Women’s reproductive autonomy: medicalisation and beyond
- Correspondence to: L Purdy Wells College, 75 Turkey Hill Rd, Ithaca, NY 14850, USA;
- Received 29 June 2004
- Accepted 25 July 2005
Reproductive autonomy is central to women’s welfare both because childbearing takes place in women’s bodies and because they are generally expected to take primary responsibility for child rearing. In 2005, the factors that influence their autonomy most strongly are poverty and belief systems that devalue such autonomy. Unfortunately, such autonomy is a low priority for most societies, or is anathema to their belief systems altogether. This situation is doubly sad because women’s reproductive autonomy is intrinsically valuable for women and also instrumentally valuable for the welfare of humankind. This paper takes for granted the moral and practical necessity of such autonomy and digs deeper into the question of what such a commitment might entail, focusing on the mid-level policy making that, at least in the US and Canada, plays a significant role in shaping women’s options. This paper examines a large teaching hospital’s policy on reduction of multifetal pregnancies. The policy permits reduction of triplets to twins, but not twins to a singleton. As there is no morally relevant difference between these two types of reduction, it is evident that inappropriate medicalisation can still limit women’s autonomy in undesirable ways.
Nothing would advance women’s welfare more than respecting their reproductive autonomy. This statement presupposes autonomy’s prerequisites, such as decent health care, education, and alternative ways of supporting themselves. By reproductive autonomy, I mean the power to decide when, if at all, to have children; also, many—but not all—of the choices relevant to reproduction. I focus here on decisions about whether and when to have children. Women should also generally determine how their pregnancy will be carried out and how the birth will happen. New technologies are, however, continually raising new questions, and reproduction both requires and affects others (children, men, society at large); many issues therefore must be examined on a case by case basis. Reproductive autonomy thus has much in common with Robertson’s notion of procreative liberty, but is not identical with it.1 The desirability of women’s reproductive autonomy is in part derived from the more general benefits of reproductive autonomy recognised by many writers in the liberal tradition. (See—for example—work by Joel Feinberg, John Robertson, and John Harris.) Such autonomy is particularly important for women, however, because reproduction still takes place in women’s bodies, and because they are generally expected to take primary responsibility for child rearing. The need to locate women’s autonomy within this broader liberal context has led to a critical rethinking of the concept of autonomy, including a need to focus on options excluded from those among which subjects may choose.2
In 2005, the factors that influence women’s reproductive autonomy most strongly are poverty, and belief systems that devalue such autonomy. Ensuring that every woman had the prerequisites for practising basic reproductive autonomy would take only a fraction of the world’s resources: but that autonomy is a low priority for most societies, or is anathema to their belief systems altogether. So poverty and anti-autonomy belief systems work together to deny women control over their lives. Although lack of access to the prerequisites for exercising autonomy is often a result of anti-autonomy belief systems, it can also be a consequence of racism or limitless greed.
Belief systems that devalue women’s reproductive autonomy are widespread. They are more or less explicit, and are based on a variety of religious and philosophical ideas. Most influential in Western societies are probably the biblical sources that depict women’s origin in Adam’s rib as a mark of their subservient nature. Also hugely influential is the Aristotelian elaboration of that nature as lacking elements of rationality, particularly those elements that legitimise individual purposes (as opposed to group function). There is, however, no shortage of other belief systems alleged to justify the subordination of women’s autonomy—where such justification is thought to be needed at all.
This situation is doubly sad because women’s reproductive autonomy is not only intrinsically valuable for women, but also instrumentally valuable for the welfare of all humankind. This point was finally recognised internationally at the 1994 population conference in Cairo where it was reasserted that only by providing women with the prerequisites for autonomy, including the prospect of security and fulfilment with few or no children, is there any hope of meeting basic human needs.3 The conference concluded that investment in health, education, and women’s empowerment is necessary to reduce the birthrate. Naturally, population issues are complicated, and a decent life for all will never be possible without much greater political and economic equality. However, not only are such changes not on the horizon in 2005, but population growth must still slow if human needs are to be met sustainably. Such a slowing of population growth would also free up women’s energy to tackle the monumental social and political problems facing humankind.
Moreover, the overall turn to the right in the last twenty years of the twentieth century is seriously eroding previous progress toward women’s reproductive autonomy. In the US members of that right wing movement adopted the traditional social views of the religious segment of society they courted. The history is somewhat different in each country, particularly Muslim countries, although each appears to have had similar outcomes with respect to women’s reproductive autonomy. Most disturbingly, these various movements to the right of the political spectrum have once again put in question the desirability of both women’s autonomy and the more general appreciation of reproductive autonomy with which it is allied. This is not to say that such reproductive autonomy was previously universally accepted, but it was (in the US and some other countries) implicit in the reigning liberal paradigm and the activist judicial elaboration of that paradigm.
