Every day an average of 80 children visit the children’s emergency department (CED) at the Starship Hospital in Auckland, New Zealand. Around 25 will be admitted and 16 will receive an intravenous cannula (IVC).

Intravenous cannulation can be upsetting for the child, the parents/caregivers and occasionally the staff. High levels of distress during such procedures can result in negative consequences ranging from hypoxaemia and haemodynamic instability in newborns to avoidant healthcare attitudes and reduced access of healthcare services in later life.1 2 Several organisations have published guidelines for the management of procedure-related pain.3^–5

The use of topical anaesthetics such as EMLA (an eutectic mixture of 25 mg/g lidocaine and 25 mg/g prilocaine; AstraZeneca, Wilmington, Delaware, USA) has been shown to reduce discomfort of such procedures and is used in our department.6 However, it requires a minimum application time of 60 min and the Starship Pain Team suggests 90 min for optimal anaesthesia for intravenous cannulation. EMLA has been shown to cause blanching of the skin, possibility making IVC insertion more difficult, and it also carries a rare risk of methaemoglobinaemia.7

Amethocaine gel 4% (Ametop; Smith and Nephew, London, UK) requires an application time of 45 min.8 It has been shown to cause erythema which is thought to be due to vasodilation, and this is hypothesised to improve the first attempt success rate for IVC insertions.9

A recent Cochrane review listed six trials consisting of 534 children aged 3 months to 15 years and reviewed the effectiveness of reducing pain in venous cannulation.9 Participant numbers ranged from 34 to 300 and the comparisons significantly favoured amethocaine over EMLA (RR 0.78, 95% CI 0.62 to 0.98).9 The Cochrane review had two unanswered questions:

(1) Would using amethocaine increase the ability to successfully cannulate veins?

(2) What are the cost implications?

A study was undertaken to determine if amethocaine increases first attempt success rates compared with EMLA, and whether amethocaine provides superior analgesia to EMLA in a children’s emergency environment. The trial also collected information to aid an economic evaluation.

METHODS

Participants and setting

A parallel, randomised, blinded, controlled study was performed at the Starship Children’s Hospital, a tertiary children’s hospital in Auckland, New Zealand. The emergency department sees 30 000 children aged 0–15 years each year. Topical anaesthetic is applied to children aged >3 months thought likely to need a non-urgent intravenous cannula. All children eligible for topical anaesthetic cream were eligible for enrolment in the study.

Interventions

EMLA cream is supplied in a 5 g tube; two sites were selected for application of topical anaesthetic cream per child and covered with a transparent adhesive dressing. Up to four sites may be provided from one tube, but generally only two sites are selected. Amethocaine (provided as Ametop) is supplied in a 1.5 g tube and this was enough for two sites.

Enrolment

The triage nurse explained the trial to the parents/caregivers and requested verbal consent. If caregivers agreed, an envelope containing the anaesthetic cream was randomly selected. The nurse then applied the cream to two vein sites which was typical practice, covered it with the dressing and noted the removal time based on the Starship Children’s Hospital Acute Pain Service recommendations (90 min for EMLA and 45 min for amethocaine). Caregivers were given written consent forms to read. The New Zealand Northern X Regional Ethics Committee, because of the time pressure during the triage process, gave approval for this method of enrolment.

Randomisation

Envelopes were randomly allocated EMLA or amethocaine from a computer-generated random numbers table. The randomisation schedule was kept in a locked room and accessed only when envelopes were prepared. Each envelope contained a tube of either EMLA or amethocaine, parents’/caregiver’s written information, a consent form for parents/caregivers, an assent form for children aged 6–12 years and an assent form for children aged >12 years. A questionnaire was also printed on the parent/caregiver form asking if the cream caused any skin changes; if so, whether this was a problem; was the cream removed early; and would they want to use the cream again. Also placed in the trial envelope were a data collection form for the doctor, two transparent dressings, two stickers for noting the removal time and some cartoon stickers for children who wanted them. Topical anaesthetic tubes were weighed before and after use.

Blinding and masking

With the exception of the triage nurse, all staff and participants were blinded to treatment allocation although staff may be able to guess the allocation from the time of removal or the skin changes. After the doctors inserted the IVC they completed the data collection form, collected the parent/caregiver signed consent form and placed both in a data collection box.

