• 
• 

Three Part Question

In[paediatric patients aged one to four years] does the [use of oral sucrose prior to painful interventions][reduce procedural pain]

Clinical Scenario

A three year old child is brought to the emergency department with pyrexia and dehydration, requiring cannulation. Routinely topical anaesthetic Ametop would be applied prior to this procedure. However a paediatrician has come to see the patient and requested giving the child sucrose. When questioned, the paediatric doctor explains it provides comfort and settles children during procedures such as cannulation. Despite years of Accident and Emergency experience using this technique for comfort in neonates you wonder whether this will be of similar effective use in older children.

Search Strategy

Medline 1946- January week 3 2016
*child, preschool/ OR pediatric ti.ab.af/ OR pediat*ti.ab/ OR*paediatrics OR *infant, new-born
AND
*sucrose OR (sucrose Adj4 solution) ti, ab
AND
*pain/ OR *analgesia/OR *pain management/ OR *therapeutics/
Limits Human and Language English

Embase 1946-January week 3 2016
*paediatrics/ OR * infant
AND
Pain/ OR *pain management/ OR *pain relief
AND
*sucrose OR (sucrose and Solution)
Limits Human and language English

The Cochrane Library
MeSH descriptor [sucrose] explode all trees
Web Search www.controlled-trials.com/search

