Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Bioethics. Author manuscript; available in PMC 2010 January 1.
Published in final edited form as:
PMCID: PMC2628951

Finding Autonomy in Birth*

The Obstetrics and Gynecology Risk Research Group, Rebecca Kukla, PhD, Miriam Kuppermann, Margaret Little, PhD, Anne Drapkin Lyerly, MD, MA, Lisa M Mitchell, PhD, Elizabeth M. Armstrong, PhD, MPA, and Lisa Harris, MD, PhD


Over the last several years, as cesarean deliveries have grown increasingly common, there has been a great deal of public and professional interest in the phenomenon of women ‘choosing’ to deliver by cesarean section in the absence of any specific medical indication. The issue has sparked intense conversation, as it raises questions about the nature of autonomy in birth. Whereas mainstream bioethical discourse is used to associating autonomy with having a large array of choices, this conception of autonomy does not seem adequate to capture concerns and intuitions that have a strong grip outside of this discourse. An empirical and conceptual exploration of how delivery decisions ought to be negotiated must be guided by a rich understanding of women’s agency and its placement within a complicated set of cultural meanings and pressures surrounding birth. It is too early to be ‘for’ or ‘against’ women’s access to cesarean delivery in the absence of traditional medical indications - and indeed, a simple pro- or con- position is never going to do justice to the subtlety of the issue. The right question is not whether women ought to be allowed to choose their delivery approach, but rather, taking the value of women’s autonomy in decision-making around birth as a given, what sorts of guidelines, practices, and social conditions will best promote and protect women’s full inclusion in a safe and positive birth process.

Keywords: birth, autonomy, cesarean section, vaginal delivery, cesarean delivery on maternal request, CDMR

Over the last several years, as cesarean deliveries have grown increasingly common,1 there has been a great deal of public and professional interest in the phenomenon of women choosing to deliver by cesarean section in the absence of any specific medical indication. Because there is no validated and universal medical code to indicate such deliveries,2 we have little sense of whether this is in fact a phenomenon that is substantially impacting the overall cesarean rate or a rare novelty that rises little above the level of urban legend.3 But the issue has sparked intense conversation, as it raises anew questions about the nature of autonomy in birth. In March 2006, the National Institutes of Health convened a panel of independent experts at a State-of-the-Science Conference to assess the available scientific evidence relevant to trends in the use of cesarean delivery, the short- and long-term benefits and risks to mothers and babies, and the ethical issues surrounding performing a so-called “cesarean delivery upon maternal request” (henceforth CDMR). The panel concluded that there was insufficient evidence to issue a recommendation concerning the relative safety of planned vaginal and cesarean births, and called for further study.4 Debate - and discomfort - ensued concerning whether it is appropriate to accede to requests for such cesareans, and whether, if so, practice should be extended to offering this option more broadly.

There is an obvious sense in which expanding women’s voluntary access to cesarean deliveries would represent an increase in the options available to women. To the extent that we understand patient autonomy as equivalent to informed choice, and expanding autonomy as a matter of expanding available choices, it may seem that any move in the direction of allowing women more delivery options is a move in the direction of enhancing women’s autonomy during birth. Indeed, several published discussions of the ethics of CDMR have cast the issue as one of balancing the principle of autonomy, which counsels in favor of giving women increased options during delivery, and the principle of beneficence, which counsels protection of the welfare of mother and child. In these discussions, it is taken as obvious that incorporating CDMR into obstetrical practice enhances women’s autonomy.5

And yet we can be certain that any new permissiveness with respect to CDMR will not be received as a coup for women’s autonomy from all quarters. Writers who are particularly attentive to the increasing medicalization and technological management of pregnancy often portray the very possibility of allowing providers to perform cesarean deliveries in the absence of medical indications, even if done in the context of acceding to a request for the procedure, as an assault on women’s autonomy, which will further take the process of birth out of women’s hands and undercut their agency. They worry that any further move towards normalizing cesarean deliveries contributes to undermining women’s confidence in their own bodies and agency, and minimizes their participation in their own birthing process.6 The International Cesarean Awareness Network, or “I-can,” goes so far as to rule out the possibility that the choice to have an elective cesarean is ever autonomous. Their online “statement of beliefs” contains the principle, “Patient-choice cesareans are unethical and immoral on the part of physicians. Women are not being fully informed of the risks of this option in childbirth, and therefore make decisions based on cultural myth and fear surrounding childbirth.”7

It is clear that there exist deep disagreements, whether implicit or explicit, as to what constitutes autonomy in the domain of birth. Whereas mainstream bioethical discourse is used to associating autonomy with having a large array of choices, this conception of autonomy does not seem adequate to capture concerns and intuitions that have a strong grip outside of this discourse. A proper ethical analysis of birth practices, then, cannot begin by trying to ‘balance’ autonomy and welfare protection, but must (among other things) first provide a more richly textured understanding of autonomy in the context of birth.

