Dear Editor

We would like to highlight the varying practices in the management of cardiac murmurs in well newborn infants. Two years after the publication of Wren's important paper on this subject[1] urging an early definitive structural diagnosis to be made on such infants we conducted a telephone survey of local hospitals to assess whether practices had changed.

Until recently it was common practice to investigate a well neonate with a 12-lead electrocardiogram (ECG), a chest radiograph (CXR), 4-limb blood pressure and oxygen saturation. If these investigations were normal, infants were discharged home, to be reviewed later. Wren's paper reminds us that there is a structural defect in 54% of newborns with a persisting murmur. Some of these structural defects (particularly outflow tract obstructions) may be fatal if not treated early. Wren recommended therefore that an early definitive diagnosis should be made within the first month of life in this group.[1,2]

We obtained data from 14 district general and 6 teaching hospitals. Whilst one third (5/14) of district general hospitals complied with recommendations two thirds (9/14) did not, primarily because echocardiography was not available on-site. In the latter group, management consisted of investigations (excluding echocardiography) prior to discharge and follow-up at 4-6 weeks in a general paediatric or neonatal follow-up clinic. Two hospitals did not routinely follow neonates with apparently "benign" heart murmurs. One hospital did not investigate well infants with a murmur prior to discharge. Teaching hospitals complied with Wren's recommendations with two thirds (4/6) obtaining an echocardiogram before discharge and the remainder obtaining one within 2 weeks.

In light of this survey we feel it is important that all perinatal units are enabled to adopt the evidence-based recommendations by the provision of adequate paediatric echocardiographic resources.

References
(1) Wren C, Richmond S, Donaldson L. Presentation of congenital heart disease in infancy; implications for routine examination. Arch Dis Child 1999;80:F49-F53
(2) Richmond S and Wren C. Early diagnosis of congenital heart disease. Semin Neonatol 2001;6:27-35

Dear Editor

We have read with interest the article by Oddie et al [1] and subsequent correspondence from Williams [2], Harding et al [3] about weighing breast fed babies. With the ongoing shortage of midwives throughout the UK, staffing levels on postnatal wards are low. We are increasingly worried about the decline in support available to mothers trying to establish lactation in the days immediately after delivery. Early postnatal discharge from maternity units is now the norm, usually within 24 hours of delivery, and by 48-72 hours after LSCS in our hospital. We report 3 breast fed term infants who have presented within the past 6 months with profound hypoglycaemia, neonatal convulsions and significant dehydration which we attribute to undernutrition in the first 3 days after birth.

All 3 infants have abnormal MRI changes compatible with white matter infarction secondary to hypoglycaemia and all are showing signs of significant neurodevelopmental delay within 6 months of birth. Neither the mothers nor midwives recognised any worrying signs in the infants' pattern of suckling, tone or state of alertness until the onset of convulsions (2 infants), and repeated cyanotic episodes (1 infant).

Case 1
A 4 day old infant (birth weight 4320g, OFC 37 cm) presented with a 48 hour history of poor feeding and jitteriness.The baby was severely dehydrated with a weight loss of 18% since birth. Convulsions commenced soon after presentation. Laboratory blood sugar was unrecordable, Hb 19.2 g/dl, Na 154 mmol/l, creatinine 116 umol/l, urea 27.8 umol/l. The baby had been discharged from hospital at 7 hours of age after a spontaneous, uncomplictaed vaginal delivery at term, and the baby was breast fed. Emergency resuscitation was required and intensive care for 4 days. No underlying infective or metabolic cause was identified, although insulin levels were not measured at the time of presentation when the infant was hypoglycaemic. Thrombophilia screen was normal. MRI scan of brain at 13 days of age revealed a widespread abnormal signal intensity throughout the white matter consistent with parasagittal infarction, a pattern recognised in hypoglycaemic brain injury. EEG at 3 weeks of age was abnormal and the infant has required ongoing anti-convulsants for clinical seizures. OFC has fallen from 90th to 25th centile by 2 months of age.

