Pregnancy Advance Directives | Cardozo Law Review

It does not help that our future selves are strangers to us. Most people don’t know what they’ll want for dinner next Tuesday, so how could they know exactly what they will want in the next decade?1

Introduction

Recent decades have seen an increase in legal protections for patient autonomy in matters of death and dying. In the 1970s and 1980s, states began to recognize the right of competent patients to terminate life-sustaining measures and to develop procedures for addressing end-of-life disputes involving incompetent patients.2 In 1990, the Supreme Court acknowledged that “[t]he principle that a competent person has a constitutionally protected liberty interest in refusing unwanted medical treatment may be inferred” from prior case law, and upheld Missouri’s procedures for making such decisions on behalf of incompetent individuals.3 Nearly all states permit patients to execute some form of advance directive to set forth their wishes for future treatment or to designate an individual to make medical decisions if the patient can no longer do so.4 At the federal level, the Patient Self-Determination Act of 1990 (PSDA) encourages patients to create advance directives by integrating educational and recordkeeping requirements into the Medicare and Medicaid programs.542 U.S.C. § 1395cc(f). Indeed, in a number of states, the protection of autonomy at the end of life has expanded to include the explicit right to choose physician aid in dying.6

Yet there is one group that has not benefitted equally from this expansion: pregnant patients. At least thirty states currently restrict the choice to withhold or withdraw life-sustaining treatments from pregnant persons who lack decisional capacity, not only by invalidating their prior advance directives, but also by prohibiting their surrogate decisionmakers and physicians from making such choices on their behalf. To make matters worse, in many states those limitations may not be clear to patients when they write those directives unless they carefully read the statutes.7

These “pregnancy restrictions” have been defended on two distinct grounds: first, that a patient who creates a directive when not pregnant might not fully contemplate how their wishes might change in the case of pregnancy; and second, the pronatalist principle that health care decisions during pregnancy should be guided by the goal of saving the life of the fetus, if at all possible.8 These explanations, however, have not been accompanied by actions designed to meet their goals: neither educational efforts to improve the quality of advance decision-making for patients of childbearing age, nor resources devoted to prenatal care that might improve fetal outcomes.9See infra Section II.B.2. Instead, these statutes simply preclude anyone—patient, surrogate, or physician—from refusing or withdrawing life-sustaining care during pregnancy. The effect is not only to deny individuals the exercise of their autonomy, but also to raise the grisly specter of incapacitated patients being reduced to incubators solely for the benefit of their fetuses.

Yet pregnancy is not the only medical condition that raises concerns about a competent patient’s present ability to predict care preferences during future incompetency. Newly diagnosed patients with dementia, particularly Alzheimer’s disease, face analogous concerns: while they currently may be competent, they will lose that capacity in the future due to the relentlessly progressive nature of the illness.10 At the time critical decisions will need to be made, the patient most likely will no longer be capable of doing so. Faced with these concerns, advocates have adapted psychiatric and mental health advance directives to focus on the types of decisions that often must be made as Alzheimer’s progresses, including instructions about care preferences, medical treatments, and financial planning.11 An important aspect of these directives is that they are completed after the diagnosis, when “the patient . . . has experience with the disease” and therefore is better able to contemplate future decisions that previously might have been unfathomable.12 This Article considers the need for a similar Pregnancy Advance Directive to safeguard the interests of a pregnant patient who faces life-threatening circumstances but has lost decisional capacity.

Part I of this Article provides a general introduction to the various types of advance directives available in the United States, including their goals and limitations. Part II provides a detailed overview of pregnancy restrictions, including comparisons of the substantive restrictions, procedural issues, and rationales for restricting the application of advance directives during pregnancy. Part III offers a critical analysis of both the scholarship addressing pregnancy restrictions and the litigation seeking to challenge the restrictions, demonstrating that the existing legal framework has not been satisfactory in resolving the issues—a situation that will only be exacerbated by the Supreme Court’s recent decision to overrule Roe v. Wade.13

Part IV offers an alternative. Taking a cue from those who advocate for the creation of special advance directives for early-stage Alzheimer’s patients, it is time to consider the creation of a Pregnancy Advance Directive: a targeted medical form addressing a patient’s wishes in the case of decisional incapacity during pregnancy, which could be completed only after the patient has become pregnant.14 Although it would not answer those critics who prioritize the interests of the fetus above all, it would address the concerns of those who fear prior directives may no longer reflect the new circumstances of pregnancy. It would also have the salutary effect of encouraging physicians to discuss these issues with their pregnant patients, leading to deeper consideration of the concerns and establishing a more detailed record in the event such a difficult decision must be made.15

A note regarding terminology. Throughout this Article, I will use the terms “pregnant women” and “pregnant patients” to refer to the pregnant individuals who are affected by these statutes. I do not do so lightly, mindful that not all pregnant persons identify as “women.”16 I use this terminology for two reasons. First and foremost, many of the statutes use gendered language, such as “a woman who is pregnant.”17 Even states that have repealed these restrictions have adopted gendered rationales: when Hawaii repealed its pregnancy restriction in 2000, for example, the legislature explained that “[a] woman should have the right to predetermine her medical treatment, including treatment during her pregnancy.”18 Using this terminology is an accurate, albeit unsatisfactory, reflection of the way in which the legal system historically has addressed pregnancy. Moreover, much of the scholarly literature has relied extensively on feminist legal theory as an analytical framework, situating the issue of pregnancy restrictions within the broader context of subordination and the devaluation of women’s autonomy.19 Adopting similar terminology is the most expedient way to engage with both the legal analysis and this scholarship. But I remain mindful of the limitations of this approach,20 and emphasize that future legislation must be written in an inclusive manner.

I. Advance Directives

The United States has a robust, albeit incomplete, history of deferring to the medical wishes of competent patients under the rubric of “informed consent.” In an oft-quoted opinion, the New York Court of Appeals noted in 1914 that “[e]very human being of adult years and sound mind has a right to determine what shall be done with his own body,” holding that surgery performed in the absence of such consent constituted an assault.21 As a corollary, “the common-law doctrine of informed consent is viewed as generally encompassing the right of a competent individual to refuse medical treatment.”22 Whether such refusal encompassed the right to decline or withdraw life-sustaining care, rather than less drastic medical interventions, remained unclear for many years. In the absence of legislative guidance, physicians feared prosecution for homicide if they withdrew life-sustaining measures.23 A series of cases in the 1970s and 1980s—including high-profile litigation involving Karen Ann Quinlan and Elizabeth Bouvia—established legal protections both for patients who wanted to exercise such rights and for the medical professionals who acceded to the requests.24 By 1990, the Supreme Court in Cruzan v. Director, Missouri Department of Health acknowledged the common law history and assumed, but did not conclusively hold, that the “Constitution would grant a competent person a constitutionally protected right to refuse lifesaving [treatment, including] hydration and [artificial] nutrition.”25

While these developments certainly were not without controversy, the far more vexing question proved to be whether incompetent patients retained similar rights of medical self-determination—and, if so, how those rights could be exercised. As Cruzan noted, “[a]n incompetent person is not able to make an informed and voluntary choice to exercise a hypothetical right to refuse treatment or any other right. Such a ‘right’ must be exercised for her, if at all, by some sort of surrogate.”26Cruzan, 497 U.S. at 280. Who were those surrogates? The presumption was that most would be family members, although patients with longstanding incompetency might be subject to legal guardianship.27 A skeptical Cruzan majority, however, noted that “[n]ot all incompetent patients will have loved ones available to serve as surrogate decisionmakers,” and that perhaps not all family members could be trusted.28Cruzan, 497 U.S. at 281. Moreover, in the absence of a statute, even family members required court appointment to this role;29 in controversial cases, that process might require multi-year litigation, adding to the trauma already experienced by the family.

