Editor’s Choice

Maternal haemorrhage at birth

Haemorrhage remains one of the most important direct causes of maternal mortality worldwide, accounting for more than one in ten of all maternal deaths. The World Health Organisation has estimated that there could be as many as 140 000 preventable deaths from pregnancy related bleeding each year. The last decade has seen significant advances in the management of postpartum haemorrhage, for example, the use of misoprostol, compression sutures and intrauterine balloons inserted into the uterus to produce a tamponade. Information about balloons has appeared in the literature in a somewhat haphazard fashion and we are therefore pleased to publish a review by Christos Georgiou (page 748) of their introduction and subsequent use. A variety of types have been described, ranging from readily available and inexpensive condom-catheters and Foley catheters, through balloons previously used for other purposes (e.g. the Sengstaken–Blakemore tube used to control bleeding from oesophageal varices and the Rusch urological balloon) to the Bakri balloon that has been specifically designed for obstetric use. The aggregated literature reported by Georgiou records 98 cases where balloons are deemed to have worked and only nine cases where they did not (in four of these placement of the balloon was unsuccessful). These figures need to be interpreted with caution because of potential reporting bias (cases where the balloons do not work may be less likely to be reported). However, given that adverse effects are few and that the effectiveness of the technique often appears striking, it may be that we will never see a prospective randomised trial (a comment often made is that there are no randomised controlled trials of parachutes either).

In the developed countries, an increasingly recognised cause of major postpartum haemorrhage is placenta accreta. This problem is likely to become more prevalent because of its relationship with previous caesarean section scars. Affected women commonly need massive blood transfusion, and the last few years has seen the growing use of intra-operative cell salvage (IOCS), reviewed by Geoghegan et al. (page 743). The importance of this technique has increased as the number of available blood donors falls, because of restrictions on donation (for example, because of the risk of variant Creutzfeldt-Jakob disease). Cell salvage allows some of the blood lost during a surgical procedure to be returned to the patient. A machine filters, washes and centrifuges aspirated blood, allowing the isolated red cell component to be reinfused. Unlike intrauterine balloons, there has been a randomised trial, although this was small, including only 68 patients. Despite the paucity of evidence, the use of cell salvage has been recommended by many major organisations, including the American Society of Anesthesiologists in the USA and the National Institute for Health and Clinical Excellence in the UK. Geoghegan et al. argue that although there are theoretical benefits, there are also potential harms and call for a large randomised controlled trial (which would need to include approximately 4500 caesarean sections) to assess their effectiveness and safety more reliably.

Vitamin supplementation and pre-eclampsia––does the background state of nutrition matter?

At the beginning of the decade, a positive pilot study of the use of antioxidant vitamins for the prevention of pre-eclampsia led to the exciting possibility that routine supplementation might be of value. Unfortunately, three large double blinded randomised controlled trials have failed to confirm this original promise. However, because most of the women entered into these trials were in developed countries and therefore relatively well nourished, the trials did not rule out the possibility that supplementation might be protective in vitamin deficient women. In this issue of BJOG, we publish a large prospective randomised trial carried out in developing countries. Villar et al. (page 780) report that daily supplementation with 1000 mg of vitamin C and 400 iu vitamin E had no significant effect on the incidence of pregnancy hypertension or indeed on any other maternal outcomes. The high rate of pre-eclampsia observed (24%) indicates that this was indeed a high-risk population and the study was accordingly adequately powered to rule out a clinically meaningful effect. The failure of all large trials to date to find antioxidants of benefit in preventing pre-eclampsia suggests that either supplementation cannot be given early enough in pregnancy to influence placentation or that free oxygen radicals are involved in the pathophysiology of pre-eclampsia but are not causal.

Viral infections and the fetus

Although humans combat many viral infections by developing antibodies, which protect future pregnancies from damage, if the initial infection occurs during pregnancy, the fetus can be damaged before the mother develops her protective response. Probably the best known teratogenic virus is rubella, because there is a vaccine for it, but in fact cytomegalovirus (CMV) is the most frequent cause of congenital viral infections. It affects up to 1% of all live births and cytomegalovirus is the leading infectious cause of sensorineural deafness and mental retardation. We publish a study by Picone et al. (page 818), which looked at the incidence of primary infection over a two-year period in a French hospital. Just over half (53.2%) of 3665 women tested were CMV-specific IgG negative at initial screen and were thus susceptible to infection. Despite being given detailed hygiene information, nine women (0.46%) had a primary infection between screening and 12 weeks of gestation and a further five (0.26%) seroconverted between 12 and 36 weeks of gestation. There were two pregnancy losses, unrelated to viral infection. Three babies showed evidence of having been infected, but only one of these had clinical signs (petechiae and unilateral hearing loss at 1 year of age). The main risk factor for infection was contact with young children (10 out of 14) and five of the mothers infected were nurses or doctors. The authors point out that the rate of seroconversion they observed was only about a quarter of that predicted from previous studies of similar populations and attribute this to hygiene education, suggesting that this might be of particular value in high risk women (sero-negative women with small children).