In this paper, we attempt to open the path for developing such an understanding. We restrict our attention here to decision-making surrounding approach to delivery,8 and more specifically to scheduled cesarean section versus planned vaginal delivery, although this is of course only one context among many where issues of women’s autonomy arise in the domain of birth. We hope that the discussion we begin here could be productively applied to debates concerning home versus hospital birth, the use of surveillance technology such as fetal monitors, decisions about vaginal birth after cesarean section (VBAC), and so forth. The point of our title is double: we wish to find autonomy in birth in the sense of conceptually clarifying it, but we do so for the purpose of understanding how medical guidelines and health policy, women’s health activism, and patient education can be structured so as to help women find autonomy in the process of birth.

Most will agree that the process by which providers and patients come to a decision about approach to delivery should ideally be one that supports patients’ autonomy and effectively promotes high-quality care and good outcomes for both mothers and babies. But if autonomy cannot be assumed to be reducible to the right to make an uncoerced choice, then we face a substantive question concerning how women’s voices and values should be incorporated into decision-making about mode of delivery. Despite increasing acceptance of the ideal that providers and patients should engage in shared decision-making,9 there has been little discussion of how and when providers and patients ought to collaborate so as to arrive at a plan of action for birth. It is widely accepted that, except perhaps in very unusual cases, patients have the right to decline invasive treatments. But beyond this limit case, there has been no systematic analysis of the role that women’s values and preferences should play in determining an approach to delivery, or of how these decisions can be made in ways that effectively promote women’s autonomy. For instance, in the December 2007 American College of Obstetrics and Gynecology Committee Opinion on CDMR, which summarizes the evidence on the clinical outcomes of cesarean delivery and focuses on clinical contexts in which the procedure should not be performed, women’s values are simply mentioned in the middle of a list of “other factors” that the provider should consider in responding to requests for cesarean delivery: “When a woman desires a cesarean delivery on maternal request, her health care provider should consider her specific risk factors, such as age, body mass index, accuracy of estimated gestational age, reproductive plans, personal values, and cultural context.”10

One fairly traditional picture of the place for patient preferences in medical decision making looks like this: It is the job of professional associations and individual providers to determine which medical options count as sufficiently safe, given the best current scientific evidence and the particular patient’s clinical profile. If more than one option is medically sound, then the patient may be offered a choice between these options, to be made on the basis of her values and preferences. Patients also may refuse medically sound interventions at their discretion, and debate may ensue over whether to let patients make some ‘bad’ choices in the name of patient autonomy. This is the framework implicit, for instance, in obstetrician-gynecologist Peter Schwartz’s analysis of the ethics of CDMR. He begins his case study commentary with a summary of the risks and benefits of scheduled cesarean sections, and then adds: “Physicians now give greater weight to the patient’s preferences, even when these run counter to the patient’s good.” He concludes that if faced with a request for CDMR, he would “strongly urge” the patient “to opt for vaginal birth. ... But if the patient were, say, thirty-five years old, having her first pregnancy, and anticipating no further pregnancies, I would feel that elective cesarean section is a reasonable alternative and would support her wishes.”11 Here, the physician determines the space of reasonable medical options in advance, and patient preferences then play a role within this frame.

But this model is deeply problematic. First of all, the concept of risk, as many have pointed out, is itself an inherently value-laden one: there is no truth of the matter about the size of a medical risk in abstraction from some set of values and preferences. For example, while it is clear that, faced with two medical treatment options, one of which carries a 50% risk of a headache the next day and the other of which carries a 3% risk of death, the second option would normally be viewed as riskier, even though it has a lower probability of a negative outcome. We have no trouble deciding that given our values and preferences, the 3% chance of death is a more pressing risk. But without our normative weighting of the relative ‘badness’ of headaches and death, there is literally no answer to the question of which option is riskier. Indeed, if risks were mere non-normative probabilities,12 they would be utterly nondirective for policy and decision-making. Thus any judgment concerning whether an option is ‘too risky’ to be medically sound is already encoding someone’s valuation of different outcomes. If we do not introduce patient values and preferences into the conversation before we make this judgment, then we are not making a ‘value-free’ assessment, but rather implicitly assuming that it is the values of the physician or the professional association that should be used to determine the size of various risks. Such a choice is not based in science, but rather represents an implicit or explicit choice to privilege one normative perspective over a variety of others.