Even more confusing was the question of how the surrogate should make decisions for the patient. As Lois Shepherd explains:

Generally, respect for the autonomy of patients who previously had, but have now lost, capacity is understood as respect for their prior decisions—whether these were fully formed expressed preferences . . . or were instead a series of choices made in the course of their pre-incapacity lives from which we draw conclusions about what they would have decided if faced with the exact question now considered.30Shepherd, supra note 25, at 1699.

In the absence of express wishes, surrogates often are asked to use a “substituted judgment” process to determine what the patient would have wanted.31Id. at 1702. Yet in some cases surrogates are tasked with using a more traditional guardianship “best interests” standard—and continued life is almost always considered to be in a patient’s best interest, unless outweighed by burdens such as unrelenting pain.32 The result has been a patchwork of requirements varying by jurisdiction, and indeed sometimes varying within a single jurisdiction depending on the patient’s diagnosis.33

A. Incompetency, Brain Death, and Life-Sustaining Care

This Article focuses on situations in which a pregnant patient is alive but not competent to decide whether to begin or continue life-sustaining treatment such as a ventilator or artificial nutrition and hydration. The American Medical Association defines “life-sustaining treatment” as “any treatment that serves to prolong life without reversing the underlying medical condition.”34 Such decisions are always fraught, pitting the fervent hopes of family and friends for an unlikely recovery against concerns over imposing treatments the patient did not want.35 The disputes become infinitely more complicated when the patient is pregnant due to the reality that terminating life-sustaining care will, in nearly all cases, result in the death of the fetus as well. Maintaining life support may not change the outcome for the patient but may allow the fetus to survive long enough to result in a live birth—a fact that became the genesis of efforts to adopt special rules for pregnancy.36

Before delving more deeply into advance directives, it is important to understand both the medical and legal implications of different brain states that may be at issue in these cases.37 This Article focuses on pregnant patients who are medically and legally alive but incompetent, with no realistic hope of recovery. Although rare overall, most reported cases have involved patients in a “persistent vegetative state” (PVS), in which the brain stem continues to function but there is no cognitive or higher brain function.38Johnson, supra note 35, at 58. The functioning brain stem continues to support autonomic functions including the cardiac and circulatory system, sleep-wake cycles, and reflex reactions such as the startle reflex—leading to potential confusion among those who observe the patient with eyes open, seemingly reacting to stimuli.39 Recent research has identified other brain states such as the minimally conscious state (MCS), in which patients retain some minor level of consciousness and have the potential for improvement, leading to the development of new testing protocols to carefully distinguish MCS from PVS.40Fins, supra note 37 (describing MCS). However, in the most well-publicized end-of-life disputes, such as the case of Terri Schiavo, autopsies have confirmed that the patient was indeed in irreversible PVS.41 These cases are the focus of this Article, as they pit the interests of the fetus against those of a living pregnant patient who previously expressed wishes regarding end-of-life care.

In contrast, some cases have involved pregnant patients who are medically and legally “brain dea[d],” having suffered the “irreversible cessation of total brain function”—both upper-level cognitive function and brain stem function.42 Although not without controversy, all fifty states have adopted brain death as a legal definition of death, even if the patient’s heart can be kept beating with mechanical assistance.43 Commentators have been more open to restrictions on terminating life support for patients who are brain dead, arguing that those patients no longer have the same interests as a living patient to balance against those of the fetus. As Bertha Manninen argues:

Because no medical treatment can be performed that would harm or benefit a cadaver, the interests of the fetus become paramount. . . .

. . . .

. . . . [I]t makes no sense to argue that a pregnant cadaver has a right to die with dignity because the person, the one’s whose dignity it is, is already dead.44Manninen, supra note 14, at 411–12.

Although the situations may look similar to the naked eye—an unresponsive, bedridden pregnant patient attached to medical equipment—they are distinct as a medical, factual, and conceptual matter.

At least one court has ruled that these cases are also distinct as a legal matter. In Munoz v. John Peter Smith Hospital, a Texas court addressed the application of the Texas pregnancy restriction to the case of Marlise Munoz, a woman who was fourteen weeks pregnant when she collapsed and was declared brain dead.45 Citing the statute, the hospital refused to remove life support; Munoz’s husband sued.46 A judge ultimately ruled that the pregnancy restriction did not apply because Marlise Munoz was legally dead under Texas law and the statute applied only to living patients, clearing the way for her to be removed from life support.47 While it is possible that other state laws would be read differently, or that amendments to the statute could change this interpretation, this Article will focus on the core target of these restrictions: pregnant patients who are alive but permanently incompetent.

B. Common Types of Advance Directives

Advance directives—written documents indicating a patient’s wishes for treatment or choosing an individual to act as a surrogate—were created as a pathway out of this morass. The laws authorizing such documents serve a number of purposes, some complementary and some contradictory. First and foremost, these statutes, often called “natural death acts,” are a clear acknowledgement by state legislators that it is legally permissible to withhold or withdraw life-sustaining care under certain circumstances.48 As a necessary corollary, the laws provide legal immunity to medical providers who follow the statutory procedures.49 Second, these statutes provide a way for patients to exercise their autonomy by making end-of-life wishes known in advance, thus “enabl[ing] persons to protect their futures by foreclosing the plans of others to determine their destinies.”50Alexander, supra note 27, at 757. Third, the statutes often set forth procedures to apply in the absence of advance directives, creating a pathway that allows medical decisions for an incapacitated patient to be made by a statutory list of default surrogates and/or physicians.51

Several types of advance directives are recognized by law. One model is the “living will,” also known as a health care “directive” or “declaration.”52 Living wills are designed to express the patient’s wishes for medical care in the event they become incompetent or unable to communicate effectively. Rather than documenting specific wishes for care, a second model, called a “medical” or “health care power of attorney” or “agency” document (HCPOA), asks the patient to designate an agent to make health care decisions if the patient lacks capacity to make or communicate such decisions.53 Most of the attention to pregnancy exceptions has focused on living wills and HCPOAs, as they are the types of documents most likely to be created by healthy individuals of childbearing age.

Advance directives are largely a creature of state law; while national models are available, states vary as to the requirements for these documents. Many states will follow directives executed in other states, as long as they are “in compliance with that law or in substantial compliance” with the law of the new state.54 It is common for advance directive statutes to include a sample or template form, although use of that form is not mandatory. In Texas, for example, a directive “may” be in the statutory form, and health care providers may not require patients to use any specific form.55 While not mandating the use of the statutory form, Idaho is somewhat more prescriptive, requiring that a directive “be in substantially the following form, or in another form that contains the elements set forth in this chapter.”56Idaho Code § 39-4510(1) (2022). As will be discussed below, the availability of template forms creates ambiguity when the pregnancy language on the form (or lack thereof) does not match that in the statute.

As befitting an issue of state jurisdiction, there is no overarching federal advance directive statute. But since 1991, the federal government has sought to encourage patients to create these documents. Under the PSDA, hospitals and other health care providers are required to provide written information regarding the patient’s rights under state law to make health care decisions, including the right to accept or refuse treatment and to create advance directives, as well as the provider’s written policies for implementing those rights; in addition, the medical record must clearly document whether a patient has executed an advance directive.5742 U.S.C. § 1395cc(f).

