A Systematic Review and Meta-Analysis of Caudal Block as Compared to Noncaudal Regional Techniques for Inguinal Surgeries in Children

This systematic review and meta-analysis were designed to compare the analgesic effectiveness and adverse effects with the use of caudal analgesia as compared to noncaudal regional analgesia techniques in children undergoing inguinal surgeries. MEDLINE, EMBASE, and CENTRAL (Cochrane) databases were searched for randomized control trials published in English language from 1946 up to 2013. Use of rescue analgesia and adverse effects were considered as primary and secondary outcomes, respectively. Outcomes were pooled using random effects model and reported as risk ratio (RR) with 95% CI. Out of 3240 hits and 24 reports for final selection, 17 were included in this review. Caudal analgesia was found to be better in both early (RR = 0.81 [0.66, 0.99], P = 0.04) and late (RR = 0.81 [0.69, 0.96], P = 0.01) periods, but with a significant risk of motor block and urinary retention. According to GRADE, the quality of evidence was moderate. Although potentially superior, caudal analgesia increases the chance of motor block and urinary retention. There are limited studies to demonstrate that the technical superiority using ultrasound translates into better clinical success with the inguinal nerve blocks.


inguinal hern
a repair with or without orchidopexy (orchiopexy) [1].For postoperative pain with these surgeries, a regional analgesic modality such as caudal analgesia (CA), inguinal and iliohypogastric nerve block (INB), or local infiltration (INF) is combined with a general anaesthetic (GA).When compared to intravenous (IV) opioids, regional techniques reduce the risk of side effects such as somnolence, respiratory depression, emesis, and ileus [2].Caudal block (CB) involves the introduction of local anaesthetic (LA) into the caudal epidural space.It requires the child to be positioned appropriately and is a common practice to administer under deep sedation or a GA.It can cause complications such as needle trauma, infection, haematoma, and inadvertent subarachnoid or intravascular injection of the LA [3].Other associated adverse effects can include urinary retention and possible motor blockade.INB-including inguinal and iliohypogastric nerve blocks can provide effective ipsilateral analgesia.A single injection often blocks both the nerves, as they lie quite close to each other, side by side.Landmark techniques can result in technical failure in up to 20% of children.This can be improved by using ultrasound guidance [4].They possibly do not affect the pain caused by visceral manipulation.INF of the wound can be done by the anaesthesiologist or the surgeon.This potentially effective, but minimally invasive procedure could offer the advantage of lower costs, time, and risks [5].Other interventions which have been compared to caudal analgesia include paravertebral block (PVB) and TAP (transverse abdominal plane) block.Although CB may be considered 2 BioMed Research International as the most potent technique, it requires trained personnel and added cost; potentially it could expose children to higher risks without any superiority in clinical analgesia.Ultrasound (US) has been shown to improve the technical efficiency and success rate of regional blocks [4].It is not established whether this translates into comparative clinical success.In clinical practice, there is still no conclusive evidence to prefer one technique over the other in consideration of their efficacy and side effects.The main objective of this review is to perform a systematic review and meta-analysis of the existing evidence to compare the analgesic efficacy and side effects with the use of caudal analgesia as compared to other noncaudal regional analgesia techniques in children undergoing inguinal surgeries.


Methodology

We performed a comprehensive search in MEDLINE, EM-BA

, and Cochra
e databases for randomized controlled trials (RCT) in English language (Appendix A).This was complemented by other databases, relevant conference proceedings, and hand check of reference lists of reviews and included RCT.Our selection criteria using the PICOT format are as follows: participants: children (0-12 years) undergoing inguinal surgeries; intervention: CB (without adjuvants); comparators: noncaudal regional techniques; outcome: effectiveness of analgesia assessed using a categorical outcome reporting as the number of children needing rescue analgesia; time point: early (<4 hrs) and late (4-24 hrs).Trials with various comparator techniques such as INB, INF, or a combination of these interventions were included.Trials using adjuvants other than epinephrine (such as ketamine, clonidine) or CB in combination with another technique were excluded.Studies with outcome reporting using "only pain scales" (continuous) were separately reported without combining them in the possible pooled estimate.As secondary outcomes, clinically relevant adverse effects as related to treatments were considered: motor block (MB), urinary retention (UR), nausea-vomiting (NV), infection at the injection site, and delayed discharge (DD).The study selection was done independently by the first two authors, and a final agreement score was calculated using a quadratic kappa weighting.Risk of bias was assessed using the Cochrane risk of bias tool.Considering the nature of interventions, blinding of physicians may not be possible.Hence, only participant blinding was considered necessary to be identified as having a low risk of bias.We considered a loss to follow-up (LTFU) of 10% or more as the threshold for attrition bias.