Search Outcome

Search Results
Medline 36 results 1 relevant all others excluded as all studies of neonates pre-term infants or infants less than one year old.
Embase 5 results none relevant
Web search. Basic search – Sucrose 9 completed trials -8 not considered excluded by age- pre-term, new-born or infant. 1 of potential interest but no publication details given.Cochrane Reviews 5 Excluded 4 – 2 specific to treatments, 2 specific to under one year old children.
Other reviews 16 none relevant excluded by age all studies for neonates, pre-term infants or new-born
Trials 1123 (sorted by relevance) one potentially relevant but excluded as specific to children aged 8-12yrs and non-needle related procedure.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Harrison,D et al 2011 issue 10 Canada	330 Children aged One to Sixteen years from four RCTs undergoing needle related procedures. Including venepuncture, heel lance, finger lance, subcutaneous or intra-muscular injection, lumbar puncture or supra-pubic bladder aspiration. Given orally administered sweet tasting substance. Excluded studied with ages less than inclusion criteria and two studies not needle related interventions.	Level 1 Systematic review of RCTs or Quasi-RCTs Four planned prospective blinded RCTs with control group and one or more treatment intervention reviewed. Meta-analysis performed for both age groups. Risk of bias assessed for each study using Risk of Bias Scale(Higgins 2011)	In young children there is no definitive outcome to support or reject the use of sucrose in this age group. For school aged children the available evidence suggests that the use of sweet tasting solution is not of sufficient benefit to advocate its use.	The two studies for toddlers and young children have conflicting results. Therefore unable to recommend the use of 12% sucrose as a pain management strategy beyond the age of twelve months. Significant heterogeneity (I2 – 93%) reported. Two studies for older children comparing analgesic effect of sweetened gum had no effect either when sweet gum chewing stopped prior to or continued during the needle related procedure. Reported zero heterogeneity for the four meta-analyses for this age group.	Low number of studies included in the review therefore is there enough data to reduce both bias and random error. The review acknowledges sensitivity analysis was not carried out due to the small number of studies. The methodological quality was tabulated using the risk of bias scale which shown that the four studies had an unclear risk of selection bias as they had not described the sequence generation. Additionally the four studies had unclear risks of reporting bias as they had not registered their original protocols therefore they were not assessed. Furthermore there was incomplete outcome data from one study even after collecting original data from the author n59 but the published report included n67. It is acknowledged that extracting small numbers of age appropriate participants from the original data could be a source of bias. This also affects the external validity of the review; due to the small samples from each study are the results generalizable to the age group.
Harrison,D et al 2015 issue 5 Canada	808 Children aged One to Sixteen years from one unpublished and seven published RCTs who undergo needle related procedures. Including venepuncture heel lance, finger lance, subcutaneous or intra-muscular injection, lumbar puncture or supra-pubic bladder aspiration. Given orally administered sweet tasting substance. Excluded studied with ages less than inclusion criteria.	Level 1 Systematic review of 8 blinded RCTs. One unpublished (including two pilot RCTs) and seven published. This review was an update of the original Cochrane review (Harrison 2011) to ascertain if any subsequent research or publication would transpose the conclusions of the original review. All studies included were prospective blinded RCTs with a control and one or more treatment intervention. Meta-analyses of the data were carried out Heterogeneity of studies including the I2 statistic and sub-group analysis determined by age was carried out. Only one of the RCTs had numbers of participants greater than fifty.	For children aged one to four there is at present insufficient evidence to recommend sucrose use as a strategy for pain management. For school aged children the author’s findings state that sucrose as a strategy should not be considered as there is no evidence to support its effectiveness in this population. And recommend that it is not a research priority for the future.	For children aged one to four years the review found conflicting results. In two studies (Yilmaz2014) (Dilli2009) the use of sucrose significantly reduced pain scores and duration of crying. The study by Yilmaz (2014) also concluded that sucrose at a higher concentration was significantly more effective compared to a lower concentration, although also concluded that the lower concentration solution was more effective than the placebo (water). However the other studies of children in this age group had no significant reduction in crying time with sucrose use. The two studies of older children that considered the analgesic effects of sucrose use by using sweet tasting chewing gum compared to unsweetened chewing gum use prior to or during the procedure had no significant effect on reported pain.	Low number of studies included in the review. Of the eight studies four were reviewed in the previous Cochrane review (2011). All studies or sub-sets reviewed had sample sizes less than fifty with only one study having a sample size greater than this. However the data analysis only included the data from 25% sucrose group with control water to compare to 12% sucrose with no treatment in another study (Dilli 2009) the justification given for this was to improve heterogeneity but acknowledges that this still remained very high. And did not include the comparison as a whole group so that judgement could be made by the reader if there was difference in the effects on outcome. The methodological quality was again tabulated using the risk of bias scale. Only two of the studies sufficiently described random sequence allocations which were therefore classified as unclear risk of bias. And four of the eight studies, including the study with the largest number of participants were deemed at high risk of selection bias as the allocation concealment was unclear. The review also reports a high risk of performance bias for example in the Harrison (2014b) study the use lollipops as the oral sucrose agent would not be able to be blinded to the personnel of the participant, furthermore the older children who were able to verbalise which type of solution they were given may have prevented efforts in blinding the person carrying out the observation. Despite efforts made to blind in these studies particularly when introducing an intervention to older children who may disclose which intervention is used there is no