Editor’s Choice

Maternal mortality—the role of the obstetrician

The ultimate failure of obstetric care is death of the baby and the mother, a double tragedy. In late 1817, there began in the UK a ‘triple tragedy’. Princess Charlotte, the only legitimate child of the Prince of Wales (and therefore heir to the British throne), ended her first full-term pregnancy. She went into labour at 7 p.m. on 3 November but she did not deliver until 9 p.m. on 5 November—and her son was stillborn. She herself died the following morning, and her obstetrician, Sir Richard Croft, shot himself 3 months later.

Although maternal death is still unacceptably common in the developing world, in the developed world most midwives, and many obstetricians, will see out their entire professional careers without caring for a woman who dies during, or in connection with, childbirth. Those of us who practise ‘high-risk obstetrics’ are unlikely to be so fortunate. I have personally cared for five women who did not survive, and have witnessed the trauma to the family concerned. Inevitably, after the event, one questions one’s management in detail over a long period, and to do so is a professional, ethical and personal duty. But some are unprepared for the level of public scrutiny that ensues. I have had my practice examined by hospital enquiries, by coroners, by the Ombudsman, by the High Court, and by the National Confidential Enquiry into Maternal Deaths. In all cases, the judgement has been that the deaths occurred despite exemplary care, and with the support of colleagues (who in turn I support), I have survived in better shape than Sir Richard Croft. However, repeated exposure to double-page colour spreads in national newspapers, listing factually incorrect allegations from aggrieved relatives (which one cannot refute because of the rules of confidentiality) is stressful. That media reports are often inaccurate highlights the importance of an objective, nonconfrontational method for ascertaining the true facts in such cases—confidential enquiry. On page 379, Professor James Walker sets out the key attributes, processes and advantages of confidential enquiry. He advocates the wider use of this approach, both in the developed world, but especially, in the developing world where maternal mortality is such a scourge. He highlights some of the pitfalls in obtaining correct data, and discusses some of the lessons learned from the Confidential Enquiry carried out in the Netherlands, reported by Schutte et al. on page 399. Worryingly, Schutte et al. report a rise in maternal mortality since the 1980s, attributed in large part to substandard management of pre-eclampsia (care was below par in 91%), and the challenge of dealing with a rapidly growing immigrant population (care was substandard in 62%). They emphasise the importance of identifying risk early, ideally preconceptionally, but at least in early pregnancy. A particular challenge relates to immigrants who for a variety of reasons often fail to engage with medical care until disease is well established. In the latest confidential enquiry in the UK, one in six of the mothers who died was unable to communicate effectively in English (the rate of poor communication was 10% in the Netherlands). Acknowledging the higher risk in immigrant communities, and making sure that education in terms of language and understanding of health issues is provided effectively, remains a challenge that politicians tend to avoid. They probably wish to escape the many booby traps laid by ‘political correctness’. It is a cruel irony that ‘respect for the cultural traditions of others’ can lead to neglect of their very real needs.

Is it all in your genes?

Gregor Mendel (1822–84) was an Augustinian priest whose now famous experiments with peas suggested a potential physical mechanism for the process of ‘natural selection by survival of the fittest’, described so well by Darwin in his Origin of species, published 150 years ago. As with so many scientific discoveries, there was an element of serendipity in Mendel’s choice to study peapod characteristics because they are inherited in a relatively simple form, making it easier to understand the underlying principles. The publication by Crick and Watson in April 1953 of the molecular structure of DNA led to the identification of the physical basis of ‘genes’. Because of the widespread belief in the ‘superiority of mankind’ over all other animals and plants, the conclusion of the human genome project that man only had about 23 000 separate genes came as a shock (it is only 50% more than the simple roundworm, and about half the number in rice). Moreover, the number of functional genes in a species is not closely linked to the total amount of DNA per cell (the C-value, the origin of which term is unclear). Many protozoans have a much higher cellular content of DNA than far more complex organisms such as humans. Initially, the part of the DNA that had no obvious function was termed ‘junk DNA’ and in the human it accounts for about 95% of DNA by weight. Gradually, it has become clear that much evolved complexity derives from the modifying effects of the DNA in the so-called ‘junk regions’ upon the function of the actual genes. Moreover, it has recently been established that genes do not exist only as single copies, but can be found multiple times in the genome, often in nontranscriptional regions (where they do not code directly via RNA into proteins). This ‘copy number variation’ is being recognised as another source of complexity in the expression of a specific gene (or group of genes). Moreover, spontaneous changes in copy number occur 1000 times more frequently than changes in the nucleotides that make up the genes (single nucleotide polymorphisms, SNPs, often known as ‘SNIPS’). On page 391, Choy et al. at the Chinese University of Hong Kong and the Harvard Medical School in the USA explain the importance of this rapidly developing area of study. Copy number variation has been linked to major birth defects, generalised learning disabilities and adult cancer. Anyone involved in fetal medicine and antenatal diagnosis should read this paper to keep themselves up to date in a rapidly expanding area of knowledge.

