Suckling at the breast is better than sweet solutions and pacifiers

EDITOR—A breastfeeding mother spontaneously comforts her distressed infant by putting him or her to the breast. It is a pity that Carbajal et al, when assessing the analgesic effects of orally administered glucose and sucrose in healthy term neonates, did not include a comparison group of infants given breast milk, which is rich in lactose and naturally sweet.1 It would also have been appropriate to compare the analgesic effect of using a pacifier with that of suckling at the breast before and immediately after the painful procedure.

Pacifiers and sugar solutions given unnecessarily to healthy neonates are not proved to be “simple and safe interventions,” as Carbajal et al state. Exclusive breast feeding (for about the first six months) is the World Health Organization's recommendation.2 Two of the evidence based “ten steps to successful breastfeeding,” developed by the WHO/Unicef Baby Friendly Hospital Initiative, are step 6 (“Give newborn infants no food or drink other than breast milk, unless medically indicated”) and step 9 (“Give no artificial teats or pacifiers, also called dummies or soothers, to breastfeeding infants”).3

Anything that may interfere with the establishment of lactation or undermine the mother's confidence in breast feeding is to be avoided. I hope that this flawed piece of research will not result in either pacifiers or sugar solutions being “widely used for minor procedures in neonates.”1

Use of pacifier may modify responses measured on rating scale

EDITOR—Carbajal et al conclude that the analgesic effects of both pacifiers and sweet solutions are clinically apparent and that pacifiers are more effective than sweet solutions alone.1 As a measure of pain they used a rating scale, douleur aiguë du nouveau-né (DAN), which has been described previously.2

This scale uses facial expression, limb movements, and vocal expression to give a score between 0 and 10. Low scores mean no or little pain, and higher scores mean that the infant experiences more pain. It is apparent that the results in the two groups treated with pacifier alone and with pacifier combined with sweet solution differ from the results in the other groups in two ways: the groups whose treatment included a pacifier have a lower mean score and show a less varied response to the stimulus of venepuncture.

I would suggest that the less varied response to the stimulus is due to the pacifier itself. The ability to express a range of facial expressions will be modified by sucking on a pacifier in a way that reduces the possible responses on the rating scale. It would be interesting to see ratings of infants who do not have venepuncture and their ratings on the rating scale with and without pacifier.

Surely evidence is not needed to justify cuddling babies in pain

EDITOR—The results of Carbajal et al's trial were not surprising.1 The study concluded that non-nutritive sucking on a pacifier (dummy)was more successful in producing analgesia in neonates during venepuncture than the use of glucose or sucrose solutions.

In the accompanying editorial Choonara states that “parents know that a crying baby needs comforting and will hold their infant close. Breastfeeding mothers will give their infants the opportunity to breast feed, even if they are not hungry.”2 Choonara tells us how the study confirms that these actions of mothers are appropriate, but I question whether we need a scientific study to support deeply engrained human responses that have been witnessed to work throughout the ages.

The paper's authors comment that the mechanism by which pacifiers induce analgesia is unknown. I would offer instinct and conditioning as two possible mechanisms. Most parents know to hold their distressed offspring, but it is equally true that children naturally seek proximity to their attachment figures when in pain or distress. The pacifier mimics the breast, and psychobiological processes associated with the mother-child relationship and reduction in distress are perhaps triggered in the infant; the relation between reduction in distress and reduction in perception of pain is well recognised.

A growing body of evidence suggests that disruption of the natural, instinctive “knowledge” and interaction between mother and child can have detrimental effects on emotional development; the effect of postnatal depression is an example.3 We can all recognise a basic need for safety and security, and it seems unsurprising that children need the same to develop emotionally.

Choonara believes that the use of interventions including cuddling for pain control needs to be evidence based. Why? As he acknowledges, parents will not change their behaviour in the face of research findings. They would be right not to: they know that cuddling works. Science does not offer us absolute truths; instead it offers us answers with a given degree of certainty. This certainty will never be large enough to justify attempted suppression of the natural, magical responses evoked in a mother by her distressed baby.

Measures of pain must be validated in young children

EDITOR—Using a behavioural pain score (facial expression, movement, and vocal expression), Carbajal et al concluded that oral sugar solutions and non-nutritive sucking were analgesic in neonates.1 Their study shows not analgesia but a reduced behavioural response to pain in neonates.

There are two main problems: firstly, what is meant by the term analgesia, and, secondly, how can you check the validity of a pain scoring system? Analgesia is defined as the absence of pain on noxious stimulation, and pain is defined as always a subjective experience.2 Subjective self report measures of pain that are used in adults are impossible to use in children before they can talk. Therefore proxy responses such as behaviour (loudness and duration of cry, facial expression, body posture, mobility, alertness), physiology (heart rate, blood pressure, respiratory rate, palmar sweating, oxygenation, intracranial pressure), and endocrine responses (concentrations of cortisol, catecholamine, glucose) are used. The difficulty is that these are all non-specific markers and are influenced by factors such as fear, anxiety, and medical problems.

The sole use of behaviour as a measure of pain can be misleading.3 “Sweet flavoured pacifiers can calm a crying baby but should never be regarded as providing major analgesia.”3 Likewise measurement of the endocrine response alone is inadequate.4 An example is that dummies (pacifiers) reduce the behavioural response to pain (sleep, alertness, crying) but do not reduce the endocrine response (cortisol concentration).5 Colloquially, we do not use the term analgesia to describe techniques such as distraction or rubbing of a sore leg, which influence only one dimension of pain.

