Fateful moments and the categorisation of risk: Midwifery practice and the ever-narrowing window of normality during childbirth

Abstract

In this article, we examine the ways in which risk is categorised in childbirth, and how such categorisation shapes decision-making in the risk management of childbirth. We consider the ways in which midwives focus on and highlight particular adverse events that threaten the normality of childbirth and the life of the mother and/or her baby. We argue that such a focus tends to override other elements of risk, especially the low probability of such adverse events, resulting in ‘an ever-narrowing window of normality’ and a precautionary approach to the management of uncertainty. We start our analysis with a discussion of the nature of childbirth as a fateful moment in the lives of those involved, and consider the ways in which this fateful moment is structured in contemporary society. In this discussion, we highlight a major paradox; although normal childbirth is both highly valued and associated with good outcomes in countries like the UK, there has been an apparent relentless expansion of ‘the birth machine’ whereby birth is increasingly defined through the medicalised practices of intensive surveillance and technocratic intervention. We explore the dynamics that create this paradox using ethnographic fieldwork. In the course of this work, the lead author observed and recorded midwives’ work and talk in four clinical settings in England during 2009 and 2010. In this article, we focus on how midwives orientate themselves to normality and risk through their everyday talk and practice; and on how normality and risk interact to shape the ways in which birth can be legitimately imagined. We show that language plays a key role in the categorisation of risk. Normality was signified only through an absence of risk, andhad few linguistic signifiers of its own through which it could be identified and defended. Where normality only existed as the non-occurrence of unwanted futures, imagined futures where things went wrong took on a very real existence in the present, thereby impacting upon how birth could be conceptualised and managed. As such midwifery activity can be said to function, not to preserve normality but to introduce a pathologisation process where birth can never be categorised as normal until it is over.

Keywords:

• risk
• risk categorisation
• risk management
• midwifery
• midwives
• childbirth

Introduction

In this article we focus on the ways in which midwives categorise risk in the context of childbirth. We start our analysis with a discussion of the nature of childbirth as afateful moment in the lives of those involved, and consider the ways in which this fateful moment has changed and is structured in contemporary society. In this discussion, we highlight a major paradox; although normal childbirth is both highlyvalued and associated with good outcomes, there has been an apparent relentless expansion of ‘the birth machine’ (Wagner 1994), where birth performance is increasingly defined using medicalised practices of intensive surveillance and technocratic intervention. We then argue that the only way to understand this paradox is through fieldwork that captures the ways in which midwives make decisions during the birthing process. In the main part of this article, we will draw on fieldwork data to explore how midwives’ talk and practice structure risk and normality.

Contemporary childbirth: Midwifery, normality and risk

Childbirth can be seen as a fateful moment in which life is changed irreversibly. If all goes well, then a healthy baby is born. But if things go wrong, then the mother and/or her baby can be seriously harmed or even die. All those involved in a birth of baby ‘must launch out into something new, knowing that a decision made, or a specific course of action followed, has an irreversible quality, or at least that it will be difficult thereafter to revert to the old paths’ (Giddens 1991, p. 114).

In premodern societies, interventions in childbirth were limited with little proven efficacy, so that the outcomes were, from a modern perspective, the product of chance. Though most mothers and their babies survived childbirth, there was a relatively high probability of adverse outcomes in comparison to today. Loudon (1993) estimated that in the early eighteenth century, 1000 women died for every 100,000 births. The death of babies is difficult to calculate as historic records are limited and inaccurate. However Davenport's (n.d.) study of records for London in the eighteenth century suggested an average of 30 stillbirths per 1000 births. Her more detailed study of the parish records of St Martin-in-Fields indicated around 60 deaths per 1000 in this location, with the rate fluctuating over the century between 40 and nearly 100 deaths per 1000 births. Thus, in eighteenth century London, based on the St Martin's figures, there was a .06 probability that the baby would not survive, and a .001 probability that the mother would not survive each birth.

In the twenty-first century, childbirth has become a vastly safer process in developed countries. The probability of dying during pregnancy and childbirth has fallen substantially. The Centre for Maternal and Child Enquiries (2011) indicated that, in the three years 2006–2008, only 261 women in the UK died as a result of their pregnancy, and that 4.67 deaths per 100,000 pregnancies could be directly attributed to the pregnancy including the childbirth. The stillbirth rate has also declined substantially, to 5.2 per 1000 births in 2007 (Confidential Enquiry into Maternal and Child Health 2009, p. 3). Thus, in countries like the UK, childbirth has become a much safer process.

