we included followed the patients for no more than a few days and did not attempt to detect long-term adverse events such as infection. Abd-Elshafy et al9 reported that blood sugar was not significantly elevated in the dexamethasone group 4 hours after the caudal block, but another researcher reported that an even smaller dose of dexamethasone (0.15 mg/kg) given to prevent PONV elevated blood glucose significantly within the normoglycemic range.39 The incidence of surgical site infection after pediatric abdominal surgery was reported to be between 0.99% and 5.1%.40–43 Given the low incidence of surgical site infection, we may need to include more patients to detect the difference.
There are several limitations in our meta-analysis. The quality of evidence of the effect of intravenous dexamethasone on analgesic duration is very low. The grade was downgraded due to limitations in study design, heterogeneity, and publication bias. All studies except 1 had an unclear or high risk of bias. I2 was 94.8%, which indicates considerable heterogeneity. We included only 6 randomized controlled trials, so a funnel plot was not drawn. Among the 6 trials, only 1 was preregistered and the risk of publication bias was high. The majority of the included trials had an unclear risk of allocation concealment and blinding of health care providers. Future studies need to take steps to prevent this bias. Another limitation was that no study has reported the effect of dexamethasone on prolonging the duration of the sensory block. Although our results showed that intravenous dexamethasone prolonged the analgesic duration of the caudal block, whether the effect of the dexamethasone was due to its intrinsic analgesic effect or if it truly prolonged the sensory block was not clarified. Further studies are required to answer that question. The fact that the starting points of analgesic duration varied among the trials was also a limitation. As we combined the MD of the duration between the 2 groups, the effect of starting point variation on our analysis should be minimal.
In conclusion, when 0.5 mg/kg of intravenous dexamethasone was administered to pediatric patients receiving caudal anesthesia, the analgesic duration was increased but the number of patients who needed rescue analgesics may not have decreased. The quality of the evidence was very low, and the difference in analgesic duration may have been changed by additional randomized control trials with low risk of bias. Additional randomized trials are necessary to investigate the effectiveness of a lower dose of dexamethasone.
We would like to thank Editage (www.editage.jp) for English language editing.
Name: Hiromasa Kawakami, MD.
Contribution: This author helped design the study, collect the literature, extract and analyze the data, and prepare the manuscript.
Name: Takahiro Mihara, MD, PhD.
Contribution: This author helped design and conduct the study, analyze the data, and prepare the manuscript.
Name: Nobuhito Nakamura, MD.
Contribution: This author helped collect and analyze the data and prepare the manuscript.
Name: Koui Ka, MD.
Contribution: This author helped prepare the manuscript.
Name: Takahisa Goto, MD, PhD.
Contribution: This author helped prepare the manuscript.
This manuscript was handled by: Richard Brull, MD, FRCPC.
1. Dalens B, Hasnaoui ACaudal anesthesia in pediatric surgery: success rate and adverse effects in 750 consecutive patients. Anesth Analg. 1989;68:83–89.
2. Schnabel A, Poepping DM, Pogatzki-Zahn EM, Zahn PKEfficacy and safety of clonidine as additive for caudal regional anesthesia: a quantitative systematic review of randomized controlled trials. Paediatr Anaesth. 2011;21:1219–1230.
3. Tong Y, Ren H, Ding X, Jin S, Chen Z, Li QAnalgesic effect and adverse events of dexmedetomidine as additive for pediatric caudal anesthesia: a meta-analysis. Paediatr Anaesth. 2014;24:1224–1230.
4. Schnabel A, Poepping DM, Kranke P, Zahn PK, Pogatzki-Zahn EMEfficacy and adverse effects of ketamine as an additive for paediatric caudal anaesthesia: a quantitative systematic review of randomized controlled trials. Br J Anaesth. 2011;107:601–611.
5. Engelman E, Marsala CBayesian enhanced meta-analysis of post-operative analgesic efficacy of additives for caudal analgesia in children. Acta Anaesthesiol Scand. 2012;56:817–832.