mention in the review if indeed any of these efforts to reduce bias were made or if the person performing the outcome measurements were blinded to the original treatment allocation. Additionally how does classifying bias as unclear risk assist the reader without obtaining the original study themselves? The study by Yilmaz (2014), although the largest sample set included in the review the original planned or actual numbers of participants enrolled for the study were not given therefore unable to determine if all the data collected was included. For the older children is the sample set for the two studies 111 and 103 school aged children respectively for each RCT, even when reviewed together is the sample big enough to ensure that an inaccurate conclusion that the intervention has no effect be made, because the updated review has not included any further studies to this age group from the original review for comparison. The author of the three pilot studies included in the review was part of the review group and her participation included the evaluation of methodological quality, the meta-analysis of the data this could lead to bias whether intentional or not, all collaborating could be influenced by this fact and could be bias to the authors research as the reviewers were not blinded to author or study reviewed. Furthermore the author herself could be unduly critical or provide a more favourable appraisal of her own work
Curtis,S et al 2012 Canada	Older infants, children and adolescents. Ranging from 0 months to 19 years, an overview of 4 Cochrane reviews, and 11 additional RCTs included by the author which were not included in the original Cochrane reviews.	Level 3 An overview to summarise the systematic reviews of the management of paediatric pain using different interventions.	The overview found no further evidence of the interventions reviewed having any further beneficial outcome. Furthermore that non-pharmacological and sucrose use in older children require further procedure specific study.	This review reports finding the use of sweet gum for toddlers or the use of 12% sucrose use for pain relief during needle related pain that there is no current evidence of benefit to support its use. The overview also considered other non-pharmacological methods and report finding potential effectiveness in other interventions such as distraction, combined cognitive-behavioural interventions, hypnosis, non-nutritive sucking and video distraction. It also finds that Amethocaine provides superior pain relief compared to EMLA for IVC procedures.	The purpose of the overview was to review Cochrane reviews to provide a broad scope of interventions that could be used during painful procedures in children outside the neonatal period but then concludes that the use of sucrose and other non-pharmacological interventions require further procedure specific research, therefore for what reasons were the results collated if only to report the findings of the studies and recommend that further research should keep the interventions separate. Furthermore it only presents parts of the original reviews and offers no explanation to how the included parts were selected or excluded thus introducing bias from the outset. The editors’ note describes the purpose to compile evidence from multiple Cochrane reviews into one accessible and usable document that it aims to focus on the treatment question of which treatment to use for a condition however the piece provided unclear answers to that question, it provides descriptive explanations of the interventions and rationale for their use and no appraisal of the reviews just descriptive representations of the findings and tables reproduced from the original review of methodological qualities of the studies reviewed. Additionally, only presents parts of the original reviews results from which the reader could draw its conclusions, by only including sections or sub-groups is this misleading to the reader. Furthermore the number of different interventions included did not make the review accessible, and although the types of outcome measures were described it was difficult to see where the comparisons lie. The search strategy was not transparent the search for systematic reviews was carried out on the Cochrane data base found 6 reviews which met the criteria of analgesia effect of an intervention on children under 19 years, 2 of which were excluded as procedure specific only two search terms were provided . Furthermore the author then reports completing an updated literature search on the Medline database for the four original reviews using their search strategies outlined and states they are given in an appendix however only provided the strategies for two of the four. Then only completed a literature review in the Embase database using the search strategy for one of the reviews demonstrating inconsistences in the search. From this search 13 additional RCTs were added to the overview additionally it was selective what data was extracted from each original review for example 3 trials were excluded from one review as they were single trials and meta-analysis had not been carried out but then uses a single trial from another review to demonstrate no significant difference in a comparison of the effectiveness of water prior to venepuncture and no treatment group. Further demonstrating selective bias. Additionally only 11 of 51 trials were included from one review as only the 11 fulfilled the criteria for older infant’s toddlers and older children. Although 3 trials from another review which included data from neonates were not excluded, this is acknowledged by the author and reports only a minimal data set however does not provide numbers or explain if this could bias results, on the contrary states that it contributes valuable information to the group of interest. The data synthesis was also debatable each of the Cochrane reviews had used different methods of reviewing results for example for dichotomous data risk ratio was given, for continuous data two reported standard mean difference and one mean difference the overview used Review Manager 5.0 to pool the data however when trials did not report means or standard deviation means were imputed with available medians and standard deviations were imputed with IQRs. Did this have significant effect upon the statistical relevance, also it was not reported which of the trials this was carried out for thus not allowing the reader to interpret the results themselves. In one summary of findings reports the removal of data of seven patients due to significant heterogeneity between the two studies however the heterogeneity remained high at I2 93% and I2 90% respectively did this adjustment have any significance? In another review an outlier trial was removed from analysis has this modified the effect on the results overall. The inclusion and exclusion of particular data without adequate explanation or justification raises concerns of the validity of the overall findings.