Other types of documents are also used to express the prior wishes of an incapacitated patient, though they are less likely to be at issue in pregnancy. Perhaps the most well-known is the “Do Not Resuscitate” (DNR) order, through which a physician enters an order—with the approval of the patient or agent—indicating that the patient should not receive cardiopulmonary resuscitation or other forms of life-prolonging care; states often provide for “portable” DNRs that can be used outside of a hospital setting, such as in a nursing home or an ambulance, although experience has been mixed.58 While DNRs tend to be fairly short (and blunt), recent years have seen the creation of a more nuanced “Medical/Physician Order for Scope of Treatment” (MOST/POST) forms,59See, e.g., N.C. Gen. Stat § 90-21.17. also known as “Physician Orders for Life-Sustaining Treatment” (POLST). Similar to a DNR, the MOST form is a clinical rather than a legal document. Developed for use by elderly, chronically ill, and terminally ill patients, MOST forms may address a wide array of medical treatment decisions and preferences based on the individual patient’s situation. The form is executed after discussions between the physician and patient or patient’s representative and is designed to accompany the patient in a variety of medical and residential settings.60 Because most patients of childbearing age do not fall into the target groups, pregnancy restrictions are not commonly found in these statutes, and this Article will not discuss them further.61

C. Criticism of Advance Directives

Although advance directives have become an integral part of our approach to end-of-life care, they are not without significant criticism. Angela Fagerlin and Carl Schneider perhaps put it most bluntly: “Enough. The living will has failed, and it is time to say so.”62 Criticism has ranged from the mundane to the profound, from the practical to the philosophical, encompassing allegations that include limited uptake, the seemingly intractable challenge of asking individuals to accurately predict their future wishes, and all-too-common barriers to carrying out those wishes once expressed.63 While a full discussion of these topics is beyond the scope of this Article, a few are particularly salient to the pregnancy debate.

The most significant criticism is a practical one: the majority of individuals, including those of childbearing age, simply do not create advance directives. Noting that “[d]eath is an unpleasant subject that most people avoid discussing or even considering,” one commentator in 1991 lamented that only 15% of respondents in a recent poll had executed an advance directive—a number he assumed would rise with the implementation of the PSDA.64Alexander, supra note 27, at 777. Yet twenty years later, a meta-analysis of studies from 2011–2016 found that only 36.7% of individuals had completed an advance directive, with little variation regardless of health status: 38.2% of patients with chronic illnesses had completed the forms, compared to 32.7% of healthy adults.65 Despite educational campaigns and the rollout of the PSDA, the majority of patients do not create these documents, seemingly rendering the debate over pregnancy restrictions irrelevant.66

A second criticism focuses on the structure of the statutes themselves. Living wills are usually limited by statute to certain narrowly defined precipitating conditions. Unless the patient meets those criteria, the document is inapplicable—no matter how analogous the situation or how likely the patient’s wishes would be the same. In North Carolina, for example, a directive is effective only if the patient (1) “has an incurable or irreversible condition that will result in the [patient’s] death within a relatively short period of time;” (2) “becomes unconscious and, to a high degree of medical certainty, will never regain consciousness;” or (3) “suffers from advanced dementia or any other condition resulting in the substantial loss of cognitive ability and that loss, to a high degree of medical certainty, is not reversible.”67 It should not be surprising that the determination of whether the living will applies is a medical one. Yet research has shown that medical predictions regarding the timing of death are highly speculative, rendering concepts such as “a high degree of medical certainty” and “death within a relatively short period of time” illusory.68 These provisions have also been criticized for singling out only these final medical treatments, rather than the multitude of other treatment decisions that must be made before the patient reaches this endpoint.69

These definitions may be so narrow that they exclude many of the common situations that patients may envision when they create these directives. The high-profile cases of Karen Ann Quinlan, Nancy Cruzan, and Terri Schiavo—which spurred broad discussion about the need for advance directives—involved patients in PVS, in which the brain stem continues to function but there is no cognitive or higher brain function.70 While some PVS patients require a ventilator, others (including Quinlan) are able to breathe on their own for years.71Johnson, supra note 35, at 69. As a result, PVS may not qualify as “an incurable or irreversible condition that will result in the declarant’s death within a relatively short period of time” as required by the North Carolina statute.72 The same might be true of Alzheimer’s disease, where a quick death or permanent loss of consciousness is unlikely.73 Yet not all statutes encompass permanent unconsciousness or irreversible lack of cognition when it will not lead to death in the short term.74

These limitations make clear that while advance directives are designed in part to enhance patient autonomy, “the assumption that the state primarily is interested in assisting patients to control their own medical destinies” rather than “assert[ing] itself in the conditions that attach to the documents” is not completely accurate.75 As Cruzan made clear, states are not required to be agnostic as to the end-of-life choices made by their citizens; the state has an interest in preserving life and guarding against abuses by surrogates, for example.76 State preferences are clear, not only in the imposition of high evidentiary burdens, as in Cruzan, but also in the narrow drafting of statutes that limit their application to specified situations. State preferences also are clear in the detailed legal formalities that apply to the execution of the documents, such as requirements for witnesses or notarization, which may create a barrier to completion.77 Those same state preferences are likely reflected in the relatively small penalties available if the patient’s directive is ignored, by both statute and case law.78

A final set of practical concerns focuses on the difficulty of getting patients to accurately assess (and express) their wishes. In part, this is a function of the forms’ limitations. But it is also a more fundamental problem: people understandably have difficulty navigating uncertainty regarding potential future diagnoses and treatments. As Fagerlin and Schneider note, “[e]ven patients making contemporary decisions about contemporary illnesses are regularly daunted by the decisions’ difficulty. . . . How much harder, then, is it to conjure up preferences for an unspecifiable future confronted with unidentifiable maladies with unpredictable treatments?”79 And because individuals do not routinely update their directives once made, a decisionmaker must assume that the patient’s preference has remained stable over time. Directives created when the patient is healthy, however, may not reflect changing wishes as the patient evolves through different life phases.80

Even when the documents exist, it is not clear whether they do a good job conveying the patient’s wishes (such as they are) to the ultimate decisionmaker. Sometimes the existence of the document is unknown or it cannot be located. Moreover, a number of studies have questioned the accuracy of surrogates’ (and physicians’) conclusions about what the patient would have wanted, even after reviewing the patient’s living will.81Id. at 35–36. As Lois Shepherd has noted, there is good reason to question the “assumptions . . . that patients have defined preferences about the decision to be made and that family members know them.”82 In short, our expectations for these forms may be too high.

These practical problems coexist with an even more daunting philosophical problem: not only may these documents fail to reflect the competent patient’s evolving wishes, but they also may not reflect the contemporaneous desires of the incompetent patient at the time the decision must be made. Scholars, particularly those well versed in disability rights, have argued that healthy individuals may misperceive their desire to live “in some condition of impairment or with a loss of function. But [living] ‘like that’ is sometimes a condition in which objectively many people would in fact like to live—or at least would like to see how matters progress . . . .”83 From that perspective, living wills are troubling because they privilege the views of a prior, competent version of the patient over those of the current, incompetent patient. As John Robertson has explained:

[T]he patient’s interests when incompetent—viewed from her current perspective—are no longer informed by the interests and values she had when competent. . . . Although still the same person, the patient’s interests have changed radically once she becomes incompetent. Yet the premise of the prior directive is that patient interests and values remain significantly the same, so that those interests are best served by following the directive issued when competent.

. . . .

. . . . Because they either confuse the present interests of an incompetent patient with interests she had when competent, or forthrightly privilege the competent person’s interest in control and certainty over the incompetent patient’s current interests, they pose a threat to incompetent patients.84Robertson, supra note 75, at 7.

Indeed, some have argued that certain diagnoses, particularly dementia, may entail such profound memory and personality changes that they create, in essence, an entirely “new person, whose connection to the earlier one could be less strong, indeed, could be no stronger than that between you and me.”85Dresser, supra note 79, at 35. Those concerns turn out to have particular salience to one condition that might not be contemplated by young, healthy patients: pregnancy at the end of life.

II. An Overview of Pregnancy Restrictions

As of 2019, thirty-nine states identified pregnancy as a condition affecting the enforceability of incapacitated patients’ advance directives and/or surrogate decision-making; the vast majority of those states restrict the choice to withhold or withdraw life-sustaining therapies from pregnant patients who lack decisional capacity, although a few states offer patients the option to specify their end-of-life wishes if pregnant.86 While these statutes are similar in singling out pregnancy as an exception to the standard advance directive rules, they differ in significant ways.