However increasing safety is not the only difference between traditional and modern childbirth. There is also increase in human agency and choice. Traditional childbirth can be seen as a natural process in so far as lack of knowledge and skills limited human capacity to influence the outcome. Women giving birth had to cope with and manage the uncertainty of the outcome using such resources as were available to them.

The growth of scientific knowledge has radically changed the nature of childbirth in contemporary society, and provides the basis for expert risk management. Such expertise has transformed the capacity for informed decision-making through the application of human agency to change the probabilities of outcomes. Risk is now central to the ‘rational’ management of uncertainty. Since the end of the nineteenth century, all childbirth in the UK has been under the state-sanctioned surveillance of experts. In the UK and elsewhere, the type of surveillance used depends on expert classification of risk. Births by ‘low risk’ mothers can be supervised by midwives, and take place in low tech facilities, even in the mother's own home. Births by ‘high risk’ mothers should be supervised by obstetricians and take place in high-tech facilities. Despite this, however, high risk birthing environments continue to be the most the most used setting for the majority of mothers in the UK, with 88.2% of births in England taking place within an obstetric-led facility in 2010–2011 (NHS Health and Social Care Information Centre 2011a). In high-tech units, obstetric surgeons can, as a last resort, use a caesarean section to remove the baby from its mother's womb. The modern use of sections started in the late nineteenth century, and is now routinised in many health care systems. In the UK, the average caesarean section rate for both 2007–2008 and 2008–2009 was 24.6% (The Health and Social Care Information Centre 2009).

The development of scientific knowledge, and with it the capacity to make decisions that influence the outcome of the birthing process in the way intended, has altered the status of birthing attendants. They have become responsible and accountable for their decisions. Birthing is no longer a purely ‘natural’ process in which the outcomes are the product of chance and adverse outcomes are unpreventable ‘accidents’. It is increasingly viewed as ‘man-made’, and therefore adverse outcomes cannot be accidental – see Green (1999) for an analysis of ways of the ways in which risk has eroded the concept of the accident – but must be the fault of those who made the decisions. As Douglas (1990) has argued, the concept of risk underpins the development of a ‘blame culture’ in which all harmful events are seen as a product of human agency, and every misfortune is someone's fault. She argues that ‘under the banner of risk reduction, a new blaming system has replaced the former system based on religion and sin’ (Douglas 1992, p. 16).

Since the probability of actual harm to the mother or the baby (such as massive haemorrhage or significant birth asphyxia) during the process of spontaneous birth is small, midwives should be able to treat mothers as being capable of birthing their offspring without undue concern for risk. Indeed, midwives in the UK describe themselves as practising within a paradigm of normality (Gould 2000, Sandall et al. 2009, Midwifery 2020). Within this framework, women and their pregnant bodies are conceptualised as being essentially competent. Such a framework positions the midwife as a facilitator whose professional understandings of the spontaneous physiological process of birth can be applied through practice to ensure that babies are born with as little disturbance and intervention as possible (Rosser 1998, Leap 2000). Midwifery discourse tends to privilege notions of birth as anormal process (Davis-Floyd et al. 2009).

It follows that the categorisation of birth as high risk should be rare and exceptional. However, as we noted above, in practice the majority of births in England take place within a high risk birthing facility regardless of the risk status attributed to the pregnancy (NHS Health and Social Care Information Centre 2011a). There are two principal ways in which the decision to categorise a birth as high risk can be taken. Before the birth starts a midwife may judge that a mother has certain characteristics that place her in the high risk category, for example, if she is above a specified age when having her first baby. Alternatively, the decision may be made during the birth process when events do not follow the normal (and prescribed) trajectory, for instance, if the dilation of the cervix falls outside the ‘normal’ range. In this paper we will focus on this second type of high risk categorisation.

In this article, we explore why midwives who are committed in principle to normal childbirth are unable to articulate and defend normality. Instead, they often highlight the dangers of birth, creating the medicalisation of birth by categorising anincreasing proportion of births as high risk. The paradox which we focus on wasarticulated by Carina, a midwife who participated in our study, in terms of ‘whatseems to be an ever-narrowing window of normality’ (extract from interview with Carina, midwife).

