Results

There were no significant differences in the characteristics of the infants in the two groups (table 1). For each feature of the NFCS, the coefficient of reliability between assessors was 0.96, 0.96, 0.94, 0.95, and 1.0 for eye squeeze, brow bulge, deepened nasolabial folds, open mouth, and cry respectively. One infant (amethocaine treated) was excluded from analysis before the code was broken because of restlessness before the venepuncture. In the five seconds immediately before needle insertion, the cumulative NFCS scores were not significantly different in the two groups (2 (0–5)v 0 (0–6) for amethocaine and placebo respectively; median (interquartile range (IQR))). In all, 16 of 19 (84%) amethocaine treated infants showed little or no pain in response to insertion of the needle compared with six of 20 (30%) in the placebo group (p = 0.001). The median cumulative NFCS score over the five seconds after needle insertion was 3 (IQR 0–9) in the amethocaine group compared with 16 (IQR 8–18) in the placebo group (p = 0.001) (fig 1). Logistic regression indicated that the cumulative NFCS score over five seconds before venepuncture did not influence the response to venepuncture. Fifteen of 19 (79%) in the amethocaine treated group compared with five of 20 (25%) in the placebo treated group did not cry at all in the five seconds immediately after needle insertion (p = 0.001). In those who did cry, the median total length of cry was 33 seconds (range 9–79) in the amethocaine group (n = 4) compared with 68 seconds (range 9–122) in the placebo treated group (n = 15). This twofold difference was not statistically significant. Fifteen of 19 (79%) samples were collected at the first attempt in the amethocaine group compared with 13 of 20 (65%) in the placebo group. This difference was not statistically significant. No local skin reactions were seen after application of amethocaine or placebo.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Cumulative neonatal facial coding system (NFCS) in the five seconds after needle insertion in infants treated with amethocaine or placebo. Horizontal bars indicate the median value in each group.

Discussion

In children and adults, topical local anaesthetics have been shown to provide safe and effective anaesthesia for a number of invasive procedures. In the newborn infant, their use has been less successful. Lignocaine and a lignocaine/prilocaine mixture (EMLA) have no demonstrable effect on the pain of heelprick.12-16Gourrier et al 17 reported a positive benefit of EMLA during venous and arterial puncture, although the study was open, uncontrolled, and non-randomised. Some 57% of EMLA treated infants compared with 18% of untreated infants showed a “weak” reaction to venous or arterial puncture. Acharyaet al 2 studied 19 preterm infants in a cross-over trial comparing EMLA with placebo. Using a combination of the NFCS and physiological variables to assess pain, they could show no effect of EMLA. Larsson et al 3 used the NFCS to assess pain of venepuncture in a similar study of 120 term infants. They found a statistically significant difference in the median pain scores after needle insertion and a reduction in the duration of cry in the EMLA treated group. However, the effect was a modest one.

Amethocaine as a 4% (w/w) gel (Ametop) has been developed for use as a topical local anaesthetic by McCafferty et al.18 It is licensed for use in term infants over 1 month of age. Structurally it has a lipophilic benzene ring attached to a tertiary amine group by an ester link chain. This differs from local anaesthetics such as lignocaine and prilocaine, which have an amide link chain. It acts by blocking the influx of sodium ions across the axon and thus the propagation of the action potential along the nerve. It is rapidly metabolised in the blood by plasma pseudocholinesterase. The plasma concentration of amethocaine and its metabolites was found to be unrecordable in seven of 10 adult subjects. In those in whom amethocaine was detected, no significant side effects were reported.19 Amethocaine for pain relief of venepuncture has been examined in three studies in children4-6 and found to be very effective. Transient erythema is common because of its vasodilator effect, in contrast with the pallor of vasoconstriction produced by EMLA.

There are no data on the clinical effectiveness of amethocaine in the newborn. In a pilot study of 20 healthy preterm infants using 1% amethocaine ointment, some evidence of percutaneous absorption was shown, with an increase in local skin blood flow, but local anaesthetic action was not assessed.20 We have shown, using the cutaneous withdrawal reflex to assess local anaesthesia, that amethocaine has a significant local anaesthetic action in the term and preterm newborn.9 The effect is detectable 30 minutes after application and lasts for up to five hours.8 A longer application time of 60 minutes results in a longer duration of action.

In this study we have shown that topical amethocaine gel has a very pronounced effect on the behavioural response of term and preterm infants to venepuncture. The effect is similar in magnitude to the reduction in pain score which has been found in studies of amethocaine gel in children. There were three amethocaine treatment failures. In one infant, the cumulative NFCS score over five seconds was 12 before venepuncture and 11 after, a treatment failure by our predetermined definition but clearly a treatment success. In another infant, the cumulative NFCS scores before and after venepuncture were 8 and 14 respectively, indicating some degree of anaesthesia. In the third infant, amethocaine had no anaesthetic effect at all. Fewer babies in the amethocaine treated group cried in the first five seconds after needle insertion, and in those who did, the total duration of cry was shorter. We deliberately chose not to use a single trained doctor or nurse to perform the venepuncture, as others have done, because we wanted the study conditions to represent the normal working practice of a neonatal unit.

The scoring system we used assesses the four most frequently occurring facial actions in response to an invasive stimulus, plus the presence or absence of crying.10 It is a modification of Grunau's original NFCS which was based on nine facial characteristics.21 ,22 The score has been validated as a tool for measuring procedural pain in the newborn10 and used for this purpose.16 The four facial actions and the cry could be clearly seen and heard on the video recordings. We found excellent agreement between the two assessors of the tapes on all five features.

Information is available on the safety of topical amethocaine gel in children but not in the newborn. Localised transient erythema occurs with a reported incidence of 6–24%. Local pruritus and oedema are less common. In our own experience, we have now applied topical amethocaine gel to 122 newborn infants from 27 to 42 weeks gestation. We have seen two instances of localised and transient erythema. In the current study, we did not observe any adverse reactions. There is no information on the metabolism of amethocaine in the newborn. It has been used in injected form in much higher doses than the topical preparation to provide regional anaesthesia in the newborn, without adverse effect.23 However, prolonged application of amethocaine gel for longer than one hour, repeated applications in a short period of time, and use in the very immature infant in the immediate newborn period should be avoided until more is known about its metabolism.

We conclude that topical amethocaine gel is highly effective in reducing the pain associated with venepuncture in the newborn infant. Further studies to assess its effect on other invasive procedures are needed.