Serious attention to patient autonomy would seem to require bringing patient values in earlier, when determining how risky various medical options are. The probability of various possible outcomes from vaginal or cesarean births such as future fertility, duration of recovery, or subsequent incontinence can be measured scientifically, but probabilities alone will not tell us how particular women value these outcomes. Moderate incontinence may be an unbearable humiliation or a debilitating incapacity for a professional dancer and an annoying but quite manageable inconvenience a woman who spends most of her time at her desk in a private office. Slow recovery may represent a crisis for a single mother with several other children at home and little or no paid maternity leave, or a relatively minor discomfort for a first-time mother with a supportive and available spouse. Thus the domain of birth and delivery decisions is one where the size of different risks will vary greatly depending on the meaning of the various outcomes for a particular patient, as these outcomes will be situated within the context of her larger narrative, needs, and projects, her family and community responsibilities and relationships, and her cultural values. In short, there is just no answer to how risky an approach to delivery is on the basis of statistics alone; women’s perspectives necessarily play a role in measuring risk in the first place.

A second point follows upon the first. Determining the size of relevant medical risks, even if done in appreciation of patients’ values, will still not determine how directive it is appropriate for providers to be, for medical risks are not the only considerations relevant to decisions about how to give birth. How much medical risk providers should be willing to regard as reasonable depends very much on the kind of decision in question, and in particular on how deeply the decision is bound up with intimate and integral features of a patient’s larger narrative, values, needs, and well being. For example, when choosing which antibiotic we should use to treat a bacterial infection, we normally ought to countenance a quite narrow range of reasonable risk; since the differences accompanying different antibiotics do not penetrate deeply into patients’ narratives and values, even a small difference in expected total risks and benefits suffices to warrant some quite directive counseling on the part of physicians.13 In contrast, when it comes to managing the course of palliative care, or addressing a severe facial disfigurement, we recognize that there will be a wider range of reasonable options. We understand that different patients may reasonably choose very different paths, including what might be viewed as relatively “risky” paths, because the choices they make about treatment have not just medical effects but wide-ranging impacts on patients’ personal narratives, their sense of identity and dignity, and on their social relationships and contexts. We do not deem a terminal cancer patient irrational for refusing an immediately debilitating treatment that may extend her life by a few months because she wants to be able to attend her grandson’s wedding next month. The more intimately and extensively a decision penetrates our lives, the more it will be attended by agent-relative, extra-medical costs and benefits that play a large role in determining the overall most reasonable course of action. When a patient is deciding how to die, how to give birth, or how to come to terms with a major disability, options that fail to minimize medical risks can still be the most rational choices, given other potential harms such as indignity, abandonment, or alienation, or benefits such as security, fidelity to loved ones, or a sense of comfort with one’s own narrative. Birth is a momentous and intimate event, and women may have a variety of reasons for rationally tolerating small increases in medical risk in order to avoid other harms or to gain other benefits.

Thus determining the range of appropriate delivery options for a particular patient requires that providers bring that patient’s voice and life circumstances into the conversation earlier than the traditional model suggests. Substantively incorporating a patient’s values into decision-making about cesarean delivery is not a matter of allowing a ‘subjective’ element into an otherwise rationalized, evidence-based set of practices. Since any assessment of the size of a risk is a normative judgment that relies on a set of values, whether explicitly or implicitly, guidelines and practices that do not interrogate these underlying values are prone to a number of distortions.14 Careful inclusion of patients’ perspectives in determining the riskiness and reasonability of different medical options will likely help correct for these distortions and add evidence-based objectivity to the process, rather than the reverse.

