• Nadine Edwards

The problems with birth-related research


The new and completely revised publication of Birthing your baby: the second stage of labour, details the benefits of physiological birthing for healthy well women and babies.

It might seem surprising, then, that not all the findings from studies into the second stage of labour clearly show this. There are many potential reasons for this. Our ways of thinking can limit how research is designed, what it looks at, what is measured, how it is measured and where it takes place. The medicalisation and mechanisation which increasingly dominate our lives have not only influenced birthing practices, but have also profoundly influenced how research is carried out and reported.

This blog post has been adapted from one of the appendices in Birthing your baby: the second stage of labour, which has been published by the BPPF.

Research is mainly done in hospitals

Most of the research on birth, including the second stage of labour, has been carried out in hospitals with high intervention rates. This means that many women had time restrictions, which might not have taken into account that individual women and their babies were well but labour was just taking more time than “allocated”. Many had inductions or augmentation of labour, epidurals, continuous electronic fetal heart monitoring, were confined to a bed, were lying on their backs and had numbers of vaginal examinations. All of these interventions and practices can disturb a mother’s labour, and some can increase the chance of her having an instrumental birth or a caesarean section.

Almost none of the studies that have looked at the second stage of labour considered the impact of the environment or of high intervention rates on their findings. Midwife and researcher Tricia Anderson (2002) likened the studying of birth in obstetric environments to studying animals in a laboratory rather than in their own surroundings. We know that birth is likely to unfold straightforwardly if the mother is in a homely environment or in her own home with people she trusts, and is unmedicated and able to move around and follow her own instincts throughout labour and birth. Introducing a single change (which is what most of the studies have done) such as extending the length of the pushing phase, or delaying pushing, or encouraging instinctive pushing, or introducing upright positions to women who are labouring and giving birth in a medicalised environment, is unlikely to make much of a difference (Goer 2018).

Lack of detail about research

It is also often unclear how these single interventions have been introduced and used. For example, research on positions for birth does not always specify whether women used these as they were bearing down and/or used these for the actual birth. Additionally, we don’t know whether women were adopting an upright position on the floor, which is often easier and more comfortable, especially if mats, mattresses, cushions or other furnishings are available, or if they are expected to do this on a high, narrow bed, which most women find less than relaxing.

How reliable is the evidence?

Currently many people believe that randomised controlled trials provide the best evidence, and in the UK and many other countries, the Cochrane reviews (which are meta-analyses and systematic reviews of randomised controlled trials), are usually seen as the most reliable source of evidence. But as reviewers have pointed out, some of the evidence on which Cochrane reviews are based is of low or even very low quality. Indeed, when researchers examined the evidence on which maternity care guidance is based in the UK, Canada and North America, they found that it was often based on low quality evidence (Ghui et al 2016, Prusova et al 2014, Wright et al 2011). This can be for lots of reasons: the number of women and babies included in the trials is often too small to detect uncommon but important outcomes, or there could be some bias because women and practitioners can’t be “blinded” to different approaches. In addition, trials are often not comparable because definitions of the interventions being tested vary between them, such as in the studies where delayed pushing could mean pushing only when the woman felt the urge to push, pushing after 30 minutes, pushing after one hour or pushing after some other defined time. The research might have been done many years ago when practices were different, there might have been missing data, or women might not have received the intervention they were supposed to and instead received a different intervention. For example, reviewers (Lemos et al 2017, Prins et al 2011) could not find good evidence for any particular pushing technique partly because women in the studies often ended up using a combination of approaches.

Research findings can be difficult to interpret with any certainty in other ways. For example, assumptions that one thing causes another may be incorrect; there might instead be a correlation or an association (Caughey 2009, Reed R 2017c, Sandström et al 2017). As Rachel Reed (2017c) explains, there is a correlation between increased ice-cream sales and shark attacks but neither of these causes the other. Both are affected by good weather. Aaron Caughey discusses the assumption that prolonged second stage causes more infections and excessive blood loss in mothers after birth. He suggests that there may be a correlation between prolonged second stage and infections and excessive blood loss, but that infection may actually lengthen second stage rather than a long second stage causing infections, and there may be an underlying reason for prolonged second stage and excessive bleeding after birth, such as reduced muscle tone of the mother’s uterus (“uterine atony”) (Caughey et al 2009). Increasingly, researchers acknowledge that it is difficult to prove causal relationships and talk more about correlations and associations.

Do we know when to intervene, and when not to?