Methods: Using ethnographic methods to access midwives’ practice and the tacit knowledge which underpins them

This paper is based on data drawn from an ethnographic study designed to explore how midwives make sense of risk. In this study, we used methodological tools that could make explicit midwives’ tacit knowledge, their common-sense understandings about risk and normality. The aim of the research was to observe and record situated midwifery talk and practice in the various clinical settings in which midwives work. The most effective way to access such activity was for the lead author (a qualified midwife) to participate in and observe birthing in different settings. This ethnographic approach was not employed in the early anthropological, positivist sense, as an attempt to capture what was ‘really out there’. Instead the lead author adopted a reflexive approach in which her identity as a midwife and a researcher was implicitly woven into the process of data collection, and also shaped the production and analysis of an ethnographic text.1

To observe midwifery practice and talk in different settings, we selected four very different settings that represent the major organisational forms for birthing and midwifery practice in the UK. These settings were: a obstetric-led unit with all the medical facilities for high risk births (3361 births per year); two midwife-led units, one located in a hospital with access to back-up medical facilities in case a birth shifted from low to high risk category (606 births per year); a free standing midwifery-led birthing unit where the reclassification of a birth into the high risk category involved a 40-minute transfer journey (378 births per year) and, finally, ahome birth service (224 births per year).

For this study, we used ethnography in its broadest sense, not so much as a set of research methods or analysis techniques, but as a ‘concern with the meaning of actions and events to the people we seek to understand’ (Spradley 1980, p. 5). As such, we considered the methods to be the most effective for achieving our desired objective – understanding situated midwives’ meaning-making. As the fieldwork and associated analysis developed, we adapted, adopted or, in some cases suspended, various research tools (Clifford and Marcus 1986, Denzin 1998, 2002). Thus, our emphasis approach changed, depending on the issues raised by the data analysis, and included a combination of:

•	participant observation (Malinowski 1932, Spradley 1980) (n = 42) of midwifery labour care with midwives of various levels of seniority and in various care settings in order to observe what actually happens in practice
•	non-participant observation (n = 15). This was mainly done in ‘behind the scenes’ National Health Service (NHS) observations, such as board meetings, staff meetings, protocol meetings and risk case reviews at unit and trust level, to gain insight into organisational issues which constrain and facilitate different kinds of practice
•	ethnographic interviews (Spradley 1979) with managers, midwives, students and maternity and midwifery pressure group members (n = 27),2 which allowed for the testing of hypotheses and the scrutiny of incidents observed during participant observation
•	text analysis (Fairclough 2001, 2003) of protocols, policy documents and key professional texts to give a broader social and cultural contextualisation to the observation and interview data.

Analysis

Since the main objective of the research was to access midwives’ intuitive knowledge, the analysis involved a careful reading and content analysis of field notes, interviews and related texts (Reissman 1993, Graneheim and Lundman 2004). Ongoing analysis was carried out alongside, and guided, the fieldwork. Following an initial reading, we undertook closer scrutiny of the texts produced within the study using conversational and discourse analysis techniques (Silverman 1988, 2004, Van Dijk 1993, 1997, Fairclough and Wodak 1997, Wodak 1999, Fairclough 2001, Gwyn 2002). This initial content analysis was checked and corroborated through the project supervision process, and was then intensified towards the end of the research, using ATLAS.ti to check for reliability and validity of the analysis; codes were networked and checked for density to ensure groundedness. In this paper, we focus on the data relating two codes – ‘normal birth’ and ‘risk’. These were both densely populated codes, although ‘normal birth’ was more complexly networked and denser than risk.

Access and ethics

We accessed the initial sample (n = 33) using a process of self-selection, following a recruitment and information campaign targeted at all midwives working in the selected sites; and then expanded participation through opportunistic, snowball techniques (Bryman 2004), with some attention to purposeful structuring to maximise diversity. We obtained written consent and sequential verbal consent from all those involved in the study, and ‘cleaned’ all transcripts and field notes by removing identifying features prior to analysis. We sought ethical approval from both national and local NHS ethics committees, and obtained full approval for the study in February 2009. The NHS Trust’s Research and Development governance team, the Head of Risk, Assurance and Legal Services and the Head of Midwifery reviewed and approved the project before we started data collection. The lead author had a NHS licence to practice for the duration of the data collection.3 All data published in this paper have been ‘cleaned’ to remove identifying features, and all names have been changed.

