This journal has long been interested in evidence-based test evaluation, and the review of tubal patency tests from a collaboration of British and Dutch experts on pp. 1313–1321 is the sort of thing we like to publish. There are many difficulties, not least because the ‘gold standard’ laparoscopy and dye is unlikely to be a perfect test itself, and because it is unethical to measure the reliability of many tests because of the risks of repeating them. Nevertheless the reviewers summarise what data there are in the best way for clinicians, namely likelihood ratios for positive and negative results.
Accidents associated with the surgical use of electrosurgery/diathermy are too common. The paper on pp. 1413–1418 from Australia suggests one preventable reason; many surgical trainees are ignorant about how such instruments should be used. The authors suggest that colleges should include electrosurgery in specialist training programmes, that surgeons should be regularly retested, and that they should practise regularly on simulators. I would only add that the instrument manufacturers themselves also have a self-interest in ensuring that their devices are used correctly and many have commendably already taken a lead in implementing such training.
Human papilloma virus
There may be many advantages from including HPV testing in cervical screening programmes but the harms need to be weighed in the balance too. The authors of the paper on pp. 1437–1443 report the anxiety and negative feelings towards sexual partners caused by women learning that they are HPV positive. It is substantial.
The systematic review on pp. 1333–1340 from Australia updates a 14-year-old one on the same topic. Since then there have been several new studies, including the PEOPLE trial which was larger than all other trials put together. It appears that delayed pushing reduces the risk of difficult instrumental delivery (defined as rotational or mid-cavity) among women with an epidural. The only real reason for doubting this before was concern that misclassification of the type of instrumental delivery might be biasing the results. The present review removes that concern since there was also a clear trend towards fewer overall instrumental deliveries and caesarean sections as well. There were no significant differences in any fetal outcome. Delayed pushing should be added to the list of effective second-stage interventions for women with an epidural.
Eating in labour
Few issues in labour management arouse such strong feelings as whether to feed the mother. Some doctors fear that it may cause gastric aspiration, others that starvation may slow labour, so it is good that researchers are at last evaluating different protocols in randomised trials. The trial on pp. 1382–1387 refutes the hypothesis that oral carbohydrate solution in the late first stage halves instrumental deliveries, but cannot rule out smaller benefits. Although there was no trend towards fewer instrumental delivery there was one towards reduced caesarean section. A previous similar trial had shown a trend towards an increase in caesarean, so the present results should not be used to alter practice, although they may guide future trials.
I hope readers will learn as much as I have from the commentary on pp. 1307–1312. We commissioned it in response to criticism of the two studies we published last year involving renal biopsy in pregnancy for research purposes. The commentary authors' main conclusion is unequivocal – the studies should not have been performed. They also argue that ideally BJOG should not have published the studies, although I think they recognise that views on this latter question may differ.
One of the reasons for insisting that research ethics committees take care with non-therapeutic research proposals is because participants often fail to grasp all the important elements of the projects they are being invited to join. On pp. 1341–1345 Sarah Kenyon and Mary Dixon-Woods report on a survey of patient understanding in the ORACLE trial. At least 14% of respondents did not seem to know what they had joined. This may not matter much for studies like ORACLE where both treatment arms are in common use and experts have judged that the question is undecided. Even participants with poor understanding are not sacrificing themselves or their babies for the benefit of others. Non-therapeutic research is different. Even if we occasionally allow altruistic women to risk their own health for the greater good, we should rarely, if ever, allow them to put their babies at risk.