Outcomes

The primary outcome for the study was successful first attempt cannulation and the secondary outcome was observed pain scores in a convenience cohort. The measurement tools were the visual analogue scale (VAS) and the Faces, Legs, Activity, Cry and Consolability Score (FLACC). The VAS scale is easy to use and the data derived can usually be analysed using parametric statistical techniques.10 The FLACC scale has been shown to be reproducible and easy to use.11 Studies show that this scale is valid and reliable in children aged 2 months to 7 years, and there is a positive correlation between FLACC scale and self-reported scales of pain.12 13 The FLACC score has reliability and validity as a measure of pain in children aged 4–19 years.14 Prior to the trial, two observers were educated in the use of the VAS and the FLACC scale and a calibration trial run was conducted. For one-quarter of the observed procedures, a second observer was planned to be present for inter-rater reliability assessment.

Staff training

Extensive staff training was conducted with examples of the paperwork required and the dosing instructions for the two topical anaesthetic creams. As new staff began work in the department, the authors introduced them to the trial and went through the recruitment and consent processes with them.

Non-participants

If any patients opted not to enter the trial they received EMLA as per existing practice. All non-trial EMLA was numbered, weighed and placed in small clear plastic bags. A sticker was placed on the outside of the bag asking why the patient was not entered in the trial. Two options were given with corresponding check boxes: “did not want to enter trial” and “could not enter patient due to work load”. An area was also left blank for any additional comments. The used tubes were collected and weighed.

Power and statistical analyses

The first attempt success rate in our department was calculated from an intravenous cannulation audit.6 It was calculated that to capture a 10% (clinically significant) increase in success rate from 70% to 80% using a power of 0.8 and a p value of 0.05, 330 children would be required in each group. The primary outcome was analysed using the χ² test. Other outcomes include reported skin changes, caregiver preference, early removal of cream and weight of cream used. Descriptive statistics were used for these outcomes.

A convenience cohort (cohort B) was observed for distress based on availability of the primary observer. The primary observer was available for a 10-week period. No sample size was calculated for this secondary outcome as the variability of the VAS and FLACC scores were unknown in the population. The mean scores were analysed with a 95% confidence interval using the Student t test if normally distributed or non-parametric tests if not normally distributed and the results were collected in an Excel spreadsheet (Microsoft, Seattle, Washington, USA). The statistical software used was JMP V.5.1 (SAS, Cary, North Carolina, USA).

RESULTS

From November 2006 to June 2007, 2837 children were enrolled and 809 were known to have had intravenous cannulation. Consent forms for 73 were not returned and data forms for 44 were not returned. A further 13 could not be identified. Complete data and consent forms were returned for 679 children. The flow of participants through the study is shown in fig 1. The outcomes of participants who were randomised but did not have data forms or consent forms completed are shown in table 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow diagram of progress through trial. CED, children’s emergency department; EMLA, eutectic mixture of 25 mg/g lidocaine and 25 mg/g prilocaine.

The baseline characteristics of the study participants are shown in table 2. Comparison results from an intravenous quality assurance project conducted in the same unit before the study6 are included in the table.

Primary outcome

The first attempt cannulation success rate for amethocaine was 75.8% compared with 73.9% for EMLA (p = 0.56, χ² test).

Secondary outcomes

Skin changes were reported by 22.5% of the amethocaine group and 12% thought this was a problem. In the EMLA group, skin changes were noted in 12.6% and 11.6% thought this was a problem. Overall, 61% of all caregivers reported they would use either anaesthetic cream again.

Only 4.4% of EMLA cream and 2.7% of amethocaine cream was removed early. The average weight of cream used from 239 tubes of EMLA was 2.9 g (95% CI 1.3 to 5.2) compared with 1.2 g (95% CI 0.6 to 1.6) from 262 tubes of amethocaine.

Sixty-five children (29 girls, 37 boys) were observed for behavioural changes. There was no statistically significant difference in the VAS scores or the FLACC scale (p>0.05 using both a parametric t test and a non-parametric Wilcoxon test). The results are shown in fig 2A and B. Sixteen children were also observed by a second observer for inter-rater reliability. The R² value for the VAS score during the cannulation was 0.87 (p<0.0001) and for the FLACC score during the cannulation was 0.86 (p<0.0001). No significant adverse affects were seen or reported.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

(A) Intravenous cannulation Faces, Legs, Activity, Cry and Consolability (FLACC) scores. (B) Intravenous cannulation visual analogue scale (VAS) scores for EMLA (an eutectic mixture of 25 mg/g lidocaine and 25 mg/g prilocaine) and amethocaine (Ametop).

