DRY ARTIFICIAL LUNG SURFACTANT IN PREMATURITY

555 TABLE I-OUTCOME IN VENTILATED INFANTS WITH

PNEUMOTHORAX, MECHANICAL VENTILATION, AND PERIVENTRICULAR HAEMORRHAGE

AND WITHOUT PVH

SIR,-The observation by Dr Lipscomb and colleagues (Feb. 21, p. 414) that pneumothorax in preterm infants receiving mechanical ventilation significantly increases the incidence of periventricular haemorrhage (PVH) accords with a current hypothesis concerning the aetiology of PVH.1 We have been doing a similar study, using a real time ultrasound sector scanner (ATL 850A), over the past year, but we have been unable to demonstrate such a strong association between PVH and either pneumothorax or all pulmonary air leaks.2 During 1980,130 infants with a birthweight of 500-1500 g were admitted to the neonatal unit at the Liverpool Maternity Hospital; 52 had been born elsewhere. 108 infants required mechanical ventilation, mainly because of respiratory distress syndrome. All infants were examined via the anterior fontanelle3,4 frequently during the first few days of life by ultrasound, and 56 infants were found to have PVH. All were in the group receiving mechanical ventilation. A similar grading system to that of Papille et al.was used, and data on birthweight, sex, incidence of pneumothorax and other pulmonary air leaks, blood gas tension, and early complications were recorded for all infants. As PVH only occurred in ventilated infants, tablerefers to, just these 108 babies. Although pneumothorax and air leaks were more frequent in infants with PVH this was not significant. Hypercapnia and metabolic acidosis were significantly more frequent in PVH infants during the first 48 h of life. Although these blood gas data were obtained from the results of intermittent sampling rather than continuous measurement in most cases, they did appear to be representative of the severity of the individual infants’ disease during this period. Equally, no significant difference in the distribution of the maximum grades of PVH was observed in the infants with a pneumothorax or without (table n). We are unable to explain the differences between the two studies with regard to the influence of pneumothorax. The incidence of pneumothorax in ventilated infants in our study is higher (32%) but not significantly so (chi2 =2-7). The birthweights are lower, but this did not seem to alter the association of pneumothorax and PVH (table III). Differences in management may be important, although similar indications for and techniques of mechanical ventilation are being used in both units. We wish to imply, not that pneumothorax plays no part in the aetiology of PVH, but that since most infants with PVH do not have a pneumothorax, other factors which may be associated with pneumothorax may be more important in the development of PVH. Amongst other possible aetiological factors, we have looked at early management complications. A common problem in ventilated infants is obstruction of the endotracheal tube. If simple measures such as suction fail to improve matters, the tube is replaced. Such replacements are recorded on the intensive care charts and the relative frequency of tube replacement in the first 2 days of life is shown in table 1. A significantly increased incidence of tube blockage is seen in the PVH group. Tube blockage causes hypoventilation with resulting hypercapnia and hypoxia. Relief of the obstruction produces a sudden improvement in ventilation and increase in cardiac output in the face of a dilated cerebral vascular bed. Such a rebound effect has been described after recovery from apnoea in the preterm infant,6 and could equally account for an increase in incidence or spread of PVH after either pneumothorax or an obstructed airway. ischaemia and the perinatal brain. Clin Devel Med 1979; no 69/70. 2. Dykes FD, Lazzara A, Ahmann P, Blumenstein B, Swartz J, Brann AW. Intraventricular haemorrhage: a prospective evaluation ofetiopathogenesis. Pediatrics 1980; 66: 42-49 3 Cooke RWI. Ultrasound examination of neonatal heads. Lancet 1979; ii: 38. 4. Cooke RWI Factors associated with periventricular haemorrhage in very low birth weight infants. Arch Dis Child (in press). 5 Papille LA, Burnstein J, Burstein R, Koffler H. Incidence and evolution of sub ependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1500 grams. J Pediatr 1978; 92: 529-34. 6 Cooke RWI, Rolfe P, Howat P. Apparent cerebral blood flow in newborns with respiratory disease. Devel Med Child Neurol 1979; 21: 154-60. 1.

Pape KE, Wigglesworth JS. Haemorrhage,

*

Worst values recorded in first 48 h of life. 1 kPa=7.5 mm

Hg.