Parvovirus B19 is another widespread human pathogen and affects the fetus in about a third of infections occurring during pregnancy. Because it infects fetal erythroid precursors cells, it causes fetal anaemia and hydrops (which can be fatal) in up to 12% of affected pregnancies. Fetal anaemia can be diagnosed effectively using ultrasound examination, for example, Doppler assessment of the middle cerebral artery. However, a specific diagnosis of B19 associated nonimmune hydrops requires laboratory tests of antibodies and/or detection of the virus in the fetus. Sometimes, diagnosis using antibodies is hampered by the immaturity of the fetal immune system, which does not always show a response. Bonvicini et al. (page 813) assess the value of virological assays in diagnosis. Just under a third of samples proved positive for B19 DNA. Interestingly, positivity rates were higher in amniotic fluid cells than in fetal cord blood samples, and as obtaining amniotic fluid is much less risky for the fetus than fetal blood sampling, they recommend amniocentesis as the primary test. Because of the efficiency of ultrasound, in most cases, cordocentesis is no longer necessary to assess fetal anaemia.

Cervical cancer

Screening for cervical cancer is widely agreed to be a major public health success in developed countries. However, in developing countries, cervical cancer remains a major cause of early death. Screening using cervical cytology is often not feasible because of an absence of cytology laboratories and trained cytologists. An alternative strategy is visual screening with acetic acid and Lugol’s iodine. Muwonge et al. (page 829) report the introduction of colposcopy and biopsy in five centres in sub-Saharan Africa. They found in an audit of 29 033 women participating in their study that 98.3% of women had a satisfactory colposcopy as judged by the proportion of women having an adequate punch biopsy and the diagnosis of a precancerous lesion, and that this approach could overcome many of the difficulties relating to cervical cancer screening in resource-poor countries.

Despite the success of screening in developed countries, treatment of cervical intraepithelial neoplasia (CIN) is not always curative. Jakobsson et al. (page 838) report that the overall mortality of a cohort of Finish women treated for CIN was 17% higher than average. While not unexpectedly the standardised mortality ratio for cervical cancer was 7.69, more surprisingly, the standardised mortality ratio for deaths related to injury was 1.31 and for medical conditions and other diseases 1.17. Having a baby post-treatment reduced the risk of death (standardised mortality ratio 0.78) except for those women who had a subsequent preterm delivery (standardised mortality ratio 2.5). It is tempting to hypothesise that the removal of larger proportions of cervical tissue is linked both to more extensive disease and to subsequent cervical dysfunction. This hypothesis is supported by the finding that the standardised mortality ratio was increased by excisional procedures, but not by ablation. The most important caveat in relation to this study is that the authors were unable to correct for many important confounders, such as smoking, alcohol consumption and socio-economic status. Thus, it remains possible that their findings reflect the effects of high risk behaviour and social disadvantage rather than the effect of the disease itself or its treatment. Arguably, treating women for CIN should be used as an opportunity for evaluating the woman’s lifestyle and providing appropriate health education.

We sometimes forget that even in developed countries, the implementation of routine population screening for CIN has been relatively recent. A national cervical screening programme in the UK was not introduced until 1988. Prior to that date, screening had been largely opportunistic and quality control was poor. Herbert et al. (page 854) report their audit of data relating to invasive cervical cancer carried out in the Southampton and South West Hampshire region of the UK. Of the 382 cases reviewed, two-thirds presented because of symptoms and only one-third were detected by a screening programme. However, there was a gratifying fall in the incidence of symptomatic cancers from 13 per 100 000 in the years before screening to 6.4 per 100 000, six to eight years following the introduction of screening. Although in women over 65 years of age, 92% of cancers were detected from symptoms rather than screening, by 1994–96, 75% of cancers in women aged 20 to 34 were being detected by screening. This should eventually lead to a fall in the incidence of invasive cancer in older women. The authors suggest that the high proportion of early stage screen detected cancers in younger age groups accounts for their better prognosis. They also suggest that early detection in young women is so important that it represents an argument against the current policy in England of not offering first screening until 25 years of age. This argument has received a big boost by the wide media coverage in the UK given to the terminal illness from cervical cancer of Jade Goody, aged only 27. Ms Goody first appeared on television in 2002 when she was a contestant on a reality TV show called ‘Big Brother 3’, which led to a lucrative career in similar programmes. Cervical screening starts at age 20 in Scotland, Wales and Northern Ireland and the health minister Ann Keen has now announced a review of English policy.

Even when women take part in screening programmes, some will develop cervical cancer between smears. In the study by Herbert et al. (page 845), 45% had been screened within the preceding 5.5 years. About half of these had had negative cytology. On review, a third of these showed high grade dyskaryosis and a further 17% showed low-grade abnormalities. Only 41% were confirmed as negative. These findings highlight the importance of quality control in any screening programme.