This paper does not attempt to update the arguments for women’s reproductive autonomy: it takes for granted the moral and practical necessity of such autonomy, and digs deeper into the question of what such a commitment might entail.
The liberal paradigm was theoretically committed to women’s reproductive autonomy. But to many feminists, that commitment looked shallow and all too readily forgotten when conflicting interests arose. See for example, John Robertson’s work.1 Despite his allegiance to what many see as an extreme position, procreative liberty, and despite a much more woman friendly approach than many writers in reproductive ethics, he compromises women’s interests at crucial junctures. Feminists—those who seek justice for women—have most reliably sought to protect and expand their reproductive autonomy, and a large and excellent feminist literature on reproduction, including its politics, and its legal and sociological contexts now exists. For a more detailed discussion of the concept of feminism see my paper in Health Care Analysis.4
Nevertheless, there is further ground to be covered. Sweeping legal decisions (both legislation and court judgments), and the day to day interactions between patients and healthcare providers have received substantial popular and scholarly scrutiny. However, somewhat less attention has been directed toward the mid level policy making that, at least in the US and Canada, plays a significant role in shaping women’s options. Other critically important dimensions of reproductive autonomy are also in need of further work. For example, for an eye opening treatment of the way race alters the politics of reproductive autonomy see the book by Dorothy Roberts.5 Also, I believe that even feminists have failed to focus sufficiently on the pronatalism and other cultural factors that can lead women to unwittingly make reproductive decisions that may not be in their own interest. The influence of other critically important elements of the overall context, such as mid-level policies, is especially noticeable given the relatively liberal constitutional frameworks within which abortion decisions are to be made in those countries. In both the US and Canada—for example, policies established by professional societies and individual hospitals greatly affect women’s reproductive autonomy. Consider how abortion access, as defined by Roe (and subsequent Supreme Court cases), has been sculpted by such policies. Roe announced that in the first trimester, a decision to abort was between a woman and her doctor. Although feminists have quite rightly objected to the medical paternalism implicit in this standard, for some women this has meant in practice that abortion is available to them on demand, without further restrictions imposed by states. However, subsequent Supreme Court decisions opened the way for numerous restrictions (such as required counselling, twenty four hour waiting periods, and special requirements for minors seeking abortions) imposed by particular states. Later in pregnancy, states could regulate abortion with an eye to women’s health (in the second trimester) or out of respect for fetal life (in the third trimester). Yet most US hospitals developed policies prohibiting abortions after a given number of weeks, some do not offer abortion at any stage of pregnancy, and many training programmes have deleted abortion from ob/gyn requirements or do not offer such training at all. So abortions are now unavailable in most US counties. Violence against abortion clinics has also reduced the number of practitioners willing to provide abortions. Likewise, in Canada, although abortion has been decriminalised for some fifteen years, professional societies and individual hospitals have stepped into what they saw as a regulatory vacuum to restrict women’s access to abortion. Inappropriate medicalisation can be a major factor in subtly altering (or even erasing altogether) women’s voices. Such medicalisation has been a cornerstone of feminist critiques of health care. At the core of these critiques is the claim that it reduces “political, personal, and social issues to medical problems, thereby giving scientific experts the power to ‘solve’ them within the constraints of medical practice”.6 “Medicalisation” has sometimes been construed as requiring that women (and other common targets of medicalisation, such as gay men) refrain from attempting to get medical help for their bodily conditions: but that conclusion begs the question whether recourse to medicine necessarily leads to harmful loss of control.7,8,9,10 Another central question is whether this loss of control is embedded in the concept of medicalisation so that it cannot be used to refer to properly autonomous heath care. At least some feminists believe it must be possible to develop health services that are respectful of women’s interests and choices.7 The difficulty is in noticing where women’s say is insidiously appropriated (whether intentionally or unintentionally), and then showing why (and how) to reframe issues so as to restore their autonomy.
Unfortunately, women’s loss of control is so longstanding and so central to so many cultures that recognising it can be difficult. Almost everyone—for example, now agrees that involuntary sterilisation policies that have targeted disadvantaged girls and women were morally impermissible. But less obviously coercive policies adopted by physicians that limited access to sterilisation for more privileged women were equally unjustifiable. Women seeking sterilisation were subjected to various limits based on their age and number of children. Who were physicians to decide how women should live their lives? Investigating the possible answers begins to lay bare the illicit connections between conceptions of women’s nature and health.i
No doubt there are dozens or even hundreds of similar examples to which historians could point.