So far, our emphasis has been on the importance of women’s voices and values in determining what counts as a reasonable approach to delivery for an individual patient. This discussion does not yet tell us how best to protect and promote women’s autonomy in the context of delivery decisions. However, the NIH State-of-the-Science Conference summary and the ACOG Committee Opinion agree that there is insufficient evidence to fully evaluate the risks and benefits of cesarean delivery on maternal request versus planned vaginal delivery, and thus that current data do not provide a basis, except in narrowly defined circumstances, for recommending one mode over the other.15 In this light, the answer to this question might seem straightforward: since in most cases, basic safety concerns do not clearly tell in one direction or the other, why not just maximize women’s autonomy and the role of their values and perspectives by simply offering them a choice of approaches to delivery? On this view, not only should providers be able to accede to requests for cesarean deliveries, but they should actively promote women’s exercise of their autonomy by letting them know up front that they can choose among approaches to delivery, and that in most cases there is no clear medical reason why they should avoid or attempt a vaginal birth.

We believe that this strategy is more ethically complex than it appears, and that in many of its incarnations it would not in fact enhance women’s reproductive autonomy.

First, and crucially, the act of offering a choice does not merely transparently convey information. Rather, it is a substantive communicative act, and as such it can have important performative effects; in particular, it can shift values and preferences at the same time as it makes room for acting on these values and preferences.16 In the case of cesarean delivery absent medical indication, we do not know what effects would result from its being offered as an option for women who are uncomfortable with the prospect of attempting vaginal delivery. Surely this depends, to some extent, on the specific context and method of presenting and discussing these options. Given that pregnant women are likely to try hard to discern the ‘best’ delivery option, subtle differences in conversational presentation could have dramatic impact on women’s choices. Even if we assume that providers could manage to present a choice between attempted vaginal birth and scheduled cesarean section in an unbiased fashion - rather than letting their presentation be distorted by concerns about liability, insurance coverage, convenience, etc. - the presentation of this choice occurs within a specific social context, namely one in which medicine’s technologically intensive and predictable approaches to managing the body tend to be strongly valued and institutionally supported. It is easy to imagine that if the offer of cesarean delivery became more routine, women who had reasons for preferring a vaginal delivery might increasingly feel that their preferences were capricious at best, and illegitimate at worst. Or perhaps the offer would have different practical effects; our point is simply that such effects can be strong, and we do not want to blithely assume that by increasing the choices that are available to patients, we in fact increase their ability to make decisions that are properly attuned to their values and narratives. We need careful empirical studies that show what sorts of procedures and conversations will best enable women to make reasonable decisions about delivery that reflect their values and needs, without increasing their anxiety.

Second, there is a sizeable body of qualitative empirical literature demonstrating that often, patients do not particularly want to be offered more choices concerning their own medical care, nor do they experience such expanded choice as an enhancement of their autonomy.17 To the contrary, they may experience it as a form of abandonment, leaving them unable to function as committed agents in their own health care.18 For instance, Kristina Orfali found that parents felt like they had much more agential control over the care of their infants in the NICU when their physicians were more directive rather than offering more choices.19 Meanwhile, Vikki Entwistle and her colleagues found that diabetes patients judged their own autonomy and involvement with their decision-making more in terms of how much communication they had with their practitioners, and the extent to which they felt that their practitioners listened respectfully and attentively to them, than by how much choice over treatment paths they were offered.20

Pregnant women are already faced with an enormous amount of complicated risk data, and they are often expected to carefully regulate their bodies through the lens of risk when it comes to what they eat, what medications they take, what cosmetic products they use, how and when they exercise, what sort of prenatal care they receive, and much more.21 The Internet abounds with interactive ‘birth plans’ offering hundreds of choice points.22 It may well be that asking pregnant women to choose their delivery approach will only add to their already sizeable perceived responsibility for making stressful choices based on complicated and equivocal information, without particularly promoting better outcomes for mothers or babies. Furthermore, pregnant women are under intense cultural pressure to make whatever choice is safest for their baby. But in the case of mode of delivery, there are huge numbers of small risks and benefits that attend both vaginal and cesarean birth, and even the experts agree that for most women in developed countries, the data marks out no path as clearly safer.23 Hence women may find this new choice difficult - indeed, unproductively difficult - to make. It is thus plausible to think that at this point, many women would not experience an offer to choose between delivery options as an enhancement of their autonomy.

On the other hand, there could be important benefits to proactively offering more delivery choices to women. As we argued earlier, birth is a domain in which personal values are often strongly held and varied, and many factors that are not narrowly medical can play a large role in determining what women fear and seek during birth. This speaks in favor of delivery guidelines and clinical practices that make room for and even elicit women’s preferences. Furthermore, women from different socioeconomic, cultural, and educational backgrounds will likely vary in their access to information about delivery options, their confidence in their own ability to digest this information and make informed judgments about their preferences, and, importantly, their sense of entitlement when it comes to making requests of their providers. Hence if providers opt to discuss a delivery option only after a patient has spontaneously expressed an interest in pursuing it, they risk creating or intensifying systematic inequalities in women’s access to this option.