6. Kim EM, Lee JR, Koo BN, Im YJ, Oh HJ, Lee JHAnalgesic efficacy of caudal dexamethasone combined with ropivacaine in children undergoing orchiopexy. Br J Anaesth. 2014;112:885–891.
7. Yousef GT, Ibrahim TH, Khder A, Ibrahim MEnhancement of ropivacaine caudal analgesia using dexamethasone or magnesium in children undergoing inguinal hernia repair. Anesth Essays Res. 2014;8:13–19.
8. Hong JY, Han SW, Kim WO, Kim EJ, Kil HKEffect of dexamethasone in combination with caudal analgesia on postoperative pain control in day-case paediatric orchiopexy. Br J Anaesth. 2010;105:506–510.
9. Abd-Elshafy SK, Yacoup AM, Abdalla EEM, El-Melegy TTH, Abd-Elsalam KAA new look on adding dexamethasone as an adjuvant to caudal bupivacaine; efficacy on postoperative pain and vomiting in pediatric patients. Pain Physician. 2016;19:E841–E852.
10. Lyders EM, Morris PPA case of spinal cord infarction following lumbar transforaminal epidural steroid injection: MR imaging and angiographic findings. AJNR Am J Neuroradiol. 2009;30:1691–1693.
11. Somanchi BV, Mohammad S, Ross RAn unusual complication following caudal epidural steroid injection: a case report. Acta Orthop Belg. 2008;74:720–722.
12. Lundblad M, Lönnqvist PAAdjunct analgesic drugs to local anaesthetics for neuroaxial blocks in children. Curr Opin Anaesthesiol. 2016;29:626–631.
13. Watkins TW, Dupre S, Coucher JRRopivacaine and dexamethasone: a potentially dangerous combination for therapeutic pain injections. J Med Imaging Radiat Oncol. 2015;59:571–577.
14. Jæger P, Grevstad U, Koscielniak-Nielsen ZJ, Sauter AR, Sørensen JK, Dahl JBDoes dexamethasone have a perineural mechanism of action? A paired, blinded, randomized, controlled study in healthy volunteers. Br J Anaesth. 2016;117:635–641.
15. De Oliveira GS Jr, Castro-Alves LJ, Ahmad S, Kendall MC, McCarthy RJDexamethasone to prevent postoperative nausea and vomiting: an updated meta-analysis of randomized controlled trials. Anesth Analg. 2013;116:58–74.
16. Henzi I, Walder B, Tramèr MRDexamethasone for the prevention of postoperative nausea and vomiting: a quantitative systematic review. Anesth Analg. 2000;90:186–194.
17. Moher D, Liberati A, Tetzlaff J, Altman DGPRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–269, W64.
18. Higgins JPT, Green SCochrane Handbook for Systematic Reviews of Interventions. 2008.Chichester, United Kingdom: John Wiley & Sons.
19. Higgins JPT, Altman DGHiggins JPT, Green SChapter 8: Assessing risk of bias in included studies. In: Cochrane Handbook for Systematic Reviews of Interventions. 2008Chichester, United Kingdom: John Wiley & Sons187–241.
20. McMaster University. GRADEpro Guideline Development Tool [Software]. Available at: www.gradepro.org
. Published 2015. Accessed February 26, 2017.
21. DerSimonian R, Laird NMeta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188.
22. Egger M, Davey Smith G, Schneider M, Minder CBias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634.
23. Wetterslev J, Thorlund K, Brok J, Gluud CTrial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis. J Clin Epidemiol. 2008;61:64–75.
24. DeMets DL, Lan KKInterim analysis: the alpha spending function approach. Stat Med. 1994;13:1341–1352.
25. Thorlund K, Engstrom J, Wetterslev J, Brok J, Imberger G, Gluud CUser Manual for Trial Sequential Analysis (TSA). Copenhagen, DenmarkCopenhagen Trial Unit, Center for Clinical Intervention Research. Available at: www.ctu.dk/tsa
. Published 2011. Accessed January 12, 2017.
26. Bangash LR, Afzal F, And HS, Ali KComparison of caudal block alone with caudal block plus intravenous dexamethasone for postoperative analgesia in children undergoing orchidopexy under general anaesthesia. Biomedica. 2014;30:267–271.
