Editor’s Choice

More on mesh for prolapse

In October last year, I expressed my concern about the vigorous promotion of mesh for vaginal prolapse while data on long-term safety were lacking. We start 2009 with support for my concern from the chairman of the research subcommittee of the British Society of Urogynaecology. On page 1, Editor Doug Tincello criticises the mesh manufacturers for not encouraging a sufficient number of prospective randomised trials, especially in relation to long-term outcome. He highlights the paradox that the European clinical trials directive, designed to protect the safety of patients entered into clinical trials, has placed such stringent quality requirements on researchers that it now acts as a major disincentive to the initiation of prospective studies. He suggests that urogynaecologists should refuse to use new meshes until appropriate studies have been carried out, arguing that this would force manufacturers to provide funding for necessary long-term evaluation. Supporting his criticism is the systematic review by Feiner et al. on page 15 of the efficacy and safety of transvaginal mesh kits in the treatment of prolapse. They identified 30 studies involving more than 2500 women, but only 3 of these studies were randomised controlled trials. Complication rates were almost one in five, with mesh erosion and dyspareunia being the most common problems. They comment that ‘Surgeons should counsel patients that device-related complications … are not rare, most are related to the use of mesh, and their management might necessitate surgical intervention under anaesthetic’.

Tincello has also contributed to the commentary by Doshani et al. on page 3, which highlights the growing contribution of qualitative research to the evaluation of surgical treatment for prolapse. A key conclusion is that although objective outcomes are favoured scientifically because they are easier to measure, it is the subjective experience of the patient that is more relevant to them as an individual. Average changes in quantitative outcome can hide important differences between subgroups of patients, which become apparent when qualitative studies are undertaken. Such studies can address the question of whether a service is being delivered in the most effective manner and can also suggest new approaches to therapy. Qualitative research is particularly useful for hypothesis generation. It can also lead to the construction of symptom scoring systems, and one such score (the pelvic organ prolapse symptom score) is evaluated in the article by Hagen et al. on page 25. The methodology of such scoring systems has developed in recent years and includes assessing features such as internal consistency between different items designed to measure the same aspect of disability (measured using Cronbach’s alpha, calculated from the pairwise correlations between items and ranging between 0 and 1), construct validity (an assessment of whether different aspects of the score theoretically linked to the same hypothesis actually correlate) and sensitivity to change (whether the measure changes significantly following an intervention). The use of the score in three separate studies was examined to assess its performance. The authors found that it performed well, and reading the article provides a good introduction to the assessment of scoring systems.

Auditing maternity care in developing countries

In the developed countries, clinical audit is now well established. In developing and resource-poor countries, there are several major obstacles to effective clinical audit of even core measures such as maternal mortality. These are discussed in the invited commentary by Kongnyuy and van den Broek on page 7. A key problem is how to obtain reliable data on mortality in non-hospital settings. On page 82, Ronsmans et al. have piloted an innovative approach, which relies on assessing the incidence of severe morbidity in hospitals, which would have resulted in death in the absence of hospital facilities. Adding the incidence of these ‘near-miss’ events to the actual number of maternal deaths results in the predicted incidence of deaths in the population having no access to hospitals (usually the rural poor). Applying this approach in two districts of West Java gave figures for maternal mortality for the populations, which appear plausible by historical and geographical comparison. Another important obstacle to audit is that staff can feel threatened by attempts to assess why adverse outcomes occur. On page 38, Richard et al. describe an audit of a 26-bed obstetric unit in a district hospital in Burkina Faso. The problems that they identified will be familiar to anyone who has carried out an audit in a maternity unit anywhere, namely poor availability and maintenance of equipment, guidelines being disregarded, poor organisation of ward rounds and failures of communication. However, they went deeper than just identifying the deficiencies in practice and assessed the attitudes of staff to the audit process. Commonly, it was seen as senior staff criticising junior staff, thereby undermining their confidence. Moreover, 40% had difficulty in grasping the concept of self-criticism. They also felt that audit too often only identified weaknesses and failed to recognize or promote good care. Fear of blame encouraged defensive record keeping, and some staff found the process humiliating. This made it difficult for them to acknowledge mistakes in public. While audit clearly has the potential markedly to improve the standard of care, if ways are not found to deal with such psychological issues sensitively, then implementation of audit will not bring about the expected benefits. There are many reports of lack of clear benefits from introducing guidelines, and for the audit cycle to be closed requires strong leadership and avoidance of the ‘blame culture’. In my own experience of clinical case review meetings and audit, it is best to present cases either with the staff involved anonymised or where that is impossible (e.g. because everyone knows who was involved) to counsel staff beforehand to help them cope with acknowledging their own errors. Such acknowledgement usually engenders sympathy rather than the criticism generated if people are perceived to be trying to evade responsibility. I find it helps if senior staff are not only sympathetic but recount similar stories from their own experience along the lines of ‘we are all human and make mistakes and the important thing is to learn from them’. Oscar Wilde once wrote in ‘The Picture of Dorian Gray’ that ‘experience is merely the name men give to their mistakes’. We have to learn to live with the guilt that comes from acknowledging our mistakes, and during this learning process, the understanding and support of colleagues are needed.