A. Types of Pregnancy Restrictions

With regard to pregnancy restrictions, one of the most influential resources has been the Uniform Rights of the Terminally Ill Act (URTIA), first adopted by the National Conference of Commissioners on Uniform State Laws in 1985.87 The original Rights of the Terminally Ill Act contained language that allowed patients to specify their wishes in case of pregnancy, an approach that drew criticism from, among others, the National Conference of Catholic Bishops.88 The 1989 URTIA revision removed the patient’s ability to choose whether to follow previously expressed wishes in case of pregnancy, clarifying that “[l]ife-sustaining treatment must not be withheld or withdrawn pursuant to a declaration from an individual known to the attending physician to be pregnant so long as it is probable that the fetus will develop to the point of live birth with continued application of life-sustaining treatment,” with an identical restriction applying to surrogate decisionmakers.89 In 1993, the 1985 and 1989 URTIAs were superseded by the Uniform Health-Care Decisions Act, which contains no pregnancy-specific language.90 The 1989 URTIA, however, still forms the basis for many state laws.

The prevailing taxonomy of pregnancy exceptions, developed by the Center for Women Policy Studies and revised in 2012, divides statutes into five categories: (1) automatic invalidation of advance directives at any stage of pregnancy; (2) restrictions modeled on URTIA, focusing on the likelihood that the fetus will develop to the point of live birth; (3) use of a viability standard to determine whether a directive is enforceable (i.e., life-sustaining measures may not be withheld or withdrawn after the point of fetal viability); (4) “silent” advance directive statutes that do not address pregnancy; and (5) statutes that provide the option to specify wishes in case of pregnancy.91 Since 2012, however, the statutes have both proliferated and evolved in ways that push the boundaries of these five categories. In 2020, Shea Flanagan offered a revised taxonomy taking account of recent statutory evolution, sorting the statutes by whether they treat advance directives as: (1) void per se; (2) void if the fetus can with some degree of certainty develop to birth; (3) void unless an “ethical condition” is met (such as physical harm or severe pain for the patient, or a condition that precludes live birth); (4) void unless the patient specifically sets out instructions “in case of pregnancy” in the directive; and (5) valid unless the fetus is viable.92 Due to the primacy of the abortion framework in their analyses, both approaches sorted viability-based restrictions into an independent category.93

For purposes of this analysis, a different taxonomy may be helpful, one less dependent on abortion jurisprudence given recent developments and more mindful of procedural as well as substantive distinctions. Substantively, statutes can be sorted as follows: (1) the law always invalidates a pregnant patient’s advance directive; (2) the law sometimes invalidates a pregnant patient’s advance directive, typically focusing on either the stage of/prospect for fetal development or the health effects of treatment on the patient; (3) the law is silent regarding the effect of pregnancy on advance directives; and (4) the law explicitly permits patients to determine what effect their advance directives will have in the event of pregnancy. To that basic taxonomy, I would add three procedural considerations: (1) are health care providers obligated to determine whether a patient who requires life-sustaining care is pregnant; (2) to which forms (e.g., living will and/or HCPOA) and to whom do the restrictions apply; and (3) are the restrictions found in the statute, in the model forms, or both?94 A statute may fall into multiple categories under this taxonomy—an analytically messier approach but one that better reflects the convoluted legal restrictions actually facing patients and physicians.95

1. Substantive Restrictions

The most intrusive form of pregnancy restriction is a provision that always invalidates a pregnant patient’s choice to withhold or withdraw life-sustaining treatment. The Texas “Directive to Physicians” statute, for example, explicitly states that “[a] person may not withdraw or withhold life-sustaining treatment under this subchapter from a pregnant patient.”96 While Texas law appears to constrain only the choice to decline life-sustaining care, Alabama’s statute explicitly invalidates the entire document: “[t]he advance directive for health care of a declarant who is known by the attending physician to be pregnant shall have no effect during the course of the declarant’s pregnancy.”97Ala. Code § 22-8A-4(e) (1975). These prohibitions apply regardless of the stage of pregnancy, regardless of the prognosis for either the fetus or the pregnant individual, and regardless of whether the patient is even aware of the pregnancy—even when the trauma occurs so early that the pregnancy is diagnosed for the first time at the hospital.

A second category of exceptions is more nuanced, invalidating a pregnant patient’s advance directive only in certain (albeit broad) circumstances that typically focus on either the prospect for fetal development or the health effects of treatment on the patient. One set of exceptions, based on the 1989 URTIA, prohibits withholding or withdrawing life-sustaining treatment if it is “probable” that continued treatment will allow the fetus to be born alive.98 In Alaska, for example:

(b) Notwithstanding any other provision of this chapter to the contrary, an advance health care directive by a patient or a decision by the person then authorized to make health care decisions for a patient may not be given effect if

(1) the patient is a woman who is pregnant and lacks capacity;

(2) the directive or decision is to withhold or withdraw life-sustaining procedures;

(3) the withholding or withdrawal of the life-sustaining procedures would, in reasonable medical judgment, be likely to result in the death of the patient; and

(4) it is probable that the fetus could develop to the point of live birth if the life-sustaining procedures were provided.99

Other states, such as Illinois, broaden the criteria by asking whether fetal development to live birth is “possible” rather than “probable.”100755 Ill. Comp. Stat. 35/3(c) (2022).

Some statutes draw more directly from pre-Dobbs abortion jurisprudence, prohibiting the withholding or withdrawal of life-sustaining care specifically if the fetus is viable.101 Prior to 2021, for example, Colorado law provided that after viability, a directive would “be given no force or effect until the patient is no longer pregnant.”102 The Delaware prohibition combines the viability and fetal development criteria, applying if “it is probable that the fetus will develop to be viable outside the uterus” with continued treatment.103 Despite having the ring of scientific certainty, however, the concept of “viability” has bedeviled abortion jurisprudence ever since it was enshrined into law in Roe v. Wade—a point Justice Alito stressed in Dobbs to explain why Roe must be overruled.104 Rather than relying on that nebulous concept, Louisiana combines fetal development with a set point in the pregnancy, mandating the continuation of life-support if an “obstetrician . . . determines that the probable postfertilization age of the unborn child is twenty or more weeks and the pregnant woman’s life can reasonably be maintained in such a way as to permit the continuing development and live birth of the unborn child.”105La. Stat. Ann. § 40:1151.9(E) (2022).

A related set of statutes takes into account the health effects of continuing life-sustaining measures on the patient as well as the fetus. Critics of early pregnancy restrictions highlighted this omission, arguing that then-current abortion jurisprudence required life-sustaining care to be terminated if continuing the pregnancy would subject the patient to pain and suffering.106 New Hampshire, for example, combines the question of whether fetal life can be maintained with concern for the patient’s wellbeing, prohibiting withholding or withdrawing treatment

unless, to a reasonable degree of medical certainty, . . . such treatment or procedures will not maintain the principal in such a way as to permit the continuing development and live birth of the fetus or will be physically harmful to the principal or prolong severe pain which cannot be alleviated by medication.107

North Dakota similarly requires that care be provided to a pregnant patient unless it will not maintain the fetus “or will be physically harmful or unreasonably painful.”108 Ohio uses a stricter standard, permitting a health care agent to refuse or withdraw care only if the pregnancy or treatment “would pose a substantial risk to the life of the principal.”109

A third category consists of so-called “silent” states where the statutes governing advance directives contain no reference to pregnancy, seemingly leaving the patient’s directive in force. Greene and Wolfe, however, note that the fact that a statute is silent does not necessarily mean the patient’s wishes will be carried out.110Greene & Wolfe, supra note 91, at 4. Those disputes may wind up in state court, a time-consuming process that may necessitate forced treatment until a resolution is reached.111Id. Moreover, many “silent” states also have conscience protections that permit medical providers to decline to participate in decisions to terminate life-sustaining measures, requiring family members to either transfer care to another provider or go to court.112 By contrast, states such as Colorado and Hawaii are “newly silent,” having explicitly repealed their prior pregnancy restrictions because of concerns about infringing on the autonomy of pregnant patients.113 Thus, some states may be silent because they have explicitly rejected prior pregnancy limitations, while others may be silent only because they have yet to enact them—suggesting that statutory silence regarding pregnancy offers no assurance that a patient’s wishes will be respected.