People—even feminists—now tend to take it for granted that (where reproductive autonomy is the professed goal) policies have been cleaned up so as to advance that goal. It is therefore something of a shock to uncover new pockets of medicalisation, such as the pregnancy reduction policy I was asked to evaluate as the bioethicist on the ob/gyn ethics committee at a large teaching hospital in 1998. Despite the hospital’s overall goal of patient autonomy (a goal enshrined in provincial law), women had no choice at all. In fact, the policy incorporated multiple, often inconsistent strands, like a twisted rock face where a region’s geological history is displayed.
The 1989 policy (reaffirmed by another committee in 1993) stated that although triplets could be reduced to twins, twins would not be reduced to singletons. Interestingly, the policy was introduced about the time that, in the wake of the state’s inability to convict Henry Morgenthaler for performing abortions, abortion became legally unregulated in Canada. Thus abortion was no longer a criminal offence. In the United States, although various legislative and judicial bodies had been eating away at Roe v Wade almost since its inception in 1973, women’s negative right to first trimester abortions was also legally protected. That is, it was in principle illegal to interfere with a woman’s right to choose to abort, although there was no positive duty to provide her with the material resources necessary to carry out her decision. Furthermore, there were in 1989 strong pro-choice movements and highly articulate defences of that position in the philosophical literature and elsewhere.
At issue was a narrow question: is there a morally relevant difference between reducing a pregnancy from triplets to twins, and reducing twins to a singleton? The answer to this question is clear: the motive, method, and consequences of the acts are the same; nor does a twin fetus suddenly sprout rights lacking in a triplet fetus. Given this point, it is contradictory to reject reduction to a singleton if reduction to twins is offered. However, both the existence of the 1989 policy and the reception of my analysis and conclusion revealed how easily decision making in the clinical environment can be detached from its philosophical and political roots. Indeed, the argumentation about the procedure was so disconnected from abortion issues that reduction did not seem to be about abortion at all, yet both hospital policy and the objections to my position were shot through with unstated assumptions about abortion.
THE MEDICAL CONTEXT
Ovulation induction and in vitro fertilisation, fertility treatments that have come to be widely used in developed nations since the 1970s, have greatly increased the incidence of multiple pregnancy. Ovarian stimulation may cause many eggs to ripen simultaneously; if a woman then has intercourse, several may be fertilised. This is what happened in the recent widely publicised cases of septuplets and octuplets. Attempts to increase the success rate of IVF may also lead physicians to place many embryos in a woman’s uterus.ii
Women are designed, however, to carry one fetus at a time, and the more fetuses, the more risky the pregnancy, both for the woman herself and for the fetuses. Where there is more than one fetus women are more likely to develop serious health problems; they are also more likely to lose the entire pregnancy. The more fetuses she carries, the greater the risk that some or all of the babies will die, or will suffer from serious disability.
Recent studies comparing the number of fetuses with pregnancy outcomes show that the best outcomes are for singleton pregnancies. According to some indicators, twin pregnancies have somewhat worse outcomes, and triplets still worse ones. Thus it is slightly riskier to be born a twin than a singleton, and still riskier to be born a triplet. The outcomes get rapidly worse with quadruplets and beyond.
Reduction—aborting one or more of the fetuses—was introduced as a way to mitigate the consequences of such higher order multiple pregnancies (so called “supertwins”), reducing risk. However, reduction itself carries about an eight per cent chance of losing the whole pregnancy, although it appears that this figure is somewhat unreliable because of small samples and lack of information about the “natural” rate of pregnancy loss in these circumstances.
The main impetus for reduction seems to have been to diminish the risks inherent in pregnancies involving more than four fetuses. Yet the statistics also show some benefit in reducing three to two, and further benefit in reducing two to one. The risk reduction at each step is small, but noticeable. The statistics here vary with the study consulted. Recent studies underline the increased risk, both to women and their offspring. For instance: “maternal mortality is sevenfold greater in multiple pregnancies than in singletons, perinatal mortality rates are fourfold higher for twins and sixfold higher for triplets”.12 Another article states that: “in France, between 1986 and 1998, triplets represented 5.6% of all the IVF babies, but accounted for 30% of the high prematurity (<33 weeks), 11% of SGA, and 15% of the perinatal mortality. For twins, the rates were respectively 37, 52, 55, and 54%.”13 Regardless of the statistics, my central point holds: moral decisions of this sort cannot be deduced from any set of statistics.
