Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
On the fallibility of experts
- Edmund Hey (5 October 2005)
-
Re: On the fallibility of experts
- Philip J Steer (7 October 2005)
-
Expert interpretation of the data?
- David JR Hutchon (9 October 2005)
|
|
|||
|
Edmund Hey, Retired Paediatrician Newcastle upon Tyne
Send response to journal:
|
Editor – Professor Steer tells us, in his recent editorial, that it “may be harmful in the long term” to give two IM injections of betamethasone to a mother facing the prospect of delivery before 39 weeks gestation.(1) It may also be harmful to listen to the views of a senior and respected expert, and better for readers to evaluate the evidence for themselves. This is not the place for a full rebuttal of all the concerns he raises. However, to take but one example, the Professor waves before us the fear that children so exposed may be at later risk of ‘hypertension’, quoting one non-random study suggesting that systolic blood pressure in those so treated may, on average, be 4 mmHg higher 14 years later.(2) However, he ignores a second study of children followed for 20 years after inclusion in a randomized trial that found the opposite of this,(3) and also fails to mention that treatment had no effect on blood pressure 30 years later in 455 of the children from the original trial on which he is commenting (119/74 v. 118/74 mm Hg).(4) A similarly misleading editorial in your journal 23 years ago(5) put back any general recognition of the potential value of such treatment for more than ten years, as the forthcoming Witness Seminar that I chaired for the Wellcome Trust Centre for the History of Medicine last year will document when published in December.(6) Of course there are risks of doing something. There are also risks associated with doing nothing. Life is risky business, and we can’t always wait for 30 years to know everything we would like to know about the balance of risk and benefit. However, thanks to the fact that Mont Liggins recruited all the women offered betamethasone in Auckland into a randomized controlled trial from the outset, we now know more about the long term effect of this intervention than we do of almost any other intervention introduced into obstetric practice in the last half century. Professor Steer says that, because a single course of betamethasone or dexamethasone “reduces neonatal mortality” (it actually halved mortality in the 2000 babies entered into such studies before 1980), “this perhaps justifies the small risk of long term side effects”. He may wish to go on waiting to see if long term problems develop, but I think I know what most women would decide when given a more balanced assessment of the known, and the still unknown, risks and benefits. 1. Steer P. Giving steroids before elective caesarean section. Neonatal respiratory morbidity is halved, but they may be harmful in the long term. BMJ 2005;331:645–6. 2. Dessens AB, Haas HS, Koppe JG. Twenty-year follow-up of antenatal corticosteroid treatment. Pediatrics 2000;105:e77. 3. Doyle LW, Ford GW, Davis NM, Callanan C. Antenatal corticosteroid therapy and blood pressure at 14 years of age in preterm children. Clin Sci 2000;98:137–42. 4. Danziel SR, Walker NK, Parag V, et al. Cardiovascular risk factors after antenatal exposure to betamethacone: 30-year follow-up of a randomized controlled trial. Lancet 2005;365:1856–62. 5. Roberton NRC. Advances in respiratory distress syndrome. BMJ 1982;284:917–8. (See also 1557-8.) 6. Reynolds LA, Tansley EM (eds). Prenatal corticosteroids for reducing morbidity and mortality in preterm birth. Wellcome Witnesses to Twentieth Century Medicine, volume 25. London: The Wellcome Trust Centre for the History of Medicine at UCL; 2005. (See www.ucl.ac.uk/histmed). Competing interests: None declared |
|||
|
|
|||
|
Philip J Steer, Professor Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH
Send response to journal:
|
Editor - I agree with Dr Hey that it is better for readers to evaluate the evidence for themselves, but sometimes they need to be provoked into doing so. He seems to have overinterpreted my editorial in thinking that I was arguing against the use of antenatal steroids to provoke pulmonary surfactant release in the fetus before 34 weeks when preterm delivery seemed likely. In fact, I was arguing against their use in later pregnancy when delaying delivery was a practical alternative. In my original draft I wrote "Antenatal administration of steroids reduces neonatal mortality in babies born before 34 weeks and this justifies the risk of long term side-effects", thus expressing my view that on current evidence such use is justified. However, I was persuaded by my editor at the BMJ that I was being too dogmatic and therefore agreed to her suggestion that the word 'perhaps' be inserted before justifies, so as to indicate proper scientific uncertainty. Whenever we administer powerful drugs to pregnant women, it behoves us always to be alert for possible long term complications in the offspring, and sometimes such complications lead to a re-evaluation of the risk-benefit ratio. It is possible that with the introduction of surfactant administered to the newborn, and the development of improved techniques of neonatal ventilation, neonatal steroid administration is not as critical as it once was in determining neonatal mortality. Moreover, it should be remembered that the diagnosis of preterm labour is far from an exact science, and many mothers in threatened preterm labour receive steroids when in fact they go on to deliver at term. Their babies are thus exposed to the potential long term effects of steroids with little possibility of benefit. Competing interests: None declared |
|||
|
|
|||
|
David JR Hutchon, Locum Obstetrician Grey Base Hospital, Greymouth, New Zealand
Send response to journal:
|
I agree with Professor Steer that we need to continue to exert caution before administering steroids for elective Caesarean Section. Are we not all experts in some way and we can all be wrong? This is why debate in these sections of the eBMJ is so valuable. If this is not the place for debate I am not sure where is. The internet versions of the two papers which are the subject of Professor Steer's editorial provide more data and I believe need to be debated in much more detail. First of all, the long term study from Auckland (1) is a great achievement although limited in value as only 20% of the original participants were involved. From the flow diagram published on the eBMJ we can see that there were 472 survivors out of a total of 560 in the treatment group and 474/582 in the placebo group. This is an odds ratio for survival in the treatment group of 1.07 (95% confidence interval 0.8 – 1.4). There was therefore no significantly improved survival in the treatment group. Again the eBMJ shows that 46 out of the 87 treated survivors reached secondary education while there were 63 out of 105 controls. Treatment did not improve the educational opportunities, the odds ratio of 0.75 (95% confidence interval 0.4 – 1.3). A reasonable evidence based conclusion would be that steroids did neither harm nor benefit. The second paper on the effect of steroids before elective caesarean section (2) also requires the e-version to determine the full story. In his editorial Professor Steer points out that admission to special care for the neonate increases parental anxiety and is costly for the NHS. So do we assume that giving steroids before elective caesarean section reduces admission to special care? Steroids did reduce the risk of being admitted to special care with a diagnosis of respiratory distress. However in the e-version we can identify the fact that a total of 26 neonates out of 467 babies were admitted in the treatment group for any reason and there were 32 admitted out of 475 in the controls. This results in a relative risk for neonatal admission following treatment of 0.82 (95% confidence intervals 0.5 – 1.37). There was therefore NO significant reduction in parental anxiety or cost to the NHS as a result of giving steroids. Data about the time in special care baby unit in days is available in table 2 of the paper version and confirms no significant difference between the two groups. Careful examination of these two papers, together with Professor Steer’s editorial, should form the basis of continued debate by any obstetrician or paediatrician with an interest in scientific truth. References 1. Dalziel S R, Lim V K, Lambert A, McCarthy D, Parag V, Rodgers A, Harding J E. Antenatal exposure to betamethasone: psychological functioning and health related quality of life 31 years after inclusion in randomised controlled trial. BMJ 2005;331:665-8 2. Stutchfield P, Whitaker R, Russel I Antenatal betamethasone and incidence of neonatal respiratory distress after elective caesarean section: pragmatic randomised trial BMJ 2005;331:662 Competing interests: None declared |
|||