Finally, some states explicitly permit patients to determine what effect their advance directives will have in the event of pregnancy. Maryland is an example of the broadest protection: the statutory health directive contains a section in both the living will and HCPOA provisions titled “In Case of Pregnancy,” which allows the declarant to instruct “[i]f I am pregnant, my decision concerning life-sustaining procedures shall be modified as follows.”114 The statute explicitly states that such instructions are optional and the form itself remains valid if that section is blank. New Jersey similarly grants a “female declarant” the broad authority to include in the directive “information as to what effect the advance directive shall have if she is pregnant.”115 Georgia, by contrast, restricts the option to choose to withhold or withdraw life-sustaining treatment in the case of pregnancy only to a fetus that is not yet viable.116

Georgia illustrates another important difference among statutes that explicitly ask patients for their pregnancy preferences: whether the choice is crafted as an opt-in or opt-out. In other words, what is the statutory default: that the patient’s wish to withhold or withdraw life-sustaining care will not be followed during pregnancy unless the patient explicitly indicates, or that the wish will be followed in all cases including pregnancy unless the patient explicitly indicates it should not be? Florida and Georgia default to a policy of nonenforcement during pregnancy unless the patient expressly indicates otherwise (or, in Florida, a surrogate receives court approval to make that decision).117 Maryland, by contrast, defaults to enforcement of prior wishes unless the declarant expressly provides otherwise.118Md. Code Ann. Health–Gen. § 5-603. The difference is not merely semantic. An opt-in statute such as Georgia’s sends a clear message that the state is not agnostic as to the preferred course of action, even as it grudgingly acquiesces to a patient’s contrary decision. Moreover, the opt-in approach heightens the risk that a patient, particularly one without legal representation, may not fully understand the importance of completing these optional sections of the form—an omission that may have crucial consequences.

2. Additional Considerations

In addition to differences in substantive limitations, pregnancy restrictions vary in other respects, including the obligations placed on health care providers to diagnose pregnancy. Restrictions vary as to whether they apply to the actions of health care providers or to the advance-directive documents themselves. There is also variation as to the target of the prohibitions, including the types of advance directives affected and the range of actors bound by the restrictions. Additionally, states differ in whether the restrictions are found in the statute and/or model form. The answers to these questions may have significant repercussions, not only for treatment decisions but also for the basic issue of whether patients are even aware such restrictions exist in the first place.

a. Providers’ Knowledge of Pregnancy

One consideration is whether the restriction applies only when the health care provider is aware the patient is pregnant, or whether the statute imposes an independent duty on the provider to check for pregnancy. Some statutes focus on the mere existence of pregnancy, such as the Texas statement that “[a] person may not withdraw or withhold life-sustaining treatment under this subchapter from a pregnant patient.”119 Others condition the limitation on the physician’s actual knowledge, applying only when the patient has “been diagnosed as pregnant and that diagnosis is known to [the] physician.”120 Other states require health care providers to determine whether a patient is pregnant before following a directive to withhold or withdraw life-sustaining care.121 Another permutation is found in Missouri, which immunizes health care providers if they are mistaken about whether the patient is pregnant—both if they comply with a DNR and withhold resuscitation from a pregnant patient “while believing in good faith that the patient is not pregnant” and, in the alternative, if they refuse to comply with a DNR and therefore provide “resuscitation to a nonpregnant patient while believing in good faith that the patient is pregnant.”122Mo. Rev. Stat. § 190.609(3) (2022).

b. Application of Restrictions

Another key distinction is the target of the prohibition: some statutes restrict the enforcement of advance-directive documents, while others restrict the treatment decisions of health care providers and surrogates. Recall the two key types of advance directives: (1) living wills, in which patients set forth their wishes for care, and (2) HCPOAs, in which patients name individuals to make decisions if the patients are incapacitated.123See supra Section II.B. Even in the absence of these documents, many statutes identify a list of default surrogates who should be consulted regarding medical decisions, both for life-sustaining and more ordinary forms of care.124 Because living wills may be limited to statutorily defined medical conditions, while HCPOAs become effective when a patient more generally lacks capacity, the latter documents apply far more broadly. Whether the pregnancy restriction applies to a living will, an HCPOA, and/or the decisions of statutorily ranked surrogates thus has very practical consequences.

Some states have answered this question by inserting the restriction into both the living will and HCPOA statutes—both invalidating the pregnant patient’s instructions and denying the agent the ability to make that decision on the patient’s behalf. The 1989 URTIA contained identical restrictions governing a patient’s health care declaration and others’ consent to withhold or withdraw life-sustaining care; indeed, harmonizing the two sections was one of the reasons for revising the earlier model statute.125 Nebraska follows that model, prohibiting providers from following a declaration to withhold or withdraw care “so long as it is probable that the fetus will develop to the point of live birth” and, in parallel, denying a health care agent the authority to make such a decision.126 Alaska reaches a similar result through a single pregnancy restriction, providing that:

[A]n advance health care directive by a patient or a decision [to withhold or withdraw life-sustaining procedures] by the person then authorized to make health care decisions for a patient may not be given effect if . . . it is probable that the fetus could develop to the point of live birth if the life-sustaining procedures were provided.127

Rather than invalidating the decision or document, Texas focuses instead on the actors, prohibiting “a person” from withholding or withdrawing life-sustaining care from a pregnant patient.128

Whether intentional or due to drafting errors, other states impose the restriction on only one category of advance directive. In Illinois, for example, the pregnancy restriction is contained in the Living Will Act but not in the Health Care Surrogate Act.129 Minnesota law is particularly confusing in this regard. A living will executed before August 1, 1998, “must not be given effect as long as it is possible that the fetus could develop to the point of live birth.”130 After August 1, 1998, however, a patient may execute a “health care directive”—encompassing not only instructions to health care providers (i.e., a living will) but also instructions for surrogates and/or the appointment of a health care agent—and that directive may contain “instructions by a woman of child bearing age regarding how she would like her pregnancy, if any, to affect health care decisions made on her behalf.”131 While health care providers “shall presume” a pregnant patient would have wanted life-sustaining care if “there is a real possibility . . . the fetus could survive to the point of live birth,” that presumption will be negated by either contrary instructions in the directive or by clear and convincing evidence of the patient’s wishes.132Id. § 145C.10(g). Similarly, some states with silent statutes permit women to create alternate instructions in case of pregnancy, but only on one type of form; Arizona, for example, offers that option on the living will form but not the HCPOA.133 In short, states differ significantly on these issues.

c. Where Are the Restrictions?

Finally, states differ in whether pregnancy restrictions are found in the statute, in the model form, or in both. The broad Texas restriction, for example, is found in both the statute and the form.134 But not all states are this explicit. As Elizabeth Villarreal argues, an unclear requirement that advance directives be “substantially the same” as a model form raises ambiguity about whether a pregnancy restriction is mandatory or whether a patient could simply opt not to include it.135Villarreal, supra note 8, at 1061. A disconnect between statute and form may create two distinct problems. A restriction contained in the statute but not in the recommended form—which may be the only document the patient actually sees—raises the disturbing possibility that patients will be unaware of the applicable pregnancy restrictions. On the other hand, a restriction that appears in the recommended form but not in the statute may be less persuasive as an indication of the legislature’s intent, and thus perhaps easier to challenge in court. The confusion means “that there is virtually no public awareness that [the restrictions] even exist, in part because there is no uniformity in the way in which pregnancy exclusion clauses are written into state statutes.”136Greene & Wolfe, supra note 91.