Summary Measures and Synthesis of Results.

For the primary outcome, the proportions of children needing rescue analgesia were compared.Rescue analgesia was considered as the administration of an analgesic medication to control pain with or without agitation.Outcomes at the 2 time points, early (<4 hrs) and late (4 hrs-24 hrs), were analysed separately as a pooled estimate for all the studies and also individually within each group.For secondary outcomes, adverse effects in

ach category were compared as proportion of
patients.For both, outcomes were pooled and reported as relative risks with 95% CI.

Synthesis was done using revman (review manager 5.2).A priori, it was decided that only if the studies are sufficiently homogeneous, outcomes would be reported as pooled effect sizes.Studies reporting only continuous outcome measures, and also studies which were substantially heterogeneous and did not fit into a particular comparison group, were analysed separately and reported, without inclusion for the pooled effect estimate.To accommodate for any unexplained heterogeneity, ra dom effects model was used for analysis.Statistical heterogeneity was calculated using Chi Square and also the  2 statistic to describe the percentage variability in individual effect estimates that could be due to true differences between the studies rather than a sampling error.We considered  2 < 40% as low, 30-60% as moderate, and >50% as substantial [6].Further, study findings have also been shown in the form of "summary of findings" table, using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) approach.The utility of "an estimate of the magnitude of intervention effect" depends upon our confidence in that estimate.GRADE incorporates the aspects of study limitations, inconsistency of results, indirectness of the evidence, imprecision, and the reporting bias [7].1).The search results are highlighted as a flow diagram in Figure 1.Out of 3240 items, 1958 reports were obtained after removing duplicates.Finally, 27 full-text articles were considered for inclusion out of which 17 were included (Table 3) [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24], and three were excluded [25][26][27].Of the remaining seven studies [28][29][30][31][32][33][34], two were journal reports and five were conference proceedings.Despite multiple attempts, we could not obtain any full study report for the above seven studies.The study selection agreement between the authors was 0.73 using quadratic kappa weighting.For quantitative analysis (metaanalysis), only 16/17 studies were included; Hannallah and colleagues reported their results only as continuous outcomes [12].


Results


Study Selection (PRISMA Flow Chart-Figure


Characteristics of Included Studies.

Other important considerations are as follows.The study by Fisher and colleagues [9] included a three-arm design with caudal compared to inguinal nerve block.It involved the use of "epinephrine with LA" in only one group of CB.For the purpose of this review, both caudal groups were combined for comparisons, as suggested by [35].The study by Tug and      category and was hence reported separately [24].Jahromi

groups [17]: infiltration and acetam
nophen suppository.We included only the patients compared under infiltration with the caudal group.There was only one included study using US image guidance for INB [8].


Risk of Bias within

Studies.The risk of bias across studies is represented in the bar graph obtained through revman (Figure 2).The risk of bias in individual studies, in specific domains, is shown in Figure 3.A majority of studies were observed as having a high risk o selection bias.Only seven studies reported the method used for sequence generation,  and only five studies reported the method used for allocation concealment.Four studies excluded patients with failed interventions from the final analysis [8,9,17,24].For primary outcome analysis, we imputed the outcome of these excluded patients.Our rationale was that failed interventions will always n

essitate rescue anal
esic.Conroy and colleagues reported that they did not follow the randomisation sequence appropriately as generated for the first 30 patients [16].Hence, we decided that a sensitivity analysis is to be carried out by excluding this study and observing the change in estimate of effect.The reporting of methodology and outcome assessment was not entirely clear in the study by Lafferty et al. [14].We could not identify any major publication bias (funnel plot-Figure 4) and no study mentioned any specific funding support.


Outcome Analysis and Results


Use of Rescue Analgesia in Early

Period: ≤4 hrs (Figure 5).In total there were 14 studies with 851 patients.Tug and colleagues compared CB with single shot lumbar (L2) PVB [24].We noted that the inclusion of this particular study resulted in heterogeneity and significant subgroup differences (test for subgroup differences: Chi = 7.66, df = 3 ( = 0.05),  2 = 60.8%).Also, in practice it is not commonly performed for inguinal surgeries in children.After its exclusion, we had 13 studies with 789 children with the overall pooled estimate favouring caudal; RR: 0.81 [0.66, 0.99],  = 0.04, with no identifiable heterogeneity ( 2 = 0) or subgr

risk reduction) was 1.38.Quality of evidence, according to
he GRADE, is moderate (Table 2).