27. Murni Sari Ahmad A, Azarinah I, Esa K, Khairulamir Z, Hamidah I, Norsidah Abdul MIntravenous dexamethasone in combination with caudal block prolongs postoperative analgesia in pediatric daycare surgery. Middle East J Anaesthesiol. 2015;23:177–183.
28. Khan AS, Waqur-ul-NisaEffect of intravenous dexamethasone in combination with caudal analgesia on postoperative pain control in paediatric orchidopexy. Ski J Med Sci. 2015;18:20.
29. Modi TM, Jain P, Zinzala B, Raval BEvaluation of effect of combination of intravenous dexamethasone and caudal ropivacaine on duration of post operative analgesia in paediatric patients. Int J Adv Case Reports. 2016;3:438–444.
30. Hargreaves KM, Costello AGlucocorticoids suppress levels of immunoreactive bradykinin in inflamed tissue as evaluated by microdialysis probes. Clin Pharmacol Ther. 1990;48:168–178.
31. Hong D, Byers MR, Oswald RJDexamethasone treatment reduces sensory neuropeptides and nerve sprouting reactions in injured teeth. Pain. 1993;55:171–181.
32. De Oliveira GS Jr, Almeida MD, Benzon HT, McCarthy RJPerioperative single dose systemic dexamethasone for postoperative pain: a meta-analysis of randomized controlled trials. Anesthesiology. 2011;115:575–588.
33. Waldron NH, Jones CA, Gan TJ, Allen TK, Habib ASImpact of perioperative dexamethasone on postoperative analgesia and side-effects: systematic review and meta-analysis. Br J Anaesth. 2013;110:191–200.
34. Steward DL, Grisel J, Meinzen-Derr JSteroids for improving recovery following tonsillectomy in children. Cochrane Database Syst Rev. 2011:CD003997.
35. Desmet M, Vanneste B, Reynvoet M, et al.A randomised controlled trial of intravenous dexamethasone combined with interscalene brachial plexus blockade for shoulder surgery. Anaesthesia. 2015;70:1180–1185.
36. Abdallah FW, Johnson J, Chan V, et al.Intravenous dexamethasone and perineural dexamethasone similarly prolong the duration of analgesia after supraclavicular brachial plexus block: a randomized, triple-arm, double-blind, placebo-controlled trial. Reg Anesth Pain Med. 2015;40:125–132.
37. Shirazi M, Mahmoudi H, Nasihatkon B, Ghaffaripour S, Eslahi AEfficacy of dexamethasone on postoperative analgesia in children undergoing hypospadias repair. Pak J Med Sci. 2016;32:125–129.
38. Smith AF, Carlisle JReviews, systematic reviews and anaesthesia. Anaesthesia. 2015;70:644–650.
39. Gnatzy R, Hempel G, Kaisers UX, Höhne CThe effect of intraoperative administration of dexamethasone for PONV prophylaxis on perioperative blood glucose level in obese and normal weight children. J Pediatr Endocrinol Metab. 2015;28:1287–1292.
40. Olsen MA, Nickel KB, Wallace AE, Mines D, Fraser VJ, Warren DKStratification of surgical site infection by operative factors and comparison of infection rates after hernia repair. Infect Control Hosp Epidemiol. 2015;36:329–335.
41. Khoshbin A, So JP, Aleem IS, Stephens D, Matlow AG, Wright JGSickKids Surgical Site Infection Task Force. Antibiotic prophylaxis to prevent surgical site infections in children: a prospective cohort study. Ann Surg. 2015;262:397–402.
42. Varik K, Kirsimägi U, Värimäe EA, Eller M, Lõivukene R, Kübarsepp VIncidence and risk factors of surgical wound infection in children: a prospective study. Scand J Surg. 2010;99:162–166.
43. Bucher BT, Guth RM, Elward AM, et al.Risk factors and outcomes of surgical site infection in children. J Am Coll Surg. 2011;212:1033–1038.e1.