Findings: Evidence from midwifery practice and discourses

Adverse outcomes, blame and risk

The midwives who participated in our study were aware of their accountability, especially in relation to adverse outcomes. As Heather indicated, positive outcomes did not attract attention or praise, whereas adverse outcomes attracted both attention and blame:

Well you see, if the outcome was fine it would never really get questioned would it? If there was a poor outcome you would be asked, ‘Why I did that?’ Good outcomes, well they never get investigated or celebrated really for that matter, it's only the, the poor outcomes. They're what everyone hears about, they're the things that make people sit up and take notice, you see. (Extract from interview with Heather, senior midwife)

While accountability and potential blame formed the backdrop for much of research, in some circumstances it was foregrounded. For example, the demeanour of one midwife changed radically during the research. When she first worked with the lead author, she was confident and bubbly. However, as the field notes indicate, following her involvement in an internal risk inquiry she lost confidence and self-belief:

Helen kept reiterating that she was nervous, explaining that whereas she had felt clinically confident in the past, recent events had made her feel ‘so shit’ that she was sometimes unable to make the simplest of decisions sometimes. The way she overcame her confidence crisis was to picture herself discussing the case with the consultant midwife – P. ‘I know this must be okay’, she told me, ‘because this is what P would say. She would say she is not in labour so I know it's okay to treat her like this’. Helen and I left the room (where a mother was labouring) so that Helen could discuss her care plan with another midwife who had just arrived at the unit. During our conversation, Helen revealed more details about the incident that seemed to be haunting her practice so much. Helen explained that she was not traumatised by the event itself, stressing, withtears in her eyes, that `I know I didn't do anything wrong. I know I am a good midwife … [I] know we are told it is not a blame culture, but this thing has been all about blame … It makes you feel like a bloody criminal! This job can be so shit sometimes'. (Extract from field notes HJ 4).

Although personal involvement in an incident and subsequent inquiry highlighted the way in which blame was allocated, those midwives with little direct experience of adverse outcomes and related inquiries were made aware, through activities such as staff training sessions, that such circumstances could happen to them one day.

Fateful moments and the risk of adverse outcomes

For both midwives and mothers, birth was a fateful moment, but what was at stake differed. The midwives who contributed to our study were aware of their accountability, and of the personal consequences associated with adverse outcomes. So, if the birth did not go to plan, then not only could the mother and baby be harmed, but also the midwife would also need to account for her actions in an inquiry which would start from the premise that errors had been made.

As Giddens (1991, p. 127) has noted, one way of managing the threat of uncertain outcomes is through denial. When risk is part of everyday activities, such as crossing a road or preparing and eating food, familiarity and habit enable individuals to deny or bracket out the threat, creating a protective cocoon of routine. For birthing mothers, one would expect the unusual and atypical nature of the event to puncture the protective cocoon of normality. However what is unusual and exceptional for most mothers, childbirth, should be normal and routine for qualified midwives. To be qualified and granted a licence to practice, they need to provide evidence that they have participated in at least 40 births. One might expect the normality and routineness of most childbirth to sustain a protective cocoon for midwives (see Menzies 1960, for a discussion of the routines and structures in general nursing as defences against anxiety).

The fieldwork produced little evidence for the presence of a protective cocoon ofroutine. Midwives indicated that in their practice they were always alert to the possibility of adverse outcomes:

We are very risk averse aren't we.? We, we will say, within the NHS, the majority will sayit [birth] is normal after the event. (Extract from interview with Susan, a senior midwife)

From this perspective, all births were potentially hazardous, and normality could only be recognised in hindsight, after a woman had given birth to her baby and was no longer in the crisis of labour. Interviews with midwives, including those who were senior and experienced, indicated that during childbirth imagined risk was ever-present in a future inhabited by potential adverse events. Such adversities which this ‘normal in retrospect’ lens highlighted did not necessarily have much connection to events in the present, especially the probability of such events. Rather, it reflected the ‘high consequences’ of these events. What was being bracketed out was not potential adverse consequences but another important component of risk, their (low) probability. Midwifery practice coalesced around an apparently irresistible desire to anticipate and avoid even the smallest possibility of an adverse outcome, even when this might involve abandoning any commitment to the notion of normality:

Maria: I always tell people that there is high risk and there is low risk but that there is no such thing as no risk … Risk is much more important even if it might not be clinically significant …

Researcher: A 1:10,000 risk, is that a high risk or low risk?

