SEARCH

SEARCH BY CITATION

Science and uncertainty

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

Arguably, the scientific method has been the most powerful mechanism for advancing technology over the last few centuries. Although technical development has been going on for millennia (e.g. the use of the wheel for transport, metals for tool-making and concrete for construction), the pace of advance accelerated dramatically with the introduction of the experimental approach. Rather than making deductions from observations made in the course of normal life (‘trial and error’) to reveal the workings of nature, experimental scientists systematically devise and investigate special situations to determine nature’s underlying laws. For example, rather than simply observe objects falling, Galileo Galilei (1564–1642) reportedly dropped equal-sized wood and metal balls from the leaning tower of Pisa, and determined from his measurements that (allowing for air resistance) all objects fall equally fast, and not according to their weight as proposed by Aristotle (384–322 bc). Until the last 50 years, the commonest model of investigation was to record a large number of observations, and from them induce a general rule to predict the outcome (inductivism). But even if a general rule explains all currently known events, this does not ‘prove’ it to be true. Before the discovery of Australia, all observed swans were white, but this did not prove that this was a rule of nature. The first black swan was discovered by Willem de Vlamingh in Western Australia in 1697. The conditionality of knowledge (things are only considered true until someone proves them wrong) was elegantly described by Karl Popper (1902–1994), who suggested that the best way to describe the scientific approach was as a competition between alternative theories. A variety of explanations for natural phenomena are put forward and then experiments are performed to try and prove them false. This approach (a sort of ‘natural selection’ among competing theories) allows apparently unlikely theories (peptic ulcers are caused by bacteria, most intrapartum pyrexias in the developed world are caused by epidurals) to be tested and eventually gain acceptance if they prove resistant to falsification. Some have suggested that modern approaches such as ‘data-mining’ or ‘genomics’ are a return to inductivism, but some would argue that this approach has not proved very productive. I still think that the best research is hypothesis driven.

We have to make clinical decisions based on our best understanding of the current evidence and as doctors give women the confidence that they are receiving the best known treatments. But as scientists, we must always remember that we may be wrong. The enemy of science is dogmatism, or unwillingness to consider points of view that we do not personally favour. Arguably an important role of a medical journal is not just to publish observations and measurements, and the conclusions deduced from them, but also to stimulate debate and encourage the emergence of new (and testable) hypotheses. This is why we prefer to select papers that have novelty—even though the hypotheses they put forward may be proved wrong—and sometimes have to disappoint the authors of worthy but less challenging articles.

An example of a paper that some editors thought was contentious—but that we decided to publish because we thought it raised an important hypothesis that needs to be tested—is that by Pasker-de Jong and colleagues on page 1080. They studied the children of women treated with labetalol or methyldopa for pregnancy-related hypertension between 1983 and 1987, and compared them with women managed by bed rest alone (a popular treatment at that time). In 1991/92, 202 children underwent developmental testing at their homes. Children prenatally exposed to labetalol had a significantly higher risk of attention-deficit hyperactivity disorder compared with the bed rest group (odds ratio 4.1, 95% CI 1.2–13.9). There was a nonsignificant trend to more sleeping problems in the children of mothers who took methyldopa compared with those of mothers prescribed bed rest (odds ratio 4.5, 95% CI 0.9–23.2). Some of our editors thought it unwise to publish a speculative paper based on nonrandomised cohorts, because inappropriate publicity might frighten some women (or their doctors) into avoiding a therapy that can, for example, reduce the incidence of strokes. But we thought it better to expose the hypothesis that such drugs might have long-term effects on the baby to a wider audience, and perhaps stimulate a more exacting scrutiny of the idea. I trust that we can rely on the critical appraisal of our readers. But just in case, we commissioned a minicommentary from editor Michael Belfort (page 1086) to sound an appropriately cautious note.

Mea culpa

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

In June’s Editor’s choice I highlighted the paper by Al-Zirqi et al. (BJOG 2010;117:809–20), which reported on the risks of uterine rupture during vaginal birth after caesarean section (VBAC). I commented on ‘…the 15.9% risk of uterine rupture following induction of labour with prostaglandins/amniotomy. Personally, I could not present a risk of one in six to mothers as ‘a low absolute risk’, as maintained by some commentators.’ Unfortunately, I had made the error of misreading a value per 1000 as a percentage; the correct figures should have been 1.59% and one in 60. A corrigendum is published on page 1172. My error is a salutary warning that we should, wherever possible, go to the primary sources for our information, and not rely on a second-hand report because of the ever-present possibility of a transcription error—from which even editors-in-chief are not immune. The controversial topic of VBAC will be addressed in our correspondence column in the near future, so keep your eyes open for some lively debate.