Comment(s)

The use of sucrose in neonates and pre-term infants is well documented, researched and widely used. The rationale for potentially extending its use beyond the neo-natal period and adopting this practice in the emergency department includes its safe, its ease of use; it is cost effective and readily available. However does it provide analgesic effects for this wide age range? Moreover is it effective enough to replace traditional methods used to provide adequate procedural pain relief? For the children attending the emergency department it is often an unplanned event, which in itself can be a new experience and source of anxiety for the child and the parent/(s) increased by the parents concern of the cause of their child’s illness or outcome of an accident. The prospect of requiring an intervention for investigations for example heel lance for blood sugar monitoring or cannulation for intravenous fluids or medication increases this anxiety not only for the child but also for the parent. Introducing an intervention which will reduce the procedural pain will also reduce the anxiety and improve their overall experience of their visit. So ensuring that the intervention is effective is a priority. Therefore three papers were appraised; two were Cochrane reviews of the use of sweet tasting solutions to reduce pain from needle related procedures. The third was an overview of different interventions used to reduce acute procedural pain in older infants, children and adolescents. The first Cochrane review of RCTs (Harrison et al 2011) included only four studies a total of 330 participants of children aged one to sixteen years comparing the use of sweetened oral solutions or substances with controls including water, non- sweetened, pacifier, distraction, no treatment, position/containment or breast feeding. So heterogeneity was high and the studies underpowered. It also excluded the effects of breast milk as not considered as suitably sweet solution. The second updated Cochrane review (Harrison et al 2015) included seven published and one unpublished RCT, four from the original review total 808 participants with the same inclusion criteria and controls as the original so heterogeneity remained high. Excluded studies from this review were all infants under 12 months and two were excluded as not needle related procedures (cold pressor tests). The administration of the sucrose although limited to the oral route is given by different means between the studies, furthermore it is only administered prior to the procedure, and the concentration of solution used is variable. Additionally the limited number of studies reviewed and low number of participants in each study means meta-analysis and forest plots produced of these results could be of minimal value and excessively bias. The overview of reviews (Curtis et al 2012) collated numerous different interventions utilised for pain management in children undergoing procedures. Although most of the evidence was collated from previous Cochrane reviews it also included a number of RCTs that had been carried out after the Cochrane publications. (None of which were for the use of sucrose). Despite the fact that this attempts to provide the most current and up to date research, the search strategies used to identify these studies, the selection process and the inclusion/exclusion criteria was not transparent and gave the reader little opportunity to exclude potential bias. In its attempts to provide an overview of current evidence, the piece is descriptive and presents numerous findings of different interventions in tables, summaries and analysis in various formats through the text and appendixes making the results not easily accessible to the reader. Furthermore in addressing a broad topic with elements of the original reviews it is unclear if this is a balanced summary of the main findings of each of the original reviews. Regarding the small section of the article which focuses on the use of sucrose it presents summaries of the quality assessments tables in an appendix from the original review however only presents a summary of findings from one RCT from a review and there is no clarity as to which review it is from or who conducted the original RCT. This table of results finds no significance of results in the use of sucrose or control with 95% CL using effective estimate of MD with I2 93% and 90% respectively. However as mentioned previously some of the data had been imputed using median results for some analysis and it is not explained whether this was carried out in this instance. As the information is not provided as to which RCT this was from the reader is unable to ensure the data presented is truly representative of the original findings. Overall from the limited amount of information detailed in this overview of sucrose use, the authors have concluded that sucrose has no significant effect in pain management and even though it acknowledges the need for further research fails to demonstrate the implications of the small number of studies (two by the same Author) and small sample sizes included in the Cochrane review. Without thorough consideration of these points the reader could be dismissive of sucrose use altogether.

Clinical Bottom Line

In agreement with the consensus findings of the reviews, there is not enough quality evidence to support the use of sucrose to provide pain relief beyond the neonatal period at present. That further larger studies with standardized controls are required before efficacy can be determined.

References

1. Harrison,D et al Sweet tasting solutions for reduction of needle-related procedural pain in children aged one to sixteen years Wiley&sons Ltd 2011 issue 10
2. Harrison,D et al Sweet tasting solutions for reduction of needle-related procedural pain in children aged one to sixteen years (Review) Wiley & sons Ltd 2015 issue 5
3. Curtis,S et al The Cochrane Library and procedural pain in children:an overview of reviews Wiley online library 2012 1363-1399