Collectively, these considerations suggest that enhancing women’s autonomy in birth is a matter neither of plainly prescriptive guidelines, which specify a “best” mode of delivery for women based solely on official medical indications, nor of following a “consumer model” in which women are simply offered a menu of options, along with litanies about their respective panoply of risks, and left to place their order. Women’s values and preferences should have an important role in determining approach to delivery, but their voices should be engaged in the course of a conversation designed to determine the best approach to delivery given their individual situation, all things considered, including medical factors such as fetal presentation and body mass index, narrative factors such as plans for future children and level of support at home during recovery, and personal values such as associating personal dignity and control with attempting or avoiding a vaginal birth. Such conversations, at a minimum, need to be sensitive to complicated and often contradictory cultural pressures such as the widespread association of technologically mediated birth with safe birth, pervasive rhetoric that treats unmedicated vaginal birth as a moral accomplishment and cesarean birth as a maternal failure, and a dominant ethic of maternal self-sacrifice. They also need to be sensitive to the fact that women differ in their preferences about the process of choice itself. Some women will value explicit and detailed presentation of options, others will feel burdened rather than liberated by the idea of detailing a “birth plan,” and still others, of course, simply don’t attach enormous significance to which orifice their baby exits, as long as it comes out safely (and may resent an implication that they should). Conversations about how to give birth - like conversations about how to die - are not governed by one-size-fits-all scripts, or routine check-off questions; instead, they ought to proceed via respectful probes and receptivity to emerging concerns, questions, and values, including receptivity to a desire not to continue the conversation at all.

We have seen that variations in personal and cultural values, narratives, and circumstances impact women’s preferences and needs. Different approaches to delivery - indeed, different approaches to framing risk, presenting options, and timing conversations - will be optimal for different women. But if we are to take seriously the project of giving women access to births that are both safe and expressive of their autonomy, we need to broaden our frame still more, and consider the range of structural constraints that may limit or foster such access. These are numerous, complex, and intertwined, and include the distribution, organization, and funding of health services, systematic social inequities, and power relationships between patients, providers, and health institutions. While it is not within the purview of this paper to recommend ways to correct unjust inequities and power imbalances, their existence has a substantial impact on when and how providers, policies, and guidelines can succeed in fostering women’s agency in the context of birth.

For example, the policies of different public and private insurers may not only limit which delivery option will be covered for women, but also substantially restrict the opportunities for discussion and dialogue between women and their caregivers. Regional disparities in health services may shape the conditions for autonomy, especially for women in communities without maternity services and whose options and preferences are structured by significant travel costs, the unpredictability introduced by distance from a given birth facility or provider, or separation from family or community. Women who are marginalized by poverty, recent immigration, insecure housing, drug and alcohol use, or domestic violence may lack the social authority needed to articulate their preferences within institutional contexts or in conversation with their providers. For example, a woman on social assistance whose claims to mothering have been called into question by the state may be particularly unwilling to assert her preferences for fear they will be judged improper, selfish, or ill-considered. Respectful, extended dialogues that situate birth within a personal narrative may be precluded for women who do not have access to continuous care and providers who have no opportunity for more than isolated and fleeting contact with particular patients. Women from some cultural or educational backgrounds may find a conversation that frames their medical care in terms of an array of risks and choices alienating or inappropriate.24 For all of these reasons, there can be no fixed formula that will tell providers what sorts of conversations and options they should provide to their patients in order to foster and protect their autonomy during birth. Women enter into clinical conversations from different social positions, and they may need different kinds of support in order to achieve autonomy. These differences among women are not necessarily manifested in the form of differences in their preferences concerning particular delivery options.