Endometrial cancer, liquid-based cytology and colposcopy

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

This month we have three articles related to neoplasia of the endometrial glands. On page 1043, Paul Symonds and colleagues discuss the optimal management of endometrial cancer—open or minimal access hysterectomy and bilateral oophorectomy plus or minus lymphadenectomy, this latter being the most contentious aspect. But the role of adjuvant therapy is not without its dilemmas. They put forward plenty of hypotheses that would benefit from being tested by further studies. On page 1051, Amit Patel and colleagues discuss the outcome of 69 women referred with ‘borderline glandular cells’ on liquid-based cytology. They hypothesise that women under 35 years of age with a satisfactory normal colposcopy can be managed conservatively, whereas women above 35 years of age should have a diagnostic large loop excision of the transformation zone. To discuss this controversial proposal, we invited a commentary by Cullimore and Waddell on page 1047. Does the multidisciplinary meeting improve the interpretation of colposcopy findings? On page 1060, Julia Palmer and colleagues suggest that when significant discrepancies exist between colposcopy, cytology and histopathology, discussion can help to avoid overtreatment and on page 1067, Philip Castle and colleagues highlight the need to develop global standards for the interpretation of cytology, irrespective of the system used for collection.

Journal Club papers

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

This month we have three papers on contentious issues accompanied by questions designed to structure discussion at a journal club. The first examines how to expedite the treatment of the partner in cases of sexually transmitted disease. On page 1074 Sharon Cameron and colleagues report their study of giving vouchers for free treatment of male partners for Chlamydia. Although 92% of women stated that they had given the voucher to their partners, and 87% said that their partners were satisfied with such treatment, only 62% of vouchers were actually redeemed. Male duplicity? And the continuing controversy regarding pregnancy outcome after a previous caesarean section is discussed by Marie Carlsson Wallin and her colleagues on page 1088. They conclude that although there is an increased risk of perinatal death and low Apgar score after a previous caesarean section, this occurred mainly when there was a medical indication for the previous operation. When there was no such indication, poor outcomes were not increased and they conclude that the effect is not the result of the operative procedure itself. Finally, is screening for vasa praevia cost effective? Lauren Cipriano and colleagues on page 1108 conclude that it is for twins, but not for singleton pregnancies. Trainee editor Dimitrios Siassakos has facilitated structured discussions on these important topics.

More controversy about alcohol consumption in pregnancy

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

In 2007, a systematic review in BJOG of the effects of low to moderate prenatal alcohol exposure on pregnancy outcome (BJOG;114:243–52) concluded ‘at low–moderate levels of consumption, there were no consistently significant effects of alcohol on any of the outcomes considered’, although this was challenged by Black and colleagues (BJOG:114:778–9). The study of O’Leary and colleagues in 2009 (BJOG;116:390–400) found that low levels of consumption did not increase preterm birth. So where are we now? On page 1139, Monique Robinson and her colleagues report that changes in the Child Behaviour Checklist over the first 14 years of life are improved by light to moderate alcohol consumption during the first 3 months of pregnancy. As this conclusion undermines current government advice in many countries that pregnant women should abstain from alcohol entirely, we commissioned two minicommentaries to discuss the findings. On page 1150, Vincent Jaddoe from the Erasmus Medical Centre in the Netherlands agrees that the results are consistent with other reports, but that ‘more information is needed on specific critical periods during fetal life and threshold levels above which alcohol consumption might have adverse effects.’ On page 1151, Sandra and Joseph Jacobson from the Wayne State University in the USA raise a point I have mentioned before in my editor’s choice, which is that ‘an advisory that low doses of alcohol in pregnancy could be beneficial may encourage drinking by pregnant women with a propensity for alcohol abuse who may not recognize when their drinking becomes excessive.’ The extent to which we should limit choices for the majority to benefit a minority is perhaps more of a political than a scientific issue.

Registration of trials involving interventions on human subjects

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

BJOG subscribes to the principles of the International Committee of Medical Journal Editors (ICMJE)—http://www.icmje.org/. This requires that all randomised trials started before 1 July 2005 should be registered retrospectively in a public clinical trials registry (e.g. the free one at http://www.clinicaltrials.org), whereas all those with more than 100 participants started after this date should have been registered prospectively. Moreover, from the 1 July 2008, in line with World Health Organization recommendations, the ICMJE now requires that ‘any research study that prospectively assigns human participants or groups of humans to one or more health-related interventions to evaluate the effects on health outcomes must have been registered prospectively’ (not just randomised trials). Journals, including BJOG, are already declining publication of nonregistered studies. One reason for requiring registration is to reduce as much as possible ‘negative publication bias’—which skews the literature in favour of studies with a positive outcome and potentially leads to unnecessary or inappropriate treatments being widely taken up. Please highlight the need for registration to all your colleagues and trainees—this could save them much anguish later, when they try to get their papers published.

BJOG—an International Journal of Obstetrics and Gynaecology

  1. Top of page
  2. Science and uncertainty
  3. Mea culpa
  4. Endometrial cancer, liquid-based cytology and colposcopy
  5. Journal Club papers
  6. More controversy about alcohol consumption in pregnancy
  7. Registration of trials involving interventions on human subjects
  8. BJOG—an International Journal of Obstetrics and Gynaecology

In the first 6 months of 2010, more than two-thirds of articles published in BJOG originated from outside the UK—from 22 countries around the globe—which is very appropriate given our international editorial team and readership. Thanks to all our contributors for the excellent papers that you continue to send us.