TABLE II-GRADE OF PVH

TABLE III-BIRTHWEIGHT AND PREVALENCE OF PNEUMOTHORAX AND PVH

(OR BOTH)

The relative importance of aetiological factors in PVH will depend on the population studied and management techniques used. In any condition in which multifactorial causation is likely, no single factor alone can be implicated in its production. Regional Neonatal Intensive Care Unit, Department of Child Health, Liverpool Maternity Hospital, Liverpool L7 7BN

RICHARD W. I. COOKE IMOGEN M. MORGAN NIGEL A. G. COAD

DRY ARTIFICIAL LUNG SURFACTANT IN PREMATURITY

SIR,—Dr Morley and colleagues (Jan. 10, p. 64) are to be commended for their efforts in developing dry artificial lung surfactant. Their preliminary results are encouraging, but we would criticise their study on two counts. The two groups are not comparable. In the treated group female babies, vaginal delivery, prolonged rupture of membranes, preeclampsia, and the use of beta-stimulants and steroids were more frequent. All these factors carry a reduced risk of the development of surfactant deficiency disease. The treated babies were resuscitated differently, in that a neonatal specialist (Dr Morley) attended their birth and gave the powdered surfactant by a method which (by moderate hyperinflation of the lungs) by itself might have conferred benefit. There was no objective assessment of the frequency of surfactant deficiency as the cause of the respiratory failure operating from birth. Leicithin:sphingomyelin ratios on gastric aspirates would probably have been too crude and phospholipid profiles of tracheal aspirates would have been far superior. Department of Obstetrics and Child Health, Saint Mary’s Hospital,

DAVID JAMES

Manchester M13 0JH

ALEX HARKES

556

SIR,-I must take exception to the study design, and therefore the conclusions, of the investigation by Dr Morley and colleagues. By failing to randomise infants they have biased their results. The treated infants were cared for differently from the controls in that one additional expert (Dr Morley) was present in the delivery room for the treated infants only; I find it very difficult to believe that his effect upon the care of the infant. In most cirknowledge and more hands result in quicker action and more rapid stabilisation of the infant. If this results in the infants entering the special care baby unit in better condition than the controls (and it should), the treated infants would be expected to have a lower mortality rate, regardless of surfactant treatment. presence had

no

cumstances more

Unfortunately, the dedication and patience lavished upon the basic biochemistry of such studies is often forgotten or exhausted when it comes to the clinical trial. It would have been possible to randomise all infants born when Morley was present, disregarding others. Even a single or double blind study would be possible by using an empty, opaque gelatin capsule for half of the infants. Even if the individual administering the medication became unblinded to the assignment, it would still be possible to avoid passing this information along to the physicians caring for the infant. It is now necessary to do this study again, properly controlled this time. How much easier it would have been to do it correctly the first time. Now scientific and ethical issues will be more difficult for the research review committee and also for the Cambridge team for they have obviously convinced themselves of this surfactant’s efficacy. This

study seems to have been approved by the maternity hospital’s ethics committee without insistence on informed consent, presumably because it would have been impossible to obtain such consent in the circumstances. This is, in my opinion, an appropriate application of the authority of such committees. Unfortunately this particular study raises the issue of whether it is ethical to do a study (even a very low risk one) if the design is faulty. I think not. Ifa study design is such that it can not prove or disprove what it has set out to test, then the risks (however small) taken by the participants are unjustifiable.

safe

bedding

for babies which has advantages

over

traditional

materials. 1,2

Natural and artificial sheepskins are very different materials. As far as we know there have been no published evaluations of the use of artificial sheepskins for nursing babies. The ones we have seen are less resilient and more slippery than the wool products, and there is a greater tendency for the fibres to come free. For these reasons we do not think they make a suitable bedding material for the newborn. We have been using three types of natural lambswool in our studies. Two have the original leather backing (’Babycare’ and ’Winganna’) and the other is made by weaving lambswool into an artificial backing (’Dermalex’). Before use, the washed mat should be brushed with a stiff wire brush which will remove any loose fibres. Occasionally, a baby seems to acquire the habit of sucking its lambswool. In such cases, as an additional precaution, a piece of cotton cloth can be placed beneath the baby’s head. Child Care and Development Group, University of Cambridge,