Vaginal birth after three caesareans is not a joke

In January, we had a systematic review of trials of vaginal birth following two caesarean sections. Just in case anyone thought two were the limit, we thought we should follow this with a paper looking at trials of vaginal birth following three previous caesarean sections. On page 422, Cahill et al. from Washington University in St Louis report on the outcome of 89 such women, from a cohort of 25 005 women with previous caesareans. They found their outcomes to be similar to 771 women with three previous caesareans who elected for a repeat operative delivery, and even more surprisingly, to those having only one previous caesarean. Of course, this was not a randomised trial, and the women electing for a trial of vaginal delivery were likely to have been a highly selected group (selected both by their own choice, and the advice of their obstetrician). Moreover, with such small numbers, the reliability of risk estimates cannot be high. Nonetheless, on the basis of these results, a choice for a trial of vaginal delivery after three previous caesareans cannot immediately be dismissed as irrational. As this is such a controversial topic, why not put this paper up for discussion at your local ‘journal club’? To help you to structure this exercise, editor Dimitrios Siassakos has prepared sample discussion points which are at the end of the paper.

Total or subtotal hysterectomy?

As I have mentioned before in my Editor’s choice, I was trained always to remove the cervix at hysterectomy when this was technically feasible. Although this often added to the difficulty of the procedure, it was ‘a given’ that it was irresponsible to leave the cervix because it could undergo subsequent malignant change. However, if the cervix is cytologically normal and negative for human papillomavirus, there is no longer such a strong case for automatically removing the cervix, and some have claimed that leaving it maintains subsequent vaginal lubrication and sexual function. In some parts of the world, subtotal hysterectomy has always been popular, for example in Scandinavia. On pages 469 and 479, we publish the results of a randomised clinical trial of total versus subtotal hysterectomy by Persson et al. from Sweden. As expected, the operation itself was significantly shorter for the subtotal procedure, but there were no obvious clinical benefits to the women either in the immediate postoperative weeks or at the 1-year follow-up. Moreover, 20% of women in the subtotal group reported bloodstained vaginal discharge or monthly bleeding after 1 year, compared with only 1% in the total hysterectomy group (this one instance was the result of a granulation polyp in the vaginal vault). These results provide useful information to share with women when obtaining informed consent for hysterectomy.

More information to inform consent for tubal ligation

Although vasectomy is quicker, safer and more effective than fallopian tube occlusion, many women who have had complications in childbirth make a personal decision that they would not wish to bear any more children, but would prefer their partner to retain the ability to father children in case of unforeseen future events such as loss of their existing family. It is important that we have accurate information to counsel such women both on the physical and psychosexual outcomes of occlusion. Evidence regarding any changes in sexual response following tubal ligation has been sparse, and so we are pleased to publish on page 463, a reassuring paper by Smith et al. from Australia. Using a telephone survey of 447 women with a previous tubal ligation, and comparing them with 2274 controls, they found no adverse effects, and possibly some benefits, for the sexual lives of women undergoing tubal ligation.

Screening for diabetes in pregnancy

It is always nice to have one’s prejudices confirmed. I have advocated since early in my training the use of an initial ‘O’Sullivan screen’ (50 g oral glucose followed by a 1-hour blood sugar estimation) at 26–28 weeks of gestation, as an alternative to screening for diabetes using a full glucose tolerance test (GTT) given to women on the basis of historical risk factors. The latter approach misses many cases and is expensive, not only because of the number of blood sugar measurements required, but also in terms of the women’s time and inconvenience (e.g. fasting, multiple venepunctures). When the glucose is given as a proprietary sugar drink 1 hour before the routine end of second trimester blood tests are taken, disruption is kept to a minimum, and only women screening positive need have a full GTT. On page 407, Meltzer et al. from McGill University in Canada report a prospective randomised controlled trial comparing the use of an initial screen with the immediate use of a full GTT. Importantly, they included costs to the women in their calculations. Two-stage screening was 20% cheaper when this element was factored in. I am pleased that in my unit we can continue our time-hallowed policy not just on the grounds of convenience, but also of cost-effectiveness!

And finally…

Do not forget our 2011 theme issue on ‘infections in pregnancy’, with a closing date for submissions of 31 May 2010. We already have submissions but do keep them rolling in—if accepted, your paper will be in prestigious company!