DISCUSSION

There was no significant difference between EMLA and amethocaine for success of first attempt intravenous cannulation. The reported blanching of the skin with EMLA and the erythema caused by amethocaine did not have a clinically significant effect on the ability to cannulate children. Compared with placebo, topical anaesthetic appears to increase the cannulation success rate in children, probably by reducing pain, anxiety and movement.15 It has been reported in adults that the skin changes are associated with perceived difficulty in cannulation.16 In a recent study of 203 children, Arendts et al17 found similar results to ours. Their study was powered not to miss a 20% difference between the groups and they obtained a 75% success rate with amethocaine and 74% with EMLA (p = 0.82, χ² test). Our study defined a clinically significant difference as 10%, and since we reached our a priori power requirement, it is unlikely that we have missed a real effect.

The behavioural observations (cohort B) showed no statistically significant difference between EMLA and amethocaine in reducing the pain and distress of intravenous cannulation in the children’s emergency department. A Cochrane review by Lander et al9 significantly favoured amethocaine over EMLA (RR 0.78, 95% CI 0.62 to 0.98) with regard to the effectiveness of reducing pain in venous cannulation. The review combined studies that used both self-reported and observational-based pain scores, and found that self-reported pain scales were the best measure of pain.9 Self-reported scales can only be reliably used by children aged about 4 years or older. As the minimum age in our trial was 3 months, it was appropriate to use observational-based pain scores. The Cochrane review conducted a separate analysis of four studies that based their pain data on observed pain scores18^–21 and concluded that the difference between the two drugs was not significant (p<0.23), “suggesting that amethocaine and EMLA were comparably efficacious in preventing pain from an observer’s perspective”.9 Our study did not find a difference in observed pain scores, which could be due to a type II error resulting from inadequate power for the observed cohort.

One of the limitations of this study was the difficulty in blinding. The two drugs must be applied for different time periods, the volume of the two creams is different and they cause different effects on the skin. This could result in performance bias if the participants, observers or doctors performing the cannulation are aware of these differences. Although only the removal time was noted on the patients’ hands, there is a possibility that the medical staff involved in the trial could have guessed which cream had been used based on this. It is, however, very unlikely that they would have deliberately failed on the first attempt based on which cream they thought had been used. There is also the possibility that observers could have biased the observed pain score if they guessed the cream used. The observers were instructed to ignore time clues and not to look at the site of the cream application to reduce bias.

Tracking the proportion of packs used monitored allocation concealment. After the discovery of tampered envelopes suggesting an attempt at allocation de-concealment, the envelopes were sealed by a non-removable label with a reminder that the amethocaine should only be used for the trial. Furthermore, the tubes were placed in boxes of equal size with stuffing so that the tube size could not be identified. The final number of used and lost tubes was identical between the groups, suggesting that allocation concealment was successful.

Another limitation is the large number of children who were randomised but for whom the forms were not returned. This is probably because they did not have an IVC inserted, although we did not directly measure this. One hundred and seventeen children were known to have had an IVC inserted but we were unable to obtain consent or find forms. Of the large number of blank forms returned, some of these would have had an IVC. Overall, we do not think this would have contributed a large bias effect to the study.

While our study did not find a difference in first attempt successful cannulation, there are other reasons why amethocaine may be preferable to EMLA. First, amethocaine has a much quicker onset of action, which is advantageous in the emergency department. This could save time in the department and therefore associated costs. Second, in the Cochrane review9 amethocaine was found to be a more effective analgesic, which we might have demonstrated had we used a self-reported scale. Amethocaine is also recommended as being able to be used down to 1 month of age compared with EMLA which we currently suggest using down to 3 months. Amethocaine does not have the same risks of methaemagobinaemia as EMLA. The disadvantages of amethocaine include the fact that a smaller area of cream is usually applied and more skin changes occur, although these rarely caused problems in our study. Prolonged application of amethocaine may rarely cause skin blistering, so some vigilance is required to ensure that children do not leave the department with amethocaine cream applied. The amethocaine brand we used is more expensive per gram than EMLA, although generic amethocaine formulations are cheaper.17

The study setting was a children’s emergency department and cannulations are performed by a wide range of medical staff. The study may not be generalisable to hospitals using specialised intravenous cannulation staff such as nurse practitioners. Our results would include a number of unwell or dehydrated children and may not be applicable to well children requiring intravenous cannulation. A longer application time for the topical anaesthetic may have altered the outcome of the trial, although prolonged application times would not be clinically practical.

CONCLUSION

Amethocaine was not more successful than EMLA for first attempt intravenous cannulation in children. Amethocaine has a number of other advantages over EMLA. An economic evaluation follows.