It has been suggested that “the ideal management of the second stage should maximise the probability of vaginal delivery while minimising the risks of maternal and neonatal morbidity and death” (Caughey 2009 p337). Thus its aim is to increase the chances of the woman having a normal birth with minimal interventions while keeping her and her baby safe by carrying out timely interventions that are needed for the individual mother and baby. How to do this remains unclear, because it is very difficult to know if and when the advantages of an intervention might outweigh the risk of harm for an individual woman and her baby. As researchers have found, we don’t really know what length of time is normal for the pushing phase of labour, especially as this varies between women (Abalos et al 2018, Holvey 2014, Kimmich et al 2018, Kopas 2014). In some studies, researchers have tried to find this out by carrying out frequent vaginal examinations, but this doesn’t improve outcomes (Downe et al 2013); might introduce infection; and can be invasive, traumatic and uncomfortable for the woman; can interfere with the progress of labour; and erroneously assumes that women’s cervixes follow a linear and unidirectional pattern (Wickham 2005a,b).

Meanwhile, the somewhat arbitrary time limits placed on the second stage of labour have led to more women being diagnosed with “dystocia” (slow or difficult labour or birth) and having interventions, especially caesarean sections (Karaçam et al 2014). While it remains “unclear what the chance of actually progressing to a spontaneous vaginal delivery is as the second stage of labour progresses” (Caughey 2009 p338), when time limits are extended more women give birth spontaneously without suffering increased adverse outcomes (Cheng et al 2014). It seems that “the prevalence of adverse outcomes depends not on second stage duration but on when the birth attendant decides time’s up” (Goer 2017) and intervenes. One detailed review found that after three to four hours of pushing, disadvantages are “strongly associated with mode of birth. That is, spontaneous vaginal birth after a prolonged second stage results in much less morbidity than instrumental birth or surgical birth after prolonged second stage.” (Kopas 2014 p269).

One size doesn’t fit all

Each woman, baby and birth is unique, and while research can provide helpful guidance, it is unlikely that standardised definitions and practices will be best for every woman and baby. Observing the woman and baby, listening to women, and combining good research with individual women’s circumstances are crucial skills. A powerful example comes from an experienced and highly skilled midwife, Mary Cronk. Mary tells of how, when she first learned about the benefits of upright birth some decades ago, she began to encourage women to be upright for labour and birth. On one occasion she was attending a woman having her tenth baby. Whenever the woman lay down on her bed during her labour and birth, Mary encouraged her to get up. Each time the woman would go back to lying down on her bed. Finally, as she gave birth, Mary noticed that her varicose veins were very swollen and realised that for this woman who had never had problems giving birth, lying down was the best position for her to be in. She commented that this was a valuable lesson about watching and supporting rather than applying research, no matter what it purports to tell us. Research (while immensely useful of course) has its limitations (de Jonge at al 2008), and many midwives have similar stories.

What do we know and not know?

The results of most research on birthing tell us about what happens to women and babies in hospital, where obstetric ideology and practices dominate and where interventions are relatively high. The research doesn’t tell us about the impact of disturbing mothers during labour and birth, though many women report that it makes it more difficult for them to stay focused, or that it slows down or stops their labours or births and leads to interventions. It doesn’t tell us about the long-term physical and mental health and well-being of the mother and her baby and the bond between them. It doesn’t tell us about what happens when women and babies are cared for by midwives working within social models in community settings where medical interventions are used less frequently. Very few studies have looked at outcomes for birthing when women are undisturbed and encouraged to respond to their bodies instinctively throughout labour and birth.

“most research has focused on obstetric areas of interest: the clinical and emergency interventions needed when complications arise. Much less research exists on enhancing the respectful, supportive woman- and newborn-centred, high quality care for all that prevents complications and enables timely response when they arise.” (Renfrew et al 2019).

Given the uncertainties and limitations of knowledge, respecting other forms of knowledge and listening to women remains crucial (Reed R 2017c). This means taking their knowledge and concerns seriously. The importance of listening to women is a key message in the recent UK Confidential Enquiry Reports on maternal and infant mortality and morbidity but is still not always heeded (Mackintosh et al 2015, Rance et al 2013). Dialogue between women and skilled, empathic practitioners, and supporting women’s decisions is a cornerstone of safety and respecting women’s rights in childbirth (Lawrence Beech 2020, www.aims.org.uk, birthrights.org.uk).

To read more about the physiology of birthing a baby, the benefits of physiological birthing, why this isn’t usual practice in many countries and more about the research and evidence, you can see/buy the paperback version of Birthing your baby: the second stage of labour by Nadine Pilley Edwards here.

It’s also available as en eBook – see here.

#politicalcontextofbirth #childbirth #research #womensexperiences #maternitycare

© 2018-2020 Birth Practice & Politics Forum