Case 2
An infant (birth weight 2550gm, OFC 33.5 cm) was transferred to NICU at 70 hours of age after a cyanotic episode on the postnatal ward. The baby had been delivered by semi-emergency LSCS because of growth restriction and CTG signs consistent with placental insufficency at 39 weeks gestation, but was in good condition at birth (Apgar scores 8 @ 1;10 @ 5 mins). Laboratory blood glucose was unrecordable on admission to NICU. Normoglycaemia was restored with an intravenous infusion of 10 % dextrose (7 mg/kg/min). Weight loss of 12% since birth was documented on admission to NICU. At 80 hours of age, a left sided focal convulsion occurred. No infective or metabolic cause was identified apart from hypoglycaemia. The baby had been breast fed. Cranial ultrasound on day 4 revealed a bright echodensity in the caudothalamic notch. MRI scan of brain on day 5 showed multiple haemorrhagic lesions in the white matter consistent with hypoglycaemic brain injury. In addition there was a small infarct in the right lentiform nucleus extending into the internal capsule. The aetiology of the infarct is uncertain and the results of the thrombophilia screen are awaited.

Case 3
A 60 hour old infant (birth weight 2760g, OFC 31.7 cm) was admitted with lethargy, poor feeding and convulsions within 9 hours of discharge home from the postnatal ward. He had been born by semi-elective LSCS because of failing fetal growth and oligohydramnios. No resuscitation was required at birth. Apgar scores were 9 @ 1;10 @ 5mins. He was breast fed. Laboratory blood sugar was unrecordable on readmission. An intravenous infusion of glucose at 13 mg/kg/min restored normoglycaemia. Weight loss of 6% from birth was documented. Infection, metabolic and thrombophilia screen revealed no other abnormality. MRI scan of brain at 11 days of age showed posterior and temporal white matter infarction consistent with hypoglycaemic brain injury. OFC has fallen from 10th to well below 3rd centile by 7 months of age and he has significant neurodevelopment delay.

These 3 infants presented with significant hypoglycaemia within 4 days of birth and all have abnormal MRI scans consistent with hypoglycaemic brain injury. Transient hyperinsulinaemia was not excluded in any of the infants as insulin levels were not measured at the time of initial presentation with hypoglycaemia.The glucose requirement to restore normoglycaemia was not particularly high, however, and as there was no recurrence of hypoglycaemia in any of the infants after commencing the infusion of glucose, transient hyperinsulinaemia seems unlikely. Two infants (cases 2 and 3) were delivered by LSCS because of signs of placental insufficiency, and therefore may have been at risk of hypoglycaemia, but as they were in good condition at birth and birthweight was not <10th centile, the mothers and infants all received "normal" postnatal care. The reason for the hypoglycaemia in case 1 is unclear. The mother showed no overt signs of gestational diabetes, and transient hyperinsulinaemia remains a possible explanation. No other underlying pathology in the infants was demonstrated. All infants are showing signs of neurodevelopmental delay and feeding problems and two have developed microcephaly. Because of the concomitant weight loss (> 12 % in cases 1 and 2), we have concluded that undernutrition whilst attempting to establish lactation is the most likely cause of the hypoglycaemia.

We are painfully aware that there is a low midwife/ mother ratio on our postnatal ward and that mothers may not always be able to receive optimal supervision and support whilst attempting to establish lactation. The use of lay breast counsellers and appropriately trained health care assistants to support breast feeding mothers, particularly primipara and mothers who may benefit from additional support after LSCS, is being explored. We support the view of Williams that experienced midwives can assess the establishment of lactation and clinical hydration of the baby by observation of feeding pattern and other clinical signs without having to resort to early weighing. However, there is a national shortage of midwives and we believe that very few maternity units have sufficient midwives to provide optimal support for all mothers to establish successful lactation. Discussion with other paediatric colleagues suggests that they are also seeing a resurgence of dehydration and hypoglycaemia in a small number of breast fed term infants.