Throughout this paper, we have seen reasons to be unsatisfied with the model that associates autonomy with the right to make a free choice from an array of pre-digested options. As Entwistle and Watt put it, autonomy does not concern merely “the patient’s use of information to consider the selection of one treatment option from a well-defined set”; it also depends upon “what [patients] think and feel about their roles, efforts and contributions to decision-making and their relationships with their clinicians.”25 We have suggested, furthermore, that autonomy also depends on concrete structural conditions that enable or impede various forms of inclusion in the decision-making process. Feminist bioethicists have argued that autonomy requires that patients both trust those who participate in their decision-making and trust their own judgment and capacities,26 that they have the ability to critically interrogate their own values and preferences, as well as the expertise of others,27 and that autonomy is a relational feature of decisions and actions that are planted in the right way within a network of social relationships and narratives.28 The picture that emerges here is one in which autonomous decisions are those that result from respectful, contextually sensitive processes of joint decision-making, reflect the considered values, needs, and narratives of the person making the decision, and are such that the person trusts the process by which the decision was made and feels that her responsible agency has been exercised. Such decision-making could be compatible with quite directive guidance from providers and trusted others. For some women, appropriate, autonomy-promoting inclusion in this process may not consist in getting to “choose for oneself,” nor even of participating in an extended conversation about delivery possibilities; for other women it will. For all women, however, finding autonomy in birth requires access to safe modes of delivery that are appropriate to their personal and social circumstances and their values and sense of integrity and dignity, and circumstances that enable them to experience themselves as the primary (albeit relationally embedded) agents of their own birthing process.

We suggest that an empirical and conceptual exploration of how delivery decisions ought to be negotiated must be guided by a rich understanding of women’s agency and its placement within a complicated set of cultural meanings and pressures surrounding birth.29 We therefore believe that it is too early to be ‘for’ or ‘against’ women’s access to cesarean delivery in the absence of traditional medical indications - and indeed, that a simple pro- or con- position is never going to do justice to the subtlety of the issue. The right question is not whether women ought to be “allowed to choose” their delivery approach, but rather, taking the value of women’s autonomy in decision-making around birth as a given, what sorts of guidelines, practices, and conversations will best promote and protect women’s full inclusion in a safe and positive birth process.


This research has been supported by a Standard Research Grant from the Social Sciences and Humanities Research Council of Canada, Princeton University, and the University of Michigan Institute for Research on Women and Gender.


*The authors of this paper form the Obstetrics and Gynecology Risk Research Group. Our group writes collaboratively and we take turns serving as lead author. Our names are listed in alphabetical rotation, beginning with the lead author.

Contributor Information

Rebecca Kukla, Rebecca Kukla, PhD, is Professor of Philosophy and Obstetrics and Gynecology and Affiliated Professor of Women’s Studies at the University of South Florida, as well as a Senior Research Fellow at Queen’s University. She is the author of *Mass Hysteria: Medicine, Culture, and Mothers’ Bodies*, among other publications. She currently holds a grant from the Social Sciences and Humanities Research Council of Canada entitled “Autonomy and the Negotiation of Information in Reproductive Health Care.”.

Miriam Kuppermann, Miriam Kuppermann, PhD, MPH, is Professor of Obstetrics, Gynecology and Epidemiology at the Medical Effectiveness Research Center for Diverse Populations at the University of California San Francisco. Her research, most of which is funded by the National Institutes of Health, focuses on informed decision making among sociodemographically diverse women in the contexts of prenatal genetic testing, mode of delivery, management of noncancerous uterine conditions and prevention of cervical and breast cancer.

Margaret Little, Margaret Little, PhD, is Associate Professor, Philosophy, and Senior Research Scholar, Kennedy Institute of Ethics at Georgetown University. She researches broadly in ethics, from metaphysical questions to applied ones. She is co-founder of the Obstetrics and Gynecology Risk Research Group.

Anne Drapkin Lyerly, Dr. Anne Drapkin Lyerly, MD, MA, is an Associate Professor of Obstetrics and Gynecology and Core Faculty in the Trent Center for Bioethics, Humanities and History of Medicine at Duke University. Her research is supported by the Greenwall Foundation’s Faculty Scholars Program and a career development award from the National Institutes of Health. She is the 2007-2009 Chair of the Committee on Ethics for the American College of Obstetricians and Gynecologists. She is co-founder of the Obstetrics and Gynecology Risk Research Group.

Lisa M Mitchell, Lisa M. Mitchell. PhD, is Associate Professor in the Department of Anthropology at the University of Victoria. Her research interests address social and cultural dimensions of reproduction, including clinical and public meanings of fetal ultrasound imaging and other forms of prenatal screening, the ways in which risks in pregnancy are conceptualized and managed by women and by caregivers, and women’s experiences of pregnancy and pregnancy loss. She is the author of *Baby’s First Picture: Ultrasound and the Politics of Fetal Subjects* (University of Toronto Press, 2001).