Maria: Depends if you are the one really doesn’t it. [Laugh] . (Extract from interview with Maria, midwife)

In the context of midwife practice and talk, low risk or normal birth, despite being a preferred outcome, appeared to have a limited temporal existence, in that it could only exist in the past, after the events of birth had concluded. The ways in which many of the midwives, particularly those in positions of authority, talked about birth indicated that fears about the possibility of things going wrong functioned to destabilise professional confidence in birth normality. Such anxieties were evident in several of the training sessions which the lead author attended, which often focussed on the ways in which things could and did go wrong:

As I looked around the room many of the midwives in the group were grimacing in horror as the session unfolded. Furthermore, the coffee break which followed this session was spent exchanging and collaborating over stories of near misses where risks lay waiting to develop into future Confidential Enquiry statistics. (Extract from field notes SD1)

The risk paradox: Midwives’ commitment to normality

The overshadowing of midwifery practice by imagined futures containing potential adverse events was paradoxical as much of their talk stressed the positive value of normal birth. Normality was consistently represented as a cultural ‘good’. Its merits were simply taken-for-granted, and this view was so deeply engrained into their shared tacit knowledge that a positive moral loading of the term was common to all midwives whom the lead author spoke to. When participants talked to her about normality, they simply assumed that she, as a fellow midwife, would share their understanding and appreciation of the term and its virtues. Explicit explanation was therefore deemed irrelevant, even comical. A belief in normality as a cultural good was a basis for identity as a midwife, something to be aspired to, and a source of professional pride and confidence:

Midwives very often come into the profession because they are women and intrinsically that they understand that birth is a normal process. (Extract from interview with Silvia, midwife)

To be a midwife was to have an undefined and indefinable belief in the possibility of normality in childbirth – a notion reminiscent of the act of faith that underpins trust. Furthermore, several of the midwives we spoke to suggested that normality and midwifery were symbiotically linked – one could recognise one through the presence of the other. And birth could remain ‘normal’ even if there was some (limited) physical intervention in the woman’s body:

Mmm, things like a stretch and sweep 4 and using entonox 5 … well they are all things done by a midwife aren’t they, so I suppose that doesn't make the birth, you know, just because a woman has those sorts of things doesn't mean her birth isn't normal, does it? … So yer, you can have midwifery care, midwifery care and normality are sort of … well they go together really don't they? They are the same … because you see, midwifery care is low risk care isn't it? Mmm … and a vaginal birth, yer normal vaginal birth, and, hopefully, a natural third stage, physiological third stage, all the stuff that can be managed exclusively by a midwife. (Extract from interview with Rachel, midwife)

Thus, for Rachael, midwifery practice was symbiotically linked to normal birth. The boundaries of normality were marked by autonomous midwifery intervention, described here as the administration of entonox and/or the undertaking of a ‘stretch and sweep’ for induction of labour. Midwifery activity, even when it is directed towards interfering with the physiological birth process or introducing pharmaceutical agents to disrupt the woman's experience of birth, coincided with normality to such an extent that they become virtually one in the same thing – a normal birth was a midwife-managed birth.

In midwife talk, the term ‘normal birth’ was frequently pre-fixed with ‘nice’. Thislexical choice had a normative function, confirming the speaker's professional allegiances, and emotionally defining normality as a professional good, an interpretive framework which the following field note entry illustrates:

In the nurses’ station [on a busy obstetric lead labour ward], Emma, a midwife was giving a history of the woman she had been caring for in ‘hand-over’ when she told the oncoming day staff: ‘Despite all that [referring to a catalogue of difficulties the mother had encountered during her labour] we did manage to get a nice normal delivery.’ The reaction of the other midwives whom this comment was aimed at was one of approval, even mild congratulation. Emma had done well – the fact that she had managed to ‘get a nice normal delivery’ reflected well on her midwifery skills. (Emphasis added. Extract from field notes E14)

Not only was ‘normal’ pre-fixed with ‘nice’. In addition, the word ‘managed’ in this context suggests that normal birth should be considered something of an achievement. Good midwifery and normality appear mutually dependent. Given that normality is a preferable outcome, and that normal birth is less hazardous, we need to consider why it is so difficult to protect in current midwifery practice.

The vulnerability of normality: Its absence in talk and practice

One of the major problems in categorising a birth as low risk or normal is that the precise definition of normal is elusive. Although normality is highly valued by midwives, they find it difficult to define (and measure). Our attempts to elicit what the participants meant by the term frequently met with laughter or expressions like ‘Oh no!’, ‘I don't know', ‘How am I supposed to answer that?', ‘That's a difficult one'; or even on one occasion, ‘You can't expect me to be able to tell you that!' (Extracts from interviews with midwives).