Healthy term breast fed babies have been shown to tolerate asymptomatic hypoglycaemia during early lactation (4), but infants at risk of hypoglycaemia because of fetal undernutrition may require additional feeding or glucose intake in the first few days after birth. Birth weight <10th centile is commonly used to identify babies who might benefit from routine blood glucose monitoring in the first few days after birth, but it is only a crude indicator of which babies may be at risk of hypoglycaemia, particularly if postnatal nutrition is inadequate. Clear guidelines for health care workers about the clinical assessment of adequacy of breast feeds to include observation of suckling pattern, state of alertness, frequency of stools and wet nappies, particularly after 12 hours of age, may help to prevent dehydration and hypoglycaemia. We would not dismiss the potential benefit of weighing some, but not all, breast fed babies on 2nd or 3rd day after birth, as this may identify babies with significant weight loss before more serious symptoms develop. The challenge remains to prevent serious complications in a minority of potentially vulnerable infants without undermining the establishment of successful breast feeding for the majority of normal term infants.

R Thomas
F Isaza
W Hyer

References (1) Oddie S, Richmond S, Coulthard M. Hypernatraemic dehydration and breast feeding: a population study. Arch Dis Child 2001;85: 318-320.
(2) Williams A . Weighing breast fed babies. Arch Dis Child Fetal Neonatal Ed 2001;86:F69.
(3) Harding D, Cairns P, Gupta S, et al. Hypernatraemia: why bother weighing breast fed babies? Arch Dis Child Fetal Neonatal Ed 2000;85:145.
(4) Hawdon J, Ward Platt M P, Aynsley-Green A. Patterns of metabolic adaptation for preterm and term infants in the first neonatal week. Arch Dis Child 1992;67:357-365.

Dear Editor

Dr Katumba-Lunyenya rightly recognises that management of common cardiological problems in neonates is beyond the resource provision for congenital heart disease specialists in the UK. The author also realises that along with the echocardiographic skills come the “innappropriate referrals”. Sadly there is no doubt that the increased reliance on investigations of all types (particularly the echocardiogram) has been mirrored by a decline in cardiac clinical skills amongst paediatricians of all grades. In some of our outpatient clinics (Yorkshire heart centre) over 50% of new referrals from specialist paediatric staff are innocent murmurs. Our data support the view that these referrals (described as innappropriate by the author) are increasing from both District and Teaching hospitals and that a high proportion come from junior doctors, the patient having been examined only once (during an intercurrent illness) and often not seen by a consultant. If we were to follow the authors advice we would be “firmly refusing” hundreds of referrals. Similar problems arise on neonatal intensive care units, and whilst the echocardiogram can be a powerful diagnostic tool in this setting, there is no evidence base to suggest that serial echocardiography improves outcome in neonates. Systolic dysfunction is rarely a cause of hypotension in children or neonates on intensive care units and ultrasound is of very limited value in the assessment of ventricular filling in this setting.

The RCP SAC in Paediatric Cardiology recognises the implications of increasing demands on paediatric cardiac services. With the Royal College of Paediatrics and Child Health a working party has been established to develop a curriculum of training for paediatricians with a special expertise in paediatric cardiology to increase the opportunities for paediatricians to develop clinical skills in cardiology (of which echocardiography is just one) relevant to the management of these patients.

Dr John Thomson
Specialist registrar and trainee representative SAC
Paediatric Cardiology

Dr John Gibbs
Consultant paediatric cardiologist and secretary SAC
Paediatric Cardiology

Both Yorkshire heart centre, Leeds general infirmary NHS Trust.

Dr Sue Hobbins
Consultant Paediatrician with special expertise in Paediatric Cardiology
and RCPCH representative on SAC in paediatric cardiology.
Bromley Hospitals NHS Trust.

Dear Editor

I would like to thank Dr Thomson and colleagues for their comments. I need to clarify on an issue they picked upon which when quoted in isolation gives the reader the wrong impression. They seem to suggest that I am advising the reader that inappropriate referrals are “firmly refused” out of hand. What I actually meant, if you read on, was that one should refuse to see referrals from other colleagues which have not been properly clinically assessed.

My colleagues have listened to me resulting in a dramatic drop in numbers of innocent murmurs presenting to my clinic and to the two outreach clinics run by two cardiologists, one from Oxford and the other from the Brompton. They are, however, encouraged to refer for echocardiography if in doubt.