The most troubling disconnect exists in states that have clear statutory pregnancy restrictions that are not referenced in state forms. Indiana, for example, has an absolute statutory prohibition stating that a living will “has no effect” during a pregnancy, yet the model living will form makes no mention of the restriction.137 One recent review estimated that sixty-nine percent of states with pregnancy restrictions failed to disclose the restrictions in their forms.138 From the perspective of patients, these

pregnancy restrictions are largely hidden from view, ensconced in statutes but not disclosed . . . . [Those advance directives] manifest false assurance and also deny women the knowledge they would need to advocate for their own future interests at a time when they have the capacity to do so.139

If the documents upon which most patients rely are silent on the issue, it is nearly impossible to argue that patients have notice of—let alone agree to—the restrictions.

A different problem exists in states where the statute is silent but the form contains a restriction. The Washington statute, for example, contains no explicit pregnancy restriction, but the statutory form states that it “shall have no force or effect during the course of my pregnancy.”140 Similar restrictions in Idaho’s statutory form, as well as guidance on the Secretary of State’s website, were challenged by a group of women who had executed advance directives that did not contain the restriction and hence did not comply with the statutory template.141 After initially rejecting a facial challenge but permitting the plaintiffs to amend their complaint to bring an as-applied challenge,142 the district court held that the statute did not require directives to include the pregnancy restriction; the statutory form was a “suggested format” rather than “the exclusive format for a directive,” and the State’s interpretation was a violation of the plaintiffs’ constitutional rights.143 As will be discussed more below, the lack of statutory language may thus weaken the force of administrative efforts to impose restrictions.144See infra Section III.B.

B. Rationales

Two primary rationales have been advanced in favor of these restrictions. The first is the pro-life—or at least pronatalist—principle that health care decisions during pregnancy should be guided by the goal of saving the life of the fetus, if at all possible. This view has had strong ties to organized religion, particularly the Catholic Church, which has helped to shape policy in some states.145 The second rationale hearkens back to the inability to accurately predict future desires, one of the major criticisms of advance directives in general: the concern that a nonpregnant patient who creates a directive might not fully contemplate how those wishes might change in the case of pregnancy. It is my contention that this latter rationale is far more insidious than the former, in part because it is couched in terms that appear to give primacy to the concept of patient autonomy. Yet it is also a rationale that might be amenable to a pregnancy-specific advance directive, as described later in this Article.146See infra Section IV.B.

1. Pregnancy Restrictions as Protection of Life

It is clear from the legislative history that pro-life considerations played a major role in debates over pregnancy restrictions. The Louisiana statute makes that explicit, stating: “It is the policy of the state of Louisiana that human life is of the highest and inestimable value through natural death. When interpreting this Subpart, any ambiguity shall be interpreted to preserve human life, including the life of an unborn child . . . .”147La. Stat. Ann. § 40:1151.9(E) (2022). That language is not only an indication of legislators’ goals, but also a clear instruction to judges regarding how the statute is to be interpreted. A Georgia trial court similarly upheld the State’s refusal to terminate life support for a brain-dead pregnant woman because the Georgia statute created a “clear implication of . . . a public policy favoring the maintenance of every reasonable possible chance for life.”148

Organized religion has played a significant role in advancing this position. The National Conference of Catholic Bishops objected to language in the 1985 URTIA that allowed patients to specify their wishes in case of pregnancy, arguing: “The State is thus placed in the position of ratifying and facilitating a decision to end the life of the child. . . . Instead of ignoring the unborn child’s independent interest in life, the law should provide for continued treatment if it could benefit the child.”149 Where states have amended their laws to uphold patient wishes, they often have done so over religious opposition. When Hawaii repealed its pregnancy restriction in 2000, for example, the Senate Standing Committee concluded that “[a] woman should have the right to predetermine her medical treatment, including treatment during her pregnancy, if she should lack capacity to make a health care decision for herself,” despite testimony from organizations including the Hawaii Right to Life, Hawaii Catholic Conference, and Christian Voice of Hawaii.150 While historically this debate has largely played out as a matter of lobbying efforts, the pronatalist approach received a significant boost from Dobbs. Although expressing no “view about when a State should regard prenatal life as having rights or legally cognizable interests,”151 the Dobbs majority chided the dissent for failing to attach significance to the protection of fetal life and highlighted claims that “many people now have a new appreciation of fetal life,” lending strong support for these historically religion-based arguments.152

2. Pregnancy Restrictions as Protection of Autonomy

A separate rationale for these restrictions is that pregnancy is a condition that might not have been contemplated at the time the patient set forth earlier wishes regarding life-sustaining care. “One possible justification for excluding pregnant women from using living wills . . . may be that the state believes women are unlikely to think about how their preferences might change during pregnancy.”153 As Sherry Colb explains:

The default life support situation in most people’s imaginations . . . may be one in which the patient is not pregnant and is therefore making a decision that is distinct from the decision of a pregnant woman.

. . . . If she were asked separately, then, whether she would want to be kept alive if necessary to sustain an existing pregnancy, her answer might well be different.154

This view differs from the pronatalist view in that it focuses not on the rights of the fetus, but on pregnant patients themselves. Rather than infringing on the patient’s autonomy, in this view, pregnancy restrictions are there to protect it: to make sure that this is really what the patient would have wanted in these circumstances, given the high stakes. The hidden nature of many pregnancy restrictions only fuels this concern; the lack of transparency supports the idea that a patient truly could not have considered the issue fully because she had no idea the restrictions even existed.155

That argument is belied, however, by the inflexible operation of the restrictions. Indeed, “the way that many of these state laws are written . . . reveals that accurately capturing a woman’s preferences cannot be legislators’ only concern,” as “[t]he laws are indifferent to whether the living will was created prior to or during a pregnancy” or whether pregnancy was explicitly contemplated.156 Even if this rationale can be taken at face value, there are far easier ways to achieve this goal short of a blanket prohibition. Most obviously, the forms could include the applicable restrictions so patients understand what they are agreeing to when they sign. Moreover, those forms could simply ask patients to contemplate their wishes should they be incapacitated and pregnant, and offer them the opportunity to include those instructions on the form—as states such as Maryland already do.157 States could offer additional planning materials that provide information on the types of pregnancy-related situations that might arise, the types of interventions that might be needed, and the (sparse) data on the likely outcome for a fetus at various stages of development.158 States could even offer resources to help patients think through the ethical implications of the choice. In short, there are concrete actions states could take to address concerns that patient wishes might change due to pregnancy and to help patients engage in a more thoughtful planning process.

And yet none of this happens. Pregnancy restrictions are not coupled with funding for programs to educate patients about the potential effects of their choices on a fetus, nor to offer them opportunities to grapple with those considerations in a nonjudgmental context. These restrictions have not been accompanied by any educational efforts to improve the quality of advance decision-making for patients of childbearing age. Nor are they accompanied by the allocation of prenatal resources that might improve fetal outcomes, let alone assistance to help anyone raise these children once born. Beyond the bare desire for the fetus to progress to a “live birth,” these states seem unconcerned with what will happen to the (potentially parentless) child afterward. In fact, “no living will restriction considers the newborn’s prognosis or quality of life after birth, though it is quite conceivable that the mandated treatment could seriously and permanently impair the fetus.”159Taylor, supra note 19, at 99. The failure to attempt any of these efforts, and instead to impose a blanket restriction on all women in these circumstances, suggests that this approach may at its core be no more than a disguised version of the pronatalist rationale, designed merely “to ensure that as many fetuses are carried to term as possible.”160Youngsmith, supra note 145, at 426.

Moreover, even if states allocated educational and support resources, the bare fact that women’s judgment is being questioned in this way should give us pause. As critics of advance directives have noted, the documents are often prepared without full appreciation for how our views may change as our life circumstances unfold.161 Yet we generally do not threaten to suspend directives on that basis, and certainly not for men.