Given this medical context, one might well be curious about the reasoning that led to the current policy. The policy seemed to focus almost entirely on risk: the risks of having triplets were regarded as sufficient to justify the risk of reduction, but not the risks of having twins.
However, this use of statistics hides a raft of important issues. First, as with all risk issues, the data do not tell you what to do. One might argue either way here:
a) the risk of twins is less than with triplets, so reduction is no longer justifiable on medical grounds, or
b) if the risk of twins is greater than for a singleton, then reduction is justifiable.
Deciding which direction to go requires further argument.
Secondly, the undiscriminating use of the word “risk” obscures more than it informs here. On the one hand it masks quite different kinds of risk—with quite different burdens—as well as differences in who bear them. Thus subtracting the risk of pregnancy loss from the risk inherent in a given multiple does not make sense, even though it may be treated as a simple piece of mathematics. The moral and emotional issues raised by losing all your fetuses are quite different from those raised by having one or more children with serious disability. Also, taking risks with your own health is different from risking the life or health of your fetuses or children.
On the other hand, and still more importantly, we live in a pluralistic world, and different players will evaluate these outcomes quite differently. Some women—even women who are undergoing fertility treatments—may prefer to risk losing a particular pregnancy rather than to risk the welfare of the fetuses that would be born; others may judge that it is preferable to try to protect the lives of all their fetuses, even if some or all might be disabled. Likewise, some women whose own health is more at risk from a multiple pregnancy might prefer risking the lives of the fetuses rather than their own health, although other women might prefer the reverse. The point here is that it should be up to women—not doctors or hospitals—to choose which possible outcomes they prefer.
It should also be noticed that if this were any other procedure, the case would need no argument. In the US and Canada, paternalism was officially rejected years ago, replaced with a contractual model of the physician/patient relationship requiring physicians to lay out for patients the possible consequences of various alternative treatments, and to let patients decide what to do, based on their own values. In the absence of the kind of statutory requirements common elsewhere in the world, this reduction policy appears to be left over from the 1930s, not 1993. This paradigm of informed consent also increases the burden of proof on any statutory regulation that limits women’s reproductive liberty—or any regulation that limits patient choice.
How could such a policy escape notice for so long? The answer seems to be obvious: unacknowledged, but very much present, is the spectre of abortion politics. The premise underlying this particular document is clearly a “moderate” position on abortion.
Why a moderate position? Because the policy accepts abortion in some cases (triplets to twins) but holds that abortion cannot be justified in others where the case is judged less compelling (twins to singleton). Making such distinctions is the hallmark of moderate positions. A conservative position would deny, and a radical position would accept, reduction no matter the number of fetuses. The arguments about risk purport to provide moral justification for the distinction made here between permissible and impermissible reductions.
Because the abortion premise is unstated, however, the argumentation and conclusion appear to be purely medical matters, the right course to be determined by risk computation. If the premise were stated, however, reduction would be set into its proper context—namely, the abortion context. That context would have raised the question of women’s autonomy, which was never mentioned in the document.
ADDITIONAL CLINICAL CONSIDERATIONS
The medical literature on reduction, and the policy alluded to, raise three more general issues that might bear on policy decisions of this sort: resource allocation, responsibility, and women’s emotional wellbeing.
Resource allocation loomed large in our ob/gyn ethics committee’s discussion of hospital policy. In Ontario, public funding of IVF was then available only on a limited basis, and there were long waiting lines. Some members of the committee argued that where a woman had used substantial social resources getting pregnant, it would be unjust to offer reduction for twins because if they lost the pregnancy and returned for further fertility treatments, treatment would be still further delayed for those waiting in line. Since success rates drop as women age, some on the list might thus be deprived of any hope of a genetically related baby.
As an adherent of quite an egalitarian notion of justice, I would agree that, other things being equal, a woman who foolishly endangers a pregnancy achieved under such conditions might reasonably be denied further fertility services. The rub, of course, is the definition of “foolish”. Preferring a singleton to twins is not foolish, however, as I will argue shortly.
Moreover, it is interesting to compare this objection to the reduction of twins with policies then in force in other departments of the same hospital. When transplant fails—for example, in some circumstances, the transplant programme does second, third, and sometimes even fourth retransplants, even where doing so means that others on the waiting list will undoubtedly die. In short, once a relationship is established with a given patient, he or she can be given priority even if others suffer as a result. Whether one thinks this transplant policy is justifiable or not—and there are good reasons for doubting that it is—this divergence is psychologically interesting. One might well wonder whether sexist assumptions play any role here, given that IVF patients are all women.