This inability to define normality was also evident in our interviews with a representative from the Royal College of Midwives (RCM). While discussing the impact of the College's ‘Campaign for Normal Birth' (Day-Stirk 2005, RCM 2010), the representative told us that, in the UK, midwives are so desensitised by over-use of the term ‘normal birth' that it has become devoid of real meaning:

We have had the normal birth debate such a long time in the UK, and people are quite … We are slightly blasé about it, and people, they sort of … they have had enough. I mean if I talk to a UK midwife about normal birth, they say, ‘Well what's that? What's normal to you was not normal to us and does it mean anything at all anymore in the context of modern obstetrics?' It is almost as if it is, I don't know, kind of a nothing, if you like. (Extract from interview with representative)

At the same time as viewing normal birth as desirable, the midwives who participated in this study struggled to conceptualise it as a concrete concept. Rather, it was frequently described as something that could only be defined in terms of theabsence of other more tangible attributes. More specifically, normality was something that revealed itself through the absence of risk indicators or specific risk management measures.

This definition by absence underpins the national ‘Normal Birth Consensus Statement' in which normal birth is defined via a series of negatives as:

Without induction, without use of instruments, not caesarean section and without general, spinal or epidural anaesthesia before or during delivery. (Maternity Care Working Party 2007)

This statement was developed, ironically, ‘to encourage a positive focus on normal birth' (Maternity Care Working Party 2007, p. 2). However, the choice of wording in this statement renders normal birth without substance. Instead, it is only present as alinguistic absence.

Given this wider context, it is not surprising that the majority of midwives who participated in this study saw normal birth in terms of what it was not, as an absence rather than a presence. For example, one midwife defined normal birth in the following way:

Yes, I mean normal birth is a labour that has had minimal intervention, I mean medical intervention, no medical intervention, yer no medical intervention. That includes epidural. (Extract from interview with Rachael, midwife)

Another midwife described normality in terms of a negative tick list, the absence of a series of complications:

Well even now, I still do it. I, I go through it and, you know, the woman's pushing, and I'm like, ‘Okay, is this all normal?' ‘Yep we've got not foetal distress; we've got noproblem with the woman's observations; erm' she has got this far and there is nothing’. It is almost like a tick-list in my mind ticking-off … There is nothing [abnormal], so it must be normal. (Extract from interview with Hannah, an independent midwife)

In these discussions, normal birth was ‘the subject that is not one’ (Butler 1999, p.2). In midwifery conversation, normality has no language of its own. It has to be defined against the dominant discourse of high risk (Kress 1989) which invokes the language of pathology and medical intervention. There were no words with which to police the boundaries of normality, no linguistic tools to protect its integrity. Normality could only be signified through absence within the privileged discourse ofrisk.

The absence of normality not only was evident in midwives’ talk, but could also be detected in the official texts designed to structure their practice, and in practice itself. For example, the Midwives Rules and Standards (Nursing and Midwifery Council 2004) define the legal framework for midwifery practice. These rules and standards delineate the midwifery `care' without any explicit reference to normality, defining midwife care as:

preventative measures, the detection of abnormal conditions in mother and child, the procurement of medical assistance and the execution of emergency measures in the absence of medical help. (Nursing and Midwifery Council 2004, p. 36)

All four of the midwifery activities listed by the Council coalesce around the language of risk. According to the statutory regulative body, midwifery activity has nothing to do with normality. Rather, it is about detection and prevention of risk: being alert to the possibility of problems; accessing medical support in order to manage risk; and being capable of managing unexpected crises while medical assistance is sought.

It is therefore hardly surprising that our observations of midwifery practice showed that midwives were on the constant look-out for abnormality and risk. Midwives routinely ‘manage’ the birthing process by measuring vital signs of boththe mother and baby, and monitoring ‘progress’ in labour – assessing uterine contraction and cervical dilatation. At the point when labour was identified, midwives initiated detailed surveillance and record keeping. Such intensive monitoring was applied to both normal/low risk and abnormal/high risk births, bringing all labouring women into visibility. The midwives involved in this study introduced this surveillance in a taken-for-granted manner. Its precise purpose wasrarely made explicit to the birthing mother. Rather, each intervention was introduced as part of the customary care plan, the purpose of which was treated asself-evident. Midwives commonly introduced monitoring activities with commentslike:

`I'm just going to have a listen in again now, just to make sure the baby is okay’. This preceded exposing the woman's abdomen to auscultate the foetal heart (Extract from field notes GT 20, author's emphasis) or ‘Can I have your arm a minute. I need to check your blood pressure'. (Extract from field notes RS1, authors’ emphasis)

Underpinning these mother–midwife interactions was the implicit assumption that repeated checking, rechecking and recording of parameters such as fetal heart and maternal blood pressure were beneficial in terms of risk management. Once the measurements were taken, they were plotted in the partogram,6 and/or written into the labour care section of the maternal notes. These measurements become central tothe categorisation of risk, with normality indicated by the absence of signs of abnormality:

I suppose [normality is] no intervention. Just letting the woman listen to her body and do it herself, yer … And well, you know, when everything is in the normal parameters; making sure, erm, like keep the woman and baby safe by making sure, you know, you are listening in every 15 minutes and that they don't come out the brackets thing, the chart thing … partogram. (Extract from interview with Harriet, a student midwife).

The midwives’ talk to pregnant women following these measurements was generally quite cheerful. However, this approach did not always allay the fears that surveillance seemed to introduce, as the following extract from the field notes suggests. Sarah, a first time mother, was undergoing a routine vaginal examination to measure the dilatation of the cervix and descent of the baby's head.

During the examination the room went very quiet. Sarah is lying flat on the bed as instructed by the midwife. No explanation is given to explain why this is necessary and no attempt is made to perform the examination in a position that might be comfortable for Sarah. It is as if any concerns for Sarah's physical or emotional comfort seem to be temporarily suspended given the seriousness of the task of finding out what is going on. The findings of the exam are not mentioned during the procedure, Sarah and her partner are left wondering and waiting, there is a palpable sense of tension. Afterwards Pauline [the midwife] explains what she found. Both parents look anxious and although the VE7 shows progress of the labour was normal, both Sarah and her partner needed to repeatedly have this confirmed. Pauline did not seem surprised by this reaction, she smiled and reiterated that everything was fine at least three times. She then left the room to record her finding in the notes and on the board. (Extract from field notes PS 14)

In this case, Sarah's labour was following the desired partogram trajectory – she had progressed according to the parameters set by the chart. Although normality was confirmed, the process introduced a sense of uncertainty. Before the examination, both Sarah and her partner had been managing the labour process effectively and pretty much independently. But when the time came to monitor the progress, to check for normality or more precisely to hunt for abnormality, their confidence in the process and their understanding of the active role they could play in that process seemed to dissipate. Indeed, although Pauline, the midwife, stressed that progress was good, Sarah responded by asking ‘Is there anything else I should be doing. Am I doing it right?’ (Extract from field notes PS14). Even when a woman’s labour fits within the partogram trajectory, the very process of monitoring progress simultaneously confirmed and disturbed normality.

Through the action of routine surveillance, midwifery activity was oriented not to confirming normality, but to searching for the absence of abnormality. This was a subtle but significantly different task which tended to privilege imagined possibilities of ‘what if things go wrong’, and thereby operated to unsettle a woman's confidence in her body's ability to birth her baby successfully. Although midwives wanted to reassure mothers, their actions tend to expose the unstable base on which understandings of normality rest, as well as the unarticulated issue of accountability if anything went wrong. The labouring woman and her birthing partner were far from oblivious to this instability. As the above quotation illustrating Sarah's need for professional reassurance suggests, parents could and did easily recognise the midwife's concern with the ever-present ‘virtual risk object’ (Van Loon 2002, Heyman et al. 2010).

Discussion

The categorisation of a birth as low or high risk had important consequences for theways in which it was managed. Categorisation as low risk enabled a normal midwife-supervised birth to be initiated, whereas placing a birth in the high risk category triggered increased surveillance and medical intervention. Midwives treated ‘normal’ birth as a self-evident good. But, because they were unable to define and measure normality, this categorisation was always tentative, and based on a provisional absence of risk indicators. Because the midwives who participated in our study found it hard to describe, talk about and measure normality and low risk, they effectively created an imagined future colonised by potential high risk that could at any moment be made visible through their continual surveillance.