Cambridge CB2 3RF

STEPHEN SCOTT MARTIN P. M. RICHARDS

SOURCES OF ERROR IN FRAGILE-X DETERMINATION

SIR,—X-linked mental retardation displaying the fragile X chromosome seems to be a very common cause of mental defect and soon it will be necessary for every cytogenetic laboratory to use this investigation and to know the possible sources of error. The letter from Dr Leversha and her colleagues (Jan. 3, p. 49) prompts us to report our experience based on about fifty cases with fra(X). Leversha et al. reported other C group chromosomes (most of them no. 6) which can be misinterpreted as fragile X and referred to the dilemma-either to band the chromosomes, with resulting poor morphology of the fragile site, or not to band, and face the possibility of an error.

SIR,-I acknowledge the criticism of Dr James and Dr Phelps but to make the following points. The trial was a preliminary study of the effects of pure artificial

We solved this problem in a different way: the fragile sites are located using unbanded chromosomes. Then the slides are destained with acid alcohol, banded,3and then the fragile sites are re-checked. Using this method we identified all fragile-like chromosomes in six carriers and in twenty-five non-carriers of fragile X. The results were similar in both groups. In about 1-2% of cells chromosome no. 6 is involved, and 0 -9% other C-group chromosomes simulate fra(X) morphology. It means that, for example, in patients with 10% of fra(X), out of ten positive cells two are possibly false positive. For patients with high level positivity this error does not change the diagnosis but in low level positive patients there is a definite possibility of diagnostic error. In a low level subject the exact identification of each fra(X) is essential.

surfactant in premature babies. We do not claim that the results are definitive or that the two groups are accurately matched. However, we do show that one small dose of dry artificial surfactant powder given at birth may have some beneficial effects and that, in consequence, this important subject warrants other well conducted clinical trials.

We do not think that C group chromosomes (and very often no. 16 and other chromosomes) with satellite-like appendix on long arm should be always considered as "fragile sites" in the strictest sense. 4-7 In contrast with real fragile sites a proportion of these can be found in nearly every subject but their level is always very low. Probably they are common chromosomal gaps located at the end of

I am flattered at the suggestion that my presence at the delivery of these babies was statistically more effective than that of my very competent neonatal colleagues. However, since I did not resuscitate the infants I do not feel I can accept this accolade.

long arms.

Department of Pediatrics, University of California, Los Angeles, U.C.L.A. Hospital and Clinics, Los Angeles, California 90024, U.S.A.

***These letters have been shown follows.-ED.L.

DALE L. PHELPS

to

Dr

Morley,

whose

reply

would like

Department of Paediatrics, University of Cambridge Clinical School, Addenbrooke’s Hospital, Cambridge CB2 2QQ

COLIN J. MORLEY

Leversha et al. mention cases where banding had shown fra(X)-like chromosomes which were actually no. 6 and state that "mental retardation might have been interpreted as X-linked". 1. Scott S, Richards M. 1028. 2.

LAMBSWOOL IS SAFER FOR BABIES

SIR,—DR Tyler and colleagues (24 Jan., p. 211) describe a preterm infant nursed

on an artificial "sheepskin" mat who experienced respiratory difficulties, apparently due to ingestion of loose fibres. While it is important to be aware of this possibility, we believe that, provided some simple precautions are taken, natural lambswool is a

Nursing low-birthweight babies on lambswool. Lancet 1979,i:

Powley M, Nye P, Buckfield P. Nursing jaundiced babies on lambskin. Lancet 1980; i

979. 3. Yunis JJ, Chandler ME. High resolution chromosome analysis in clinical medicine. Progr Clin Pathol 1977; 7: 267-88. 4. Giraud F, Ayme S, Mattei JF, Mattei MG. Constitutional chromosomal breakage. Hum Genet 1976; 34: 125-36. 5. Quack B, Nantois Y, Mottet J, Noel B. Lacune stereotypée constitutionelle des chromosomes humains J Génét Hum 1978; 26: 55-67. 6. Sutherland GR. Heritable fragile sites on human chromosomes Am J Hum Genet 1979; 31: 125-48. 7. Hecht F, Kaiser-McCaw B. The importance of being a fragile site. AmJ Hum Genet

1979; 31: 223-25