After all, there are many conditions that the patient on life support might not have anticipated at the time she created her living will—she might have expected to be older, or she might not have realized that one of her children would very much like her to remain on life support. These events are no more foreseeable than a pregnancy, yet they do not ordinarily unravel a living will or health care proxy.162Colb, supra note 154.

Singling out women for this extra level of “protection” signals deep distrust of a woman’s choices, especially in circumstances where she does not choose to put her fetus above all else. While couched in a concern that the woman be fully able to exercise her autonomy, this rationale is in fact an insidious threat to that autonomy—a limitation imposed to “protect” choice that in reality offers no escape. As Colb notes, “it would seem most prudent to assume that unless she has listed exceptions . . . , she really did mean what she said.”163

III. Challenges to Pregnancy Restrictions

Given the breadth of these restrictions, it is no surprise they have been challenged in both the literature and courts. The scholarly literature has focused on the constitutional implications of these restrictions on procreative liberty, drawing from three primary contexts: the abortion debate (now forever changed post-Dobbs), case law recognizing the right to autonomy in medical decision-making, and cases addressing broader autonomy concerns. None of these arguments, however, have succeeded in convincing legislatures to do away with these restrictions, and after Dobbs there is understandable concern that they never will. Similarly, the few cases that have challenged state law have faltered on procedural matters, such as lack of standing, lack of ripeness, or lack of a justiciable controversy. Nevertheless, it is important to understand the rationale behind these challenges in order to determine whether other approaches might be more useful.

A. Critical Scholarly Analysis

Much of the scholarly literature to date has focused on the constitutional implications of pregnancy restrictions. The abortion debate has loomed large in this narrative, offering support both to those who challenge the infringement on the pregnant patient’s autonomy and those who seek to protect fetal life. A second line of commentary draws from case law recognizing the right to autonomy and bodily integrity in the context of informed consent and medical decision-making.164 While this scholarship has illuminated various problems with pregnancy restrictions, none of these approaches has proven satisfactory.

1. Pregnancy Restrictions and Abortion Law

The primary framework for analysis of pregnancy restrictions, as with seemingly all issues involving pregnancy, has been that of abortion.165 While a full discussion of the tortured history of abortion law in the United States is beyond the scope of this Article, for the past fifty years—until the June 2022 opinion in Dobbs—the concept of fetal viability has been key. The landmark 1973 case Roe v. Wade adopted a trimester framework to balance the interests of the pregnant woman and the state as the pregnancy progressed: the state’s ability to regulate the procedure was highly restricted in the first two trimesters, and only after viability could the state “regulate, and even proscribe, abortion” for the purposes of promoting an “interest in the potentiality of human life.”166 Nineteen years later, in Planned Parenthood of Southeastern Pennsylvania v. Casey, the Supreme Court rejected the Roe trimester framework but retained viability as the chronological turning point.167 In place of the trimester analysis, Casey adopted the “undue burden” test, holding that “[a]n undue burden exists, and therefore a provision of law is invalid, if its purpose or effect is to place a substantial obstacle in the path of a woman seeking an abortion before the fetus attains viability.”168Id. at 878. As late as June 2020, the Court confirmed that the undue burden test governed the analysis of substantive abortion restrictions.169

Under that framework, critics argued that pregnancy restrictions—particularly those that operated as complete prohibitions—imposed an undue burden by “wholly prevent[ing] a woman from exercising her right to abortion whether the fetus is developed to 22 weeks or simply two days.”170Greene & Wolfe, supra note 91. If the state could not bar a pregnant woman from seeking an abortion prior to viability, then logically, the state could not force her to remain pregnant during the same timeframe. As Katherine Taylor argued, “if the woman may abort the fetus before viability, surely she also may refuse life-sustaining medical treatment mandated solely for the purpose of saving the life of the previable fetus,” particularly if her choice is made clear in advance.171 Indeed, the salience of the longstanding abortion analysis is reflected in the many states with pregnancy restrictions that apply only after viability—a clear attempt to fit pregnancy restrictions into the Roe model.172

Commentators representing a variety of viewpoints have argued that abortion is the relevant framework for analyzing pregnancy restrictions. “Most analysts addressing the pregnancy restrictions simply conclude . . . that because the state may usually prohibit abortion after fetal viability, it also may constitutionally compel the woman pregnant with a viable fetus to undergo life-prolonging medical treatment.”173Taylor, supra note 19, at 118. One critic explained that “[b]y invalidating a pregnant woman’s living will for the sake of an unborn child, pregnancy exclusions are in effect anti-abortion measures.”174 Others focused on whether the pregnancy was early enough that the woman legally could have chosen to have an abortion, although the fact that she did not seek an abortion was argued both in favor of and against the restrictions.175 And even some scholars who favored following the patient’s explicit wishes would have allowed the state to restrict those wishes to the period permitted for abortions under state law.176

However, this analysis was upended in June 2022 when the Supreme Court issued its opinion in Dobbs v. Jackson Women’s Health Organization, a challenge to a Mississippi statute prohibiting abortion after fifteen weeks.177 Although the Court could have upheld the statute by setting aside the viability rule while keeping the Roe balancing approach—as Chief Justice Roberts suggested in his concurrence—the majority instead held that Roe had been wrongly decided and overruled Roe and Casey in their entirety.178 Rejecting the idea that Roe could be separated from the viability rule, the majority held that the Court had erred by applying an incorrect analysis to determine that abortion was a fundamental right “‘deeply rooted in [our] history and tradition’ and . . . essential to our Nation’s ‘scheme of ordered liberty.’”179 In the absence of a fundamental right protected by the Constitution, Roe had “short-circuited the democratic process by closing it to the large number of Americans who dissented in any respect” from the decision.180Id. at 2265. Moreover, Roe and Casey could not be saved by the concept of stare decisis: the error was too egregious, the reasoning too flawed, the viability and undue burden analyses unworkable, the spillover effect too great on other areas of law such as standing, and the “reliance” interests too intangible to justify upholding the precedent.181Id. at 2265–77. Instead, the issue was one that should be left “to the people and their elected representatives,” with resulting legislation analyzed under the lower rational basis review standard that governs other health and welfare laws.182Id. at 2259, 2284.

The overruling of Roe had been anticipated as the Dobbs case wound its way through the federal courts, and many states had prepared by enacting ever-stricter abortion restrictions. In May 2021, Texas enacted the Texas Heartbeat Act, which not only outlawed abortion after detection of a fetal heartbeat—approximately five to six weeks—but also created a novel enforcement mechanism relying on private suits against persons who provide or aid an abortion.183 Facial challenges to the Act have thus far been unavailing, although more challenges are expected.184 Legislators in other states proposed similar bans, including an Idaho law that not only bans abortions but also seems to allow both the pregnant patient’s family members and a rapist’s family members to sue the doctor who performed the procedure.185

Indeed, by June 2022, thirteen states had adopted so-called “trigger bans”—abortion prohibitions that would be automatically “triggered” if Roe were to be overruled.186 Laws in some states, such as Kentucky, went into effect automatically after the Dobbs opinion was issued.187 Others, as in Texas, automatically went into effect thirty days after the decision.188 The remaining laws required an additional procedural step, such as certification by the state attorney general that Roe had been overturned, before becoming effective.189 As of November 2022, abortions were unavailable in fourteen states due to a combination of statutory bans and clinic closures, despite voters rejecting stricter bans (and in some cases supporting additional abortion protections) in recent elections.190 While some of these new abortion laws have been challenged under the relevant state constitutions,191 many will go into effect as planned. For the purposes of this analysis, the most significant factor is the overlap between states that have pregnancy restrictions and those with trigger laws, including Texas, Kentucky, Louisiana, and Missouri.192 The chances of those state legislatures removing pregnancy restrictions is slim to none, and there is a realistic concern that states with looser pregnancy restrictions may move to adopt stricter, near-complete Texas-style prohibitions. While the effect of Dobbs on pregnancy restrictions is currently unknown, the decision is unlikely to bode well for opponents.

