C reactive protein and neutrophil band forms in neonates.

P. Etches,N. Finer
DOI: https://doi.org/10.1136/adc.68.5_Spec_No.619-b
1993-05-01
Archives of Disease in Childhood
Abstract:EDITOR,-Professor McClure acknowledges that surfactant treatment benefits most of its recipients but rightly questions whether it should be given to the smallest, sickest babies. I Meta-analyses of more than 30 randomised controlled trials enrolling over 6000 preterm babies have shown a consistent reduction in neonatal mortality of about 40% and of pneumothorax of between 40% and 70%.2 Many of these trials excluded babies weighing less than 700 or 750 g at birth and some excluded babies with congenital anomalies, low Apgar scores, and severe grades of intraventricular haemorrhage. We agree with Dr Morley who indicates in his commentary that it is only by including extremely small or immature babies in trials that we can learn how (or if) their outcome can be improved. However we would also stress the need at randomisation to document their initial severity of disease, including the degree of hypothermia and asphyxia. Four trials have looked at the effects of surfactant therapy for babies of <750 g or <27 weeks' gestation; three used prophylactic surfactant-5 and one surfactant treatment of established respiratory distress syndrome.6 The results are summarised in the table. Prophylactic surfactant reduces the odds of neonatal mortality by about 40%, which is in keeping with the meta-analysis from more mature babies.2 The risk of intraventricular haemorrhage is not significantly increased, but there is a trend that needs to be studied further. When Survanta is used to treat respiratory distress syndrome in babies of 600-700 g there is no clear reduction in neonatal mortality,6 which is in keeping with the findings of Kendig et al,7 using calf lung surfactant extract, who showed that in babies <26 weeks' gestation survival was better for prophylaxis (64/85 (75%)) than for rescue treatment (39/72 (54%); p<0Q01). As Dr Morley points out in his reply the data are now beginning to support the use of surfactant, preferably by prophylaxis in even the most immature of babies. Some clinicians may accept this evidence and continue to treat these babies while others might call for further and larger randomised controlled
What problem does this paper attempt to address?