Use of Rescue Analgesia in Late

Period: 4-24 hrs (Figure 6).In total there were 9 studies with 597 patients.Excluding one study [24], for reasons of heterogeneity, resulted in 8 studies with 532 children.Overall pooled estimate favours the benefit of analgesia from caudal; RR: 0.81 [0.69, 0.96],  = 0.01, reaching statistical significance.The ARR was 7.8.Quality of evidence, according to the GRADE-SOF, is moderate (Table 2).


Side Effects.

Motor blockade (Figure 7) was observed in 24/239 children in the CB group compared to 6/230 children in the comparator group: 6 studies with 469 children; RR = 2.59 [1.29, 5.20],  = 0.007.Urinary retention (Figure 8) was observed in 32/219 children in the CB group compared to 13/210 children in the comparator group: 5 studies with 459 children: RR = 2.23 [1.27, 3.91],  = 0.005.NV was observed to be similar in both caudal and noncaudal groups.

Only Lafferty and others reported infection in one child belonging to the INF [14]; and only Fell and others reported delayed discharge in three and one, respectively, in CB and INF [15].Although there were clearly more side effects with CB, the quality of evidence, according to GRADE, was very low, except for NV.Reasons for downgrading the evidence is provided within the SOF table (Table 2).


Description of Results within Individual Subgroups (Table 1) 3.5.1. CB versus INB.

We identified five studies, out of which four were included in the meta-analysis.Except Hannallah and colleagues (orchidopexy only) [12], the studies included patients fr

injected range
from 0.7 to 1 mL kg −1 (CB) and from 0.1 mL to 0.4 mL kg −1 (INB).Only Fisher and colleagues [9] used epinephrine mixed with bupivacaine in one arm of their caudal patients.We combined them together as belonging to CB.All except one performed both their interventions before surgery [9].Hannallah and colleagues reported their pain scores only in "median range, " caudal (1.0, 6) and N block (1.0, 6), and did not report the use of rescue analgesic in the two groups separately [12].


CB versus INF.

We identified 6 studies in total, but only two studies [9,[12][13][14][15]17] provided analgesia outcomes for both time periods.Both interventions were performed after surgery in two studies [13,14,17]; however the other 4 studies performed caudal preoperatively and infiltration postoperatively [1

[15][16]18].Exc
pt for Lafferty and colleagues (only orchidopexy) [14], all included hernia surgeries only.All used bupivacaine in a concentration of 0.25% for CB and 0.25%-0.5% for INF.The volume ranged from 0.7 to 1.0 mL kg −1 (CB) and from 0.2 to 0.7 mL kg −1 (INF).Only Conroy and colleagues used epinephrine along with bupivacaine [16].Variations of the infiltration techniques involved infiltration of the wound site through the skin and infiltration of fascia or aponeurosis before closure.No study used image guidance.


CB versus Combined INB and INF.

Five studies were identified.Tobias and colleagues also performed a laparoscopic inspection of the other side [19].Bhattarai and colleagues [23], all placed their CB before surgery.However, the timing of INB and INF was variable.Epinephrine was used in three of the studies along with bup

acaine.Compared to other groups,
bupivacaine concentration used was 0.25% in all studies except 0.2% for CB by Splinter and colleagues [20].The volume ranged from 1 to 1.25 mL kg −1 (CB) and from 0.3 to 1 mL kg Additional Analysis (Sensitivity Analysis).Although we considered concentration of LA, timing of interventions as related to surgery, and the use of image guidance as potential sources of heterogeneity, we did not have sufficient number of studies to carry out further subgroup analysis.

(1) Conroy an colleagues had noted that the randomisation was not done appropriately for the first 30 patients [16].Sensitivity analysis showed that the pooled effect size for the overall estimate and the subgroup (CB versus INF) estimate was not affected much for the early use of rescue analgesia.

(


Discussion 4.1.Summary of Evidence.Our results show that CB is superior compared to the group of noncaudal regional analgesic int