Pregnancy and suicide—the depressing statistics

Recent improvements in ascertaining pregnancy-related deaths in the UK for the purposes of the confidential enquiry have highlighted suicide as a major cause of maternal mortality. Similarly, in recent years, the family morbidity related to pregnancy-related depression has become better recognized. However, although validated screening tools for depression exist, they currently seem to have little impact on improving outcome. This problem is examined by Mitchell and Coyne in their commentary on page 11. They point out that many professionals fail to administer screening because they do not feel competent to deal with depression once it is recognised and are unsure how to obtain further help for the women affected. They also suggest that in a variation of ‘Catch 22’, many women identified as being either at risk or actually depressed are unwilling to admit their condition or cooperate with offers of help. It seems likely that a change in society’s attitudes to depression and a willingness to direct more resources to this important area will be necessary before the screening tools can become truly effective.

Uterine artery embolisation for intractable postpartum haemorrhage

In recent years, BJOG has published a number of articles reporting developments in the control of postpartum haemorrhage, particularly in relation to surgical techniques and uterine compression sutures. We like to publish these because most of our readers will be able to implement them. However, we also need to recognize the role of other specialists, for example interventional radiologists. Uterine artery embolisation for postpartum haemorrhage has had a mixed press, with some small series reporting a high success rate, whereas others find less benefit. It requires advanced technology and equally advanced practitioners of interventional radiology to be effective. Even when the facilities and practitioners exist, they are not always available around the clock. Because experience with the technique remains relatively limited, we thought it important to publish the serious complications described by Maassen et al. on page 55. They report a series of 11 cases, following which one woman developed a vesicovaginal fistula and another required major surgery to rescue her leg from the consequences of avascular necrosis. Embolisation should probably remain a technique of last resort when simpler techniques fail and life-threatening haemorrhage continues.

Improving the diagnosis of preterm labour

Many women who present with preterm contractions, appearing to be in preterm labour, subsequently deliver at term. In such cases, detecting the presence of fibronectin in vaginal fluid has proved to be a sensitive and specific marker of progressive labour. But what of women who present not with contractions but with ‘losing water down below’? Currently, speculum examination following recumbency for several hours, to see if there is a pool of fluid in the posterior fornix, is the most reliable method of assessment—but there are still many false positives and negatives. Assessment of the level of lactate in the vaginal fluid may clarify the prognosis in such cases. Wiberg-Itzel et al. on page 62 report a strong inverse correlation of the lactate level with the number of hours to the onset of labour. A level of 4.5 mmol/l provided a suitable cutoff level with an 87% accuracy in predicting the onset of labour within 48 hours. However, their study was relatively small (only 81 women), and their promising results will need to be confirmed by others.

Pre-eclampsia again

Until reliable tests in early pregnancy that predict the later onset of pre-eclampsia are available, risk assessment will continue to require careful history taking. This month we publish the articles refining the risks associated with physical activity in early pregnancy and the effects of recurrent miscarriage and infertility. We know that exercise generally has beneficial effects on the cardiovascular system, and thus one’s prior hypothesis would probably be that it would protect against pre-eclampsia. In fact, the study by Østerdal et al. reported on page 98 failed to show any protection from exercise, and high levels of physical activity actually increased the risk of severe pre-eclampsia (although the effect was modest, with odds ratios of approximately 1.7 and absolute risks of less than 1.3%). On page 108, Trogstad et al. report that women who had ever been treated for infertility were significantly more likely to develop pre-eclampsia, especially if they had had three or more consecutive miscarriages (odds ratio 2.4, 95% confidence intervals 1.11–5.18). Pre-eclampsia remains a mystery wrapped in an enigma.

Nifedipine for external cephalic version?

With the demise of beta-sympathomimetics as tocolytics, nifedipine has been touted as an unlicensed but cheap alternative for external cephalic version. On page 74, Tan and Collaris found no significant difference in the success of version or in caesarean section rates between nifedipine and terbutaline. However, success rates were lower with nifedipine, and as pointed out in a Mini Commentary by Editor Amarnath Bhide, the numbers studied were insufficient for a proper noninferiority study. If you are not clear about the differences between superiority, equivalence, and noninferiority trials, you should read his very lucid account.