Evaluation of Different Approaches to Obstetric Care: Parts I and II.

Sir, We are grateful to Howie et al (19776) for giving us an opportunity to correct a misapprehension which they may share with others. The statements on pages 926 and 932 of our study indicate that we analyzed our perinatal deaths by clinical cause, but found no significant differences between the two obstetric teams except in respect of rhesus haemolytic disease. The perinatal mortality rates experienced by initially normal patients (Part 11) were derived using small numerators, and the slightly higher rate among Team B’s patients is explained by a chance excess of deaths associated with congenital abnormality. We have shown elsewhere (Chalmers et al, 1977) the extent of bias likely to have been operating in the study by McNay et al (1977). In addition, we would like to comment on the suggestion by the Glasgow group that, just as the death rate from leukaemia is the appropriate measure of the effect of anti-leukaemic therapy, so also the ‘mature, unknown’ perinatal death rate is the key factor in any assessment of the value of induction of labour. While it is true that these measures constitute an essential element in such evaluations, the inescapable fact remains that both anti-leukaemic therapy (Malpas, 1977) and induction of labour (Pinkerton et al, 1975; Meagher, 1976; Hack et al, 1976) can lead to death from other causes. We have not found evidence of such adverse effects in our own studies, and apparently none of the 12569 inductions reviewed by McNay et a1 (1977) were implicated in any of the perinatal deaths they studied. However, we remain firmly of the opinion that those who advocate an extended use of induced labour should demonstrate that the advantages of such an approach outweigh the disadvantages. This will only be possible through adequately designed, prospective research (such as that currently used to evaluate leukaemia therapies), and we are pleased to note that the Glasgow team have elsewhere expressed their agreement with this view (Howie et al, 1 9 7 7 ~ ) . Yours, etc,