The second problem is that scoring systems can be tested for internal validity such as consistency between different observers and showing that there is an increased response to what is perceived as increasingly painful circumstances. Rarely, pain scores are validated against other pain scoring systems, but we have no gold standard for comparison.

This is more than semantics. We need to validate our clinical scoring systems against all dimensions of pain, including behaviour, physiology, endocrine response, development, culture, and environment. Until we have better measures of pain in children before they have can talk we should be wary of concluding more than the evidence shows. If all that sugar and sucking does is reduce the external expression of pain then we are treating ourselves, the carers, rather than the children.

Trial of drug treatment to alleviate pain in neonatal intubation is needed

EDITOR—Carbajal et al address the issue of neonatal pain relief,1 and in the accompanying editorial by Choonara health professionals are encouraged to study the painfulness of clinical procedures and use measures to prevent pain.2

Neonatologists agree on the importance of pain relief for newborn babies undergoing invasive procedures, such as placement of a chest drain. Analgesia decreases the incidence of accidental extubation and provides pain relief for ventilated newborn babies.3 Yet for the most painful procedure—intubation—neonatologists in the United Kingdom have been cautious. American and Australian neonatal intensivists have been using intravenous drugs for some time when patients need intubation.

Evidence suggests that intubation done when the neonate is awake is associated with higher spikes in intracranial pressure than that done when general anaesthesia is given.4 Even awake neonates who have been paralysed have shown significant increases in mean arterial blood pressure and intracranial pressure, indicating the effect of pain.

Adequate anaesthesia prevents swings in blood pressure and thus potentially the development of intracranial haemorrhage. Although one study showed that there was no hypertensive response in neonates who were intubated while awake, changes in intracranial pressure were not measured.5

So that we know whether or not to follow the practice used abroad, a trial of drug treatment to prevent pain in neonatal intubation is necessary.

Authors' reply

EDITOR—Currently, most minor procedures in neonates are performed with little or no analgesia. We found that pacifiers can effectively relieve pain and that the analgesic effect is synergetic with sucrose. To our knowledge, this latter effect, which has practical implications, had not been reported before our study. Recently, Blass and Watt reported the same effect.1

Campbell wonders why we did not include breast milk or suckling at the breast before and immediately after painful procedures. Our study included six groups, and it would have been difficult to add more—firstly, because of the masking constraints imposed by a randomised double blind study, and, secondly, because we aimed to compare the analgesic effects of non-nutritive sucking with those of sugary solutions. A single study cannot answer several questions at once.

Analgesia induced by milk has been shown in newborn infants.2 This effect was not related to lactose and was modest compared with that of sugar. We do not agree with Campbell when she states that pacifiers and sugar solutions given for analgesic purposes are given unnecessarily. We adhere to the “ten steps to successful breastfeeding” and consider that, with regard to step 6, relief of neonatal pain with pacifiers and sugar solutions is medically indicated. We do not think that occasional use of pacifiers or small volumes of sterile glucose or sucrose solutions for a painful procedure should be regarded as equivalent to frequent or routine use.

Blomstrand suggests that the less varied response to venepuncture in infants treated with a pacifier may be due to these infants' inability to express a range of facial expressions during the treatment. This hypothesis can be rejected for at least two reasons. Firstly, when one evaluates facial expressions on the rating scale that we used the modification of only one of eye squeeze, brow bulge, or nasolabial furrow is enough to determine the intensity of this item. Eye squeeze and brow bulge are not incompatible with sucking. Secondly, infants who suck a pacifier make several pauses between sucking bursts, which gives them the opportunity to express grimacing.

Morris asks if evidence is needed to justify cuddling a baby in pain. Obviously not. However, studies that determine the efficacy of behavioural interventions in preventing pain in neonates are welcome for, as Choonara states,3 they can encourage health professionals to modify their behaviour. Morris's hypothesis offering instinct and conditioning to explain the mechanism by which pacifiers induce analgesia is interesting. Blass and Watt have suggested that antinociception and pain blockades induced by orogustatory and orotactile mechanisms are likely to be occurring at the level of the dorsal horn of the spinal cord.1

Rogers's concern about the difference between analgesia and reduced behavioural response to pain is theoretically valid. As he states, pain has been defined as a subjective experience. This definition has led to many advances, but it challenges our understanding of pain because it does not apply to living organisms that are incapable of self report. This includes neonates and older infants and many adult patients.4

The biological and behavioural reactions to pain are evident in term and preterm neonates. The fact that neonates' expression of unpleasantness does not fit within the strict definition of pain contributes to the failure to recognise and aggressively treat pain in children.4 Increasing evidence supports the specificity of facial expressions as a manifestation of pain in neonates. As we stated in our paper, we assumed that the more pronounced the facial expressions, limb movements, and vocal expressions the greater the pain in the neonates.

Rashid makes an important point. Endotracheal intubation is a powerful noxious stimulus with potential adverse effects. Although premedication is mandatory for endotracheal intubation in adults, most neonatal units do not sedate neonates before intubating them. Recently, Bhutada et al showed that the heart rate and blood pressure of neonates who are premedicated with thiopental before intubation remain nearer to baseline values than do those of similar infants not given premedication.5