Elizabeth M. Armstrong, Elizabeth Mitchell Armstrong, Ph.D., M.P.A., is Associate Professor of Sociology and Public Affairs at Princeton University. She has research interests in public health, the history and sociology of medicine, medical ethics, and the sociology of pregnancy and childbirth. She is the author of *Conceiving Risk, Bearing Responsibility: Fetal Alcohol Syndrome and the Diagnosis of Moral Disorder* (Johns Hopkins University Press, 2003), the first book to challenge the conventional wisdom about drinking during pregnancy.

Lisa Harris, Lisa H. Harris, MD, PhD, is an Assistant Professor in the Departments of Obstetrics and Gynecology and Women’s Studies at University of Michigan, and Director of University of Michigan’s Fellowship in Family Planning. In these various capacities she serves as a clinician, a teacher and a researcher. Her clinical work includes all aspects of general obstetrical and gynecological care, with a focus on miscarriage and family planning. In her research she focuses on interdisciplinary approaches to women’s reproductive bodies.


1. Rates in North America are currently between one quarter and one third of all deliveries. According to the National Center for Health Statistics, the 2006 cesarean rate in the United States was 31.1%, which represents a 50% increase over the past decade. [accessed 3/9/2008]. ( Armson BA. Is Planned Cesarean Childbirth a Safe Alternative? CMAJ. 2007:176. [PMC free article] [PubMed]
2. Declercq E, Menacker F, MacDorman M. Rise in ‘No Indicated Risk’ Primary Caesareans in the United States, 1991-2001: Cross Sectional analysis. British Medical Journal. 2005;330:71–2. [PubMed]MacDorman M, et al. Infant and Neonatal Mortality for Primary Cesarean and Vaginal Births to Women with ‘No Indicated Risk,’ United States, 1998-2001 Birth Cohorts. Birth. 2006;33:175–82. [PubMed]
3. A few celebrities, including, Brittany Spears and Posh Spice garnered a lot of publicity, most of it negative, for choosing to deliver by cesarean section. It would be unwise to infer anything from a few incidents of celebrity behavior to the practices of women in general.
4. National Institutes of Health State-of-the-Science Conference Statement on Cesarean Delivery on Maternal Request; Bethesda, MD: NIH. 2006; [accessed 7/20/2007]. [PubMed]
5. See for instance Minkoff H, Chervenak FA. Elective Primary Cesarean Delivery. NEJM. 2003;348:946–50. [PubMed] and Schwartz P. Is the patient always right? Hastings Center Report. 2004;34:13–4. [PubMed]
6. For example, see Katz Rothman B, Simonds W, Meltzer B. Laboring On: Birth in Transition in the United States. Routledge; New York: 2006. Bergeron V. The Ethics of Cesarean Section on Maternal Request: A Feminist Critique of the American College of Obstetricians and Gynecologists’ Position on Patient-Choice Surgery. Bioethics. 2007;21:478–87. [PubMed] and Wagner M. Born in the USA: How a Broken Maternity System Must Be Fixed to Put Women and Children First. University of California Press; Berkeley: 2006. among others.
7. [accessed 11/19/2007].
8. We use delivery approach rather than delivery mode, as it is not possible to choose to have a vaginal delivery; one can only plan a vaginal delivery, as there will inevitably be some women for whom an attempt at vaginal birth will end in a cesarean delivery.
9. Charles C, Gafni A, Whelan T. Shared Decision-Making in the Medical Encounter: What does it Mean? (Or it Takes at least Two to Tango) Social Science and Medicine. 1997;44:681–92. [PubMed]Emmanuel E, Emmanuel L. Four Models of the Physician-Patient Relationship. JAMA. 1992;267:323–9.
10. ACOG Committee Opinion Number 394: Cesarean Delivery on Maternal Request,” Obstetrics and Gynecology. 2007;110:1501–04. [PubMed]
12. And it is clear that they are not, if only from the fact that we never talk about the ‘risk’ of positive outcomes - i.e. the ‘risk’ of having a healthy baby.