Normality was absent from both official prescriptions about midwifery practice and midwives’ talk. This absence is not just semantic, which would be disturbing enough given the moral loading of the term in midwifery talk. It is absolute.8 As we noted, the Nursing and Midwifery Council's Midwives Rules and Standards require midwifery activity to focus on imagined futures where the possibility of pathology is ever-present, at the expense of the mostly much more probable alternative future inhabited by normality. This precautionary approach to risk management disregards the probabilities of events and ‘casts the future principally in negative, potentially catastrophic terms’ (Alaszewski and Burgess 2007, p. 349). As Heyman et al. (2010, pp. 22–24) have argued the answer to the question of contingency, i.e. ‘What might happen’ is ‘Absolutely anything!’. The data presented in the present paper show how midwives selectively populate the infinity of possibility with what might go wrong. This interpretive lens is shaped by the organisational and wider culture in which they operate and discounts the small magnitude of most of the relevant probabilities.

Furthermore, the inability to articulate and defend normal or natural birth provides the basis of the blame culture within midwifery. As Douglas and Wildavsky (1982, p. 35) have argued:

Blameworthiness takes over at the point where the line of normality is drawn. Each culture rests upon its own ideas of what ought to be normal or natural. If a death is held to be normal, no-one is blamed.

The challenge for midwives is that, despite their efforts and commitment, ‘blame-free’ birth does not exist. All births are supervised by experts, and when something goes wrong, a search and inquiry starts to identify who and what is to blame. This organisational context provides the context for midwife talk and practice.

Midwives working in the birthing environment contemplate two possible imagined futures. In one, the baby is born through the natural process of spontaneous delivery and unnecessary medical interventions pose an unacceptable risk of iatrogenic harm. In the other, nature fails, threatening the health of mother and/or baby, and serious harm might occur without timely intervention involving technological procedures. Importantly, both of these imagined futures are value-laden, with the former considered by midwives as the most desirable to both mother and midwife (Newburn 2006). As the evidence presented suggests, the latter, although less desirable, represents the more persuasive of the two imagined futures within the current birthing climate. In this climate, caesarean section rates have risen sharply, both nationally (Mander 2008, NHS Information Centre 2009) and globally (World Health Organization 2009); and 97% of women end up giving birth within ahospital environment ‘just in case’ (Devries et al. 2001, NHS Information Centre 2011b). As Murphy-Lawless (1998, p. 21) has pointed out, this anxiety about risk not only disempowers both midwives and birthing mothers:

The tendency has … increasingly been to define every aspect of pregnancy and birth in terms of risk in a mistaken attempt to cover all possible eventualities. In this sense, the entire female body has become risk-laden.

Conclusion

In this article we have shown how the categorisation of risk shapes, and is shaped, by the social context for decision-making. As normality lacks any language of its own through with which midwives can defend its boundaries, it is easily subsumed by the linguistically and culturally more secure notion of risk. Through the analysis of published texts and midwifery activity, we have shown how midwives create an ever closing window of normality in which all births are categorised as risky. Within a linguistic context where normality and unassisted safety could only be envisaged as the non-occurrence of unwanted futures,imagined futures where things go wrong took on a very real existence in the present, thereby impacting upon how birth could be conceptualised and managed. As such, midwifery activity functions not to preserve normality, but to introduce apathologisation process where birth can never be imagined to be normal until it isover.

Notes

1. This section provides only a brief descriptive account of methods. Detailed discussion of the methodological implications of the research design, in terms of author impact and construction of identity, translation of culture, sequential consent, etc., has been presented elsewhere and is beyond the remit of this paper.

2. Ten midwifery managers, 10 midwives, two student midwives, two independent midwives and threepressure group representatives.

3. The first author is a registered midwife, but for the purposes of the study is licensed to practice as a maternity care assistant.

4. Stretch and sweep is a procedure where a midwife or doctor will ‘sweep’ a finger around the cervix during an internal examination. The aim is to separate the fetal membranes from the cervix, leading to a release of prostaglandins and subsequent onset of labour (National Institute for Health and Clinical Excellence 2008, p. xii).

5. Both interventions into the birth process are done by midwives without any recourse to the multidisciplinary team. These are what might be called midwifery interventions and, assuch, are seen not as interventions at all, but as part of a process for facilitating normal birth (Annandale 1988).

6. The partogram, or picture of labour, is a universal chart designed in the 1970s for recording observations of mother and baby, including contraction pattern rate and strength, cervical dilatation, etc.

7. Vaginal examination.

8. The text being analysed here is the printed 2004 version. It should be noted that the online version has an update to include a more up-to-date International Confederation of Midwives definition which does include reference to normality. The modality of this reference, however, is significantly reduced as the word is sandwiched between other risk-orientated concerns and appears in a list of five activities, four of which coalesces around risk and abnormality.