Editor's choice

The evidence-base for labour induction around term, such as it is, relies almost exclusively on trials which excluded women with a uterine scar. This matters now that around 20% of pregnancies are delivered by Caesarean section. The systematic review from Oregon (pp. 1007–1015) reminds us how little reliable evidence there is to guide decision-making in this area; there have only been two small trials and a small number of observational studies of, at best, moderate quality. These suggest that induction in this situation increases the chance of having another Caesarean, and there are essentially no data on fetal outcomes, so I guess the message is: avoid induction if possible, or better still do a good quality randomised trial. Even under general anaesthesia endometrial ablation is cheaper than hysterectomy, and it would be cheaper still in outpatient settings under local anaesthesia. On pages 1109–1116 we report a trial from Aberdeen comparing post-menstrual phase microwave ablation without endometrial preparation under sedation and local anaesthesia, with conventional microwave ablation using endometrial preparation under general anaesthesia. The latter technique was more acceptable, equally effective and unsurprisingly, cheaper. On pages 1117–1120, a group from the West Midlands report going a step further with a different technique-thermal ablation. They trained general practitioners to do the procedure under local anaesthesia in a community hospital. There was no control group but the procedures seemed to go well. It is hardly surprising that purchasers of gynaecological services are looking to reduce their costs. Professor Gordon Smith from Cambridge has been using Scottish health system data to elucidate obstetric risk factors. A few years ago he demonstrated an excess of delivery-related perinatal deaths in second twins compared with their first-born co-twins. On pages 1139–1144 he extends the analysis to test the hypothesis that twin delivery by planned caesarean section reduces this difference. It does. Of course this does not prove that we should deliver all twins by planned caesarean section; we might, for example, end up increasing prematurity-related deaths from over-diagnosis of false labour. Fortunately a randomised trial, the Twin Birth Study, co-ordinated from Toronto, http://www.crwh.org/programs/tbs.php) is currently recruiting. Some people have long suspected that gastroschisis has an environmental aetiology, and others have wondered whether maternal hair analysis might one day be a useful way to measure peri-conceptional drug exposure. Now a group from Galway and London have put the two ideas together (pp. 1022–1025). Four out of twenty-two women whose fetus had gastroschisis and none out of 25 controls had evidence of recreational drug ingestion at the time of conception. The case is far from proven for all sorts of reasons. Not only are the numbers small, but the drug ingestion may be only a marker for the real cause and there remain many practical difficulties with hair analysis, but we will surely hear more of this. We could prevent obstetric anal sphincter injury by doing more caesareans if we only knew who was going to get such an injury. Unfortunately present risk factors cannot tell us (pp. 1066–1069). Alternatively we could do mediolateral rather than midline episiotomies. Apparently many doctors and midwives angle their incision much nearer to the midline than is generally recommended (pp. 1156–1158). Finally we report a long-term study of obstetric risk factors for prolapse in later life. Way back in 1952 the late Professor James Walker set up a card index of detailed obstetric information for women who delivered in his hospital in Dundee, and it later became possible to link these data to the Scottish Morbidity Record. This allowed his successors to relate pelvic floor repair surgery to the mode of delivery many years earlier. Caesarean section was protective, but nothing else seemed to matter much.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.