13. Even here, however, there can be exceptions. Imagine that a patient must choose between a more effective antibiotic to be taken every six hours and a somewhat less effective antibiotic to be taken once a day. Normally, swallowing a pill is a minor enough affair that it would make sense for a physician to be fairly directive about prescribing the more effective medication. But if the patient has mild dementia and his son, who lives in a different neighborhood, will need to supervise each dose, then it may well be reasonable to choose the less effective medication and thereby make the daily burden more manageable and autonomy-preserving for both father and son.
14. Lyerly AD, Mitchell LM, Armstrong EM, Harris LH, Kukla R, Kuppermann M, Little MO. Risks, Values, and Decision Making Surrounding Pregnancy. Obstetrics and Gynecology. 2007;109:979–84. [PubMed]
15. Both NIH and ACOG do recommend against CDMR prior to 39 weeks’ gestation or without verification of fetal lung maturity, in cases where women desire several children, and when motivated by the unavailability of anesthesia.
16. Lichtenstein S, Slovic P. The Construction of Preference. Cambridge University Press; New York: 2006. Caterer C, Loewenstein G, Rabin M. Advances in Behavioral Economics. Princeton University Press; Princeton, NJ: 2003. Hahnemann D, Tversky A. Choices, Values, and Frames. Cambridge University Press; New York: 2000. Schwartz B. The Paradox of Choice: Why More is Less. Harper; New York: 2005.
17. Orfali K. Parental Role in Medical Decision-Making: Fact or Fiction? A Comparative Study of Ethical Dilemmas in French and American Neonatal Intensive Care Units. Social Science and Medicine. 2004;58:2009–22. [PubMed]Skea Z, et al. Women’s Perceptions of Decision-Making about Hysterectomy. British Journal of Obstetrics and Gynecology. 2004;111:133–42. [PubMed]Entwistle V, et al. Which Surgical Decisions should Patients Participate in and How? Reflections on Women’s Recollections of Discussions about Variants of Hysterectomy. Social Science and Medicine. 2006;62:499–509. [PubMed]Schneider C. The Practice of Autonomy: Patients, Doctors, and Medical Decisions. Oxford University Press; New York: 1998.
18. Kukla R. Conscientious Autonomy: Displacing Decisions in Health Care. Hastings Center Report. 2005;35:34–44. [PubMed]McLeod C. Self-Trust and Reproductive Autonomy. MIT Press; Cambridge, MA: 2002.
19. Orfali 2005.
20. Entwistle V, et al. Involvement in Treatment Decision-Making: Its Meaning to People with Diabetes and Implications for Conceptualization. Social Science and Medicine. online preprint, available October 23, 2007. [PubMed]
21. Duden B. Disembodying Women: Perspectives on Pregnancy and the Unborn. Harvard University Press; Cambridge, MA: 1993. Armstrong EM. Conceiving Risk, Bearing Responsibility: Fetal Alcohol Syndrome and the Diagnosis of Moral Disorder. Johns Hopkins University Press; Baltimore: 2003. Kukla R. Mass Hysteria: Medicine, Culture, and Mothers’ Bodies. Rowman and Littlefield; Lanham, MD: 2005.
22. For instance see [accessed December 1, 2007].
23. As an anonymous reviewer correctly pointed out, the safety of both vaginal births and cesarean sections will depend on the facilities in which they occur, the skills of the particular birth attendant, the health status of the mother, and so forth.
24. See for instance Duden 1993.
25. Entwistle VA, Watt IS. Patient Involvement in Treatment Decision-Making: The Case for a Broader Conceptual Framework. Patient Education and Counseling. 2006;63:268–278. [PubMed]
26. O’Neill O. Autonomy and Trust in Bioethics. Oxford University Press; New York: 2002. McLeod 2002.
27. Kukla 2005; Kukla R. How Do Patients Know? Hastings Center Report. 2007;37:27–35. [PubMed]
28. Donchin A. Autonomy and Interdependence: Quandaries in Genetic Decision Making. and McLeod C, Sherwin S. Relational Autonomy, Self-trust, and Health Care for Patients who are Oppressed,” In: Mackenzie C, Stoljar N, editors. Relational Autonomy, Feminist Perspectives on Autonomy, Agency and the Social Self. Oxford University Press; New York: 2000. both. Lindemann Nelson H, Lindemann Nelson J. The Patient in the Family. Routledge; New York: 1995.
29. The Obstetrics and Gynecology Risk Research Group consists of two philosophers, two obstetricians (one of whom also has a doctorate in history), a sociologist, an anthropologist, and an epidemiologist/decision analyst. We are beginning to undertake the interdisciplinary research project described here.