I was a little disturbed by the statement made by Dr Thomson et al that systolic dysfunction rarely causes hypotension in the neonate in intensive care. I can not presume to comment on cause of hypotension in bigger children in PICU, as I never scan this group of patients. There is, however, no doubt that several authorities believe that poor myocardial contractility together with "systemic steal" through the often silent patent ductus arteriosus, and abnormal peripheral vasoregulation are the three commonest contributors to hypotension in neonates especially the extreme preterm during the first 24 to 72 hours of extrauterine life [1-8]. Hypovolaemia, however, is the rarer cause of hypotension in this setting. A D Gill and A M Weindling [2] recognised the contribution of poor myocardial contractility to hypotension years ago. There is a massive volume of authoritative literature to support the above statements and the February 2001 issue of Seminars in Neonatology co- edited by Nick Archer and Nick Evans covers this topic extensively as well as providing a huge source of references. Dr Thomson et al do not give any references to support his view and I am not personally aware, though I stand to be corrected, of any such literature in the last few years.

Unlike in the adult and bigger child where it is easy to invasively assess haemodynamic changes on the ICU, this is not as easily done in the neonate and echocardiography provides a useful non-invasive tool for this purpose on the NICU. The duct is also usually "silent" in the first three to five days even when open and can only be diagnosed by echo. Recent evidence suggests that ibuprofen is at least as effective as indomethacin in closing the duct without the co-morbidity (decreased cerebral, renal and gastrointestinal perfusion etc.) associated with indomethacin [9,10]. This would make medical closure of the duct much more acceptable even for those paediatricians who have rightly so, hitherto, been worried about the complications associated with indomethacin therapy.

If one accepts the fact that hypotension is associated with increased morbidity and mortality [2] and that silent ducts can cause problems in preterm neonates, then one can reasonably hope that "sequential echocardiography"[7] during the first 72 hours, and later as necessary, might reduce morbidity and mortality by improving assessment and management of these problems.

Moreover Paediatricians and Neonatologists are already voting with their feet and the Echo course at my tertiary centre is usually oversubscribed by 100% because they are beginning to recognise usefulness of echocardiography in neonatology.

I am pleased that the RCPCH SAC in Paediatric Cardiology is looking at training paediatricians to develop the skills of cardiology but I sincerely hope that the curriculum will include echocardiographic assessment of the duct and hypotension in the sick preterm infant and its impact on management. Not to include it will be a lost opportunity, which I predict will only have to be redressed later because of the mounting weight of evidence that echocardiography, is a very useful tool in this setting.

References
(1) Jasper Katumba-Lunyenya. Neonatal/Infant echocardiography the non- cardiologist: a personal practice, past present and future. Arch Dis Child Fetal Neonatal Ed 2002: 86: F55-57.
(2) AB Gill, AM Weinding. Echocardiographic assessment of cardiac function in shocked very low birthweight infants. Arch Dis Child Fetal Neonatal Ed 1993; 68: 17-21.
(3) Nick Evans. Diagnosis of patent Ductus arteriosus in the preterm newborn. Arch Dis Child Fetal Neonatal Ed. 1993; 68: 58-63.
(4) Jon Skinner. Diagnosis of patent ductus arteriosus. Semin Neonatology 2001; 6: 49-61.
(5) Evans N, Moorcroft J. Effect of patency of the ductus on blood pressure in very preterm infants. Arch Dis Child 1992; 67: 1169-1173.
(6) David B Knight. The treatment of ductus arteriosus in preterm infants. A review and overview of randomised trials. Semin Neonatal 2001; 6: 63-73.
(7) Martin Kluckow, Nick Evans. Low systemic blood flow in the preterm infant. Semin Neonatal 2001; 6: 75-84.
(8) Istvan Seri. Circulatory support of the sick preterm infant. Semin Neonatal 2001; 6: 85-95.
(9) B Van Overmeire et al. A comparison of Ibuprofen and Indomethacin for closure of Patent Ductus Arteriosus. New England Journal of Medicine; 2000; 343:674-681.
(10) Jayesh Patel, Idris Roberts, Dennis Azzopardi, et al. Randomised Double – Blind Controlled Trial Comparing the Effects of Ibuprofen with Indomethacin on Cerebral Haemodynamics in Preterm Infants with Patent Ductus Arteriousus. Paediatric Research 2000; 47: 36-42.