Regardless of the fate of any particular state proposal, the message is clear: abortion is no longer a satisfactory framework to use in challenging pregnancy restrictions. For critics of pregnancy restrictions, however, reliance on Roe and its progeny may long have been ill-advised. As much as pregnancy restrictions appear similar to abortion restrictions, the analogy is and always has been inapt. A patient seeking to withhold or withdraw life-sustaining care during pregnancy is not seeking to terminate the life of the fetus, as in an abortion, but rather to make a decision regarding the patient’s own medical care. The effect on the fetus is clear, but it is secondary: “That she is pregnant does not convert her choice into an abortion, since her objective is not to kill the fetus, but, rather, to stop existing on life support.”193 Sherry Colb uses the Catholic doctrine of “double effect” to frame this difference, explaining that “the fetus’s death is a collateral, rather than an inherent, consequence of her decision to reject life support,” similar to a pregnant cancer patient’s decision to undergo chemotherapy to save her own life even if it risks fetal harm.194 Treating pregnancy restrictions as a subcategory of abortion restrictions is analytically incorrect, and conflates the critically distinct analyses of abortion and end-of-life decision-making.195 One can only hope that it does not turn out to be an irrevocable mistake.

2. Pregnancy Restrictions, Autonomy, and Bodily Integrity

A second line of commentary draws from case law recognizing rights to autonomy and bodily integrity in the context of medical decision-making. Scholars argue that a pregnant patient’s right to control their medical destiny includes the right to decline medical treatment, even if that treatment will end their life (and, by extension, that of the fetus). Pregnancy should not alter that balance, even in extreme cases.

Scholars have grounded these arguments in various strains of analysis, beginning with the fundamental right to privacy that formed the basis for groundbreaking case law regarding reproductive freedom—an analysis inextricably linked to the abortion debate.196 Privacy is also linked to the concept of bodily integrity, including the individual’s basic right to consent to (or refuse) medical treatment.197 The Dobbs dissenters noted the importance of this analysis:

Everyone, including women, owns their own bodies. So the Court has restricted the power of government to interfere with a person’s medical decisions or compel her to undergo medical procedures or treatments.

. . . . There are few greater incursions on a body than forcing a woman to complete a pregnancy and give birth.198

Katherine Taylor focuses not on privacy but on the way in which these restrictions infringe on women’s liberty interests, notably the ability to exercise prospective autonomy by setting forth wishes in advance.199 Still others argue that singling out women for these restrictions constitutes an illegal form of gender discrimination under the Equal Protection Clause.200 As with the abortion discussion above, however, these arguments have not been successful in overturning pregnancy restrictions.

A strong line of cases supports the right to medical decision-making. In Cruzan, the Supreme Court recognized not only the longstanding common law right to informed consent, which it described as “firmly entrenched in American tort law,” but also the “logical corollary . . . that the patient generally possesses the right not to consent, that is, to refuse treatment.”201 The petitioners had argued that “the forced administration of life-sustaining medical treatment . . . would implicate a competent person’s liberty interest.”202Id. at 279. Although the Court acknowledged that prior case law could support such a right, it did not explicitly so hold, stating instead that “for purposes of this case, we assume that the United States Constitution would grant a competent person a constitutionally protected right to refuse lifesaving” care.203Id. Several years later, the Court reiterated that conclusion in Washington v. Glucksberg, noting that “[w]e have also assumed, and strongly suggested, that the Due Process Clause protects the traditional right to refuse unwanted lifesaving medical treatment” but declining to extend that right to encompass physician assistance with suicide.204 In short, “[b]odily integrity, the interest in avoiding forced physical invasions, arguably is the most fundamental of liberties,” and pregnancy restrictions violate that right.205

Cruzan’s dicta regarding the rights of competent patients, however, does not directly answer the question of how to address patients who are incompetent—the specific subject of the pregnancy restrictions. In 1976, the New Jersey Supreme Court held in In re Quinlan that a patient in a “non-cognitive, vegetative existence” retained a privacy right to terminate life-support measures, and set forth procedures to be used by her family members and physicians to carry out that decision on her behalf.206 The Cruzan Court stopped short of declaring such a federal right, recognizing a liberty interest in the decision but permitting the state to choose “to require clear and convincing evidence of the patient’s wishes” and to “choose to defer only to those wishes, rather than confide the decision to close family members.”207Cruzan, 497 U.S. at 286–87. Even under that standard, however, a pregnant patient’s advance directive would appear to be a decisive indication of her wishes—and yet that is precisely what these laws restrict.208

The Cruzan Court recognized that the patient’s liberty interest must also be balanced against relevant state interests, including not only a general interest in protecting human life, but also a more specific interest in “safeguard[ing] the personal element of this choice through the imposition of heightened evidentiary requirements.”209Cruzan, 497 U.S. at 279–81. This suggests that the right to medical self-determination may, in rare circumstances, be outweighed by a compelling state interest in keeping the patient alive, including interests in “preserving life, preventing suicide, maintaining the ethical integrity of the medical profession, and protecting third parties.”210 Whether any of those interests are strong enough to outweigh a pregnant woman’s explicit written directive is the key question, to which there is no clear legal answer—although Dobbs suggests that a fetus may well be considered a third party in need of protection.211

The most relevant analysis comes from the case of In re A.C., in which the D.C. Court of Appeals held that a lower court had erred in permitting a hospital to perform a Caesarean section (C-section) on a dying pregnant woman, without her clear permission, in an effort to save her fetus.212 A.C. was twenty-five weeks pregnant when her cancer recurred and she was diagnosed as terminally ill. She initially consented to palliative treatment to permit the fetus to reach twenty-eight weeks’ gestation; as her condition deteriorated rapidly over the next few days, and she was intubated and in and out of consciousness, her answers became more equivocal.213In re A.C., 573 A.2d at 1238–39. The hospital requested a declaratory judgment permitting it to perform the C-section, despite “no evidence . . . showing that A.C. consented to, or even contemplated, a caesarean section before her twenty-eighth week of pregnancy.”214Id. at 1239. After a hearing at the hospital, a judge granted the order, finding that A.C. likely would die within two days, that the fetus had a fifty-to-sixty percent chance of surviving an immediate C-section, “that because the fetus was viable, ‘the state has [an] important and legitimate interest in protecting the potentiality of human life,’” and that any “delay would greatly increase the risk to the fetus.”215 The C-section was performed but the baby lived only a few hours, and A.C. died two days later.216Id. at 1238.

On appeal, the D.C. Court of Appeals found that the trial court had erred in ordering the C-section without A.C.’s consent or the consent of her surrogate, holding:

[I]n virtually all cases the question of what is to be done is to be decided by the patient—the pregnant woman—on behalf of herself and the fetus. If the patient is incompetent or otherwise unable to give an informed consent to a proposed course of medical treatment, then her decision must be ascertained through the procedure known as substituted judgment.217Id. at 1237.

Rather than asking what A.C.’s wishes would have been if she were competent, the lower court instead “undertook to balance the state’s and [fetus’] interests in surgical intervention against A.C.’s perceived interest in not having the caesarean performed.”218Id. at 1252. The Court of Appeals declined to determine whether the state’s interests could ever override those of a pregnant woman, but went on to emphasize “that it would be an extraordinary case indeed in which a court might ever be justified in overriding the patient’s wishes and authorizing a major surgical procedure such as a caesarean section.”219Id. This disapproval of the court-ordered C-section, despite the fact that A.C. was dying and the procedure represented virtually the only means to potentially save the fetus’ life, suggests those circumstances would be rare indeed. If the state’s interest was not sufficient under those circumstances, it is difficult to argue that it