In addition, no one raised the fact that twins are likely to use more medical resources than singletons, and that, with respect to some indicators, the risks for twins are closer to the risks for triplets than to the risks for a singleton. Surely those resources are as relevant to the discussion as the resources required for fertility services. Perhaps the compartmentalisation of hospital services hid that point from the members of our committee.
Another objection to providing reduction of twins was that patients know infertility treatment might lead to a twin pregnancy, and so they should accept twins. This is a new twist on the old rejection of abortion based on the view that if women have sex so too should they be prepared for babies. However, society applies the more general principle upon which this objection is based—that it can never be morally appropriate to attempt to prevent undesirable consequences of actions—very inconsistently: other values such as concern for welfare determine where it does not hold. In any case, this position proves too much as it would also justify refusing to reduce any multiple pregnancy.
Last but not least, there was resistance to reduction based on beliefs about women’s emotional wellbeing. There is evidence that women find reduction emotionally painful, which is not surprising given that they are in fertility programmes to become pregnant. A study by Berkowitz et al shows that more than 65% of such women recalled acute feelings of emotional pain, stress, and fear during the procedure; 70% mourned for the lost fetuses, and their grieving lasted on average 3.2 months. Furthermore, 37% of the patients experienced an anniversary grief reaction, and although persistent depressive symptoms were generally mild, 17.6% reported lingering moderately severe feelings of guilt and sadness, and moderate levels of anger. Despite these mixed reactions, however, 93% said they would undertake the procedure again.14 These statistics underline the importance of counselling before and after any such procedure: but using them to deny women reduction would be seriously paternalistic. It also ignores the possibility that women who are refused reduction may have their lives significantly altered for the worse in ways they would not have chosen if given the opportunity.
SETTING REDUCTION IN THE ABORTION CONTEXT
Analysing the medical objections to reduction of twins thus shows that they are by no means the merely technical calculations that appear to lead so inexorably to a policy that precludes choice on the part of women.
I have hinted at some of the reasons why women should not be denied reduction of twins. First, the disaggregation of “risk” reveals the necessity for making choices involving legitimate differences in values. Second, the autonomy model of informed consent requires women to have choices about their treatment and precludes healthcare providers from making paternalistic decisions about their welfare. Thirdly, sexist assumptions may lead to unjustifiable inconsistencies in hospital policies about resource allocation that disadvantage infertile women.
Why has decision making about reduction tended to ignore these issues? I believe that the answer is, at least in part, because the medical environment focuses on pregnancy as a medical condition that is relieved by birth. But pregnancy and birth are not just—or even primarily—a medical matter: they are about shaping a life and creating a family.
I suspect that most people would concede that being presented with three new babies at a time would stretch the financial and personal resources of almost any family. The image of two new babies is, however, far less likely to trigger this kind of sympathy or understanding. After all, the average family with children has at least two of them, so the financial and personal burdens of twins might seem less daunting. Perhaps it is a bit much for both to come at once, but that is surely a relatively trivial difference.
Is it so trivial? For women in less than optimal circumstances, it certainly is not. Many families are living on the financial edge, and statistics suggest that the double shift for women is still common as many men are still not sharing domestic chores equally. Even for women in optimal circumstances, having twins rather than a single child (perhaps followed a few years later by another) can vastly change options. Many of the most desirable jobs are still designed for those with undemanding family commitments, and that second baby might deprive a woman of a position she could do brilliantly with just one baby, but not two. It is hardly up to third parties to rule on the importance of such considerations.
Reproductive autonomy takes these answers seriously, and they need to be reflected in relevant policy. iii
So we need to be much more attentive to the way policies such as the one I describe here can stealthily limit women’s autonomy, without making the kind of waves more public limits do.
Naturally, this work is necessary, but not sufficient, to ensure women’s reproductive autonomy. Removing external limits on decision making simply creates opportunities for genuine autonomy. Far more work needs to be done to help women surmount the internalised constraints on their choices arising from such cultural factors as pronatalism, geneticism, and sexism. Some informed consent theorists are grappling with those issues, but they must be left for another day.
↵ii For discussion of these issues contact the author to see my unpublished paper, Could there be a right not to be born an octuplet.
↵iii As it turned out, this particular policy revision became moot when the relevant services were moved to another hospital. I have no information about what became of it there.