lving INF, INB, or th
ir combination, demonstrated by the significantly reduced need for rescue analgesic during both early and late periods.However, the ARR (absolute risk reduction) was only 1.58 for the early period, compared to 7.94 for the late period, indicating that the benefits are perhaps more appreciable in the later period.Among the side effects, motor block and urinary retention were significantly more common with the caudal group with an ARR of 7.44 and 8.42, respectively.NV was found to be similar.Individually, among the subgroups, the need for rescue analgesia was less with CB compared to INB, and the combined INB with INF.However, the reduction did not achieve statistical significance.For provision of postoperative pain relief in inguinal surgeries in children, regional procedures are preferred because of several advantages over parenteral analgesics [36].Caudal analgesia has been widely used, and, because of the ease of administration, it is the most commonly used neuraxial block for children [37].In children, most regional procedures are done under GA or heavy sedation [38].The relative risks and benefits of CB as compared to less demanding techniques such as INB and infiltration are unclear.Complications could arise as a result of LA used or because of the nature of the regional technique [39].The potential for harm is perhaps more with a neuraxial block.Our review shows that, despite the common practice of CB, there are limited studies.The exclusion of reports which used adjuvants did not seem to affect the study results.All of those excluded study reports, except a single study [40], had studied the effect of adjuvants when used in the caudal space without actually comparing it with other comparator regional techniques.The excluded study was a pilot study by Ivani and colleagues who studied the use of ropivacaine mixed with clonidine and compared between CB and INB, with children aged 1-7 years undergoing inguinal surgeries.The pain scores were similar, with 6/20 and 11/20 children needing rescue analgesia in INB and CB, respectively [40].A systematic review also did not find convincing evidence for the use of nonopioid additives in elective outpatient surgery involving children [41].

In our review, most studies suffered from the risk of selection bias or did not specify the method of sequence


Total events

Heterogeneity:  2 = .00;  2 = 2.24, df = 5 (P = 0.81); I 2 = 0% Test for overall effect: Z = 2.69 (P = 0.007) Test for subgroup differences:  2 = 1.37, df = 2 (P = 0.50); I 2 = 0%  [5].The risk of technical failure exists with both techniques.INB can also suffer from a success rate of only 70%-80% [43].The use of US could potentially improve the precision of both CB and INB.It has been shown that the success rate of caudal injection [44,45], as well as INB [4], could be better using US guidance.Despite this, we only found 2 studies comparing US-INB to CB [8,29].Although not directly applicable to our results, we explored for other studies on the possible use of US-INB in children.Apart from Willschke and colleagues [4], we only found 4 others.Two of them looked at the exact site of injection and plasma levels of ropivacaine, respectively [46,47].Another study looked at the addition of US guided INB with CB. [48]; pain scores were found to be significantly different; however, the amount of rescue analgesic used was not.Ghani and colleagues compared US-INB with US-TAP block and found that US-INB was superior [49].Although it is acceptable to appreciate the superior technical efficiency of INB using US guidance, given the limited evidence, it cannot be extrapolated to infer a superior clinical effectiveness as compared to CB for inguinal surgeries in children.One must also keep in mind that the plasma levels of LA were found to be significantly higher with US guided blocks than landmarkbased [47]; it has significant implications on doing a rescue block or any additional local infiltration.


Limitations.

There were fewer studies, mostly with smaller sample sizes.Most studies were rated high for selection bias.There were no uniform, reliable, and validated outcome measures and the thresholds used for providing rescue analgesia were variable.It can also be argued that a network meta-analysis or multiple treatment comparison would have been a better approa

.However, the
e are limitations to interpretation or inferences drawn from such an analysis as they could be prone to a higher degree of heterogeneity and invalid conclusions [50].


Conclusions

Caudal block provides superior analgesia requiring less rescue analgesic, with higher chances of motor block and urinary retention.There seems to be little advantage of combining both INB and INF as compared to CB; by requiring more volume, this may even cause harm by potentially increasing the chances of LA toxicity.As shown in the attached SOF table, the e

dence level
or the analgesic requirement is moderate and future studies looking to evaluate this comparison will have an important impact on the confidence of this estimate.More comparative studies are required to demonstrate that better technical efficiency, with the use of US-INB, translates into superior clinical effectiveness, as compared to CB alone.Apart from larger sample sizes, studies should use well defined criteria for measurement of these outcomes.The basis for the assumed risk (e.g., the median control group risk across studies) is provided in footnotes.The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).CI: confidence interval; RR: risk ratio.GRADE Working Group grades of evidence:

high quality: further research is very unlikely to change our confidence in the estimate of effect; moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to chang

e; very low quality
we are very uncertain about the estimate.

1

There was no appropriate concealment in the majority of the studies.

2

Based on available studies, funnel plot looks symmetrical.

3

None of the studies were industry funded.

4

No uniform criter a were considered for assessment of motor blockade.

5

Wide confidence interval.

6

Sample size too low to detect a true difference.

7

No uniform criteria used for assessment of urinary retention.

8 Several confounders were not controlled appropriately.

9

No appropriate concealment or random sequence generation.



Funnel plot was used to look for any publication bias.Considering the clinically heterogeneous comparator techniques, we decided a priori to subgroup them into CB versus INB, CB versus INF, CB versus combined INB and INF, and CB versus others.Other potential sources of heterogeneity considered were concentration of LA, timing of interventions as related to surgery, and the use of image guidance.


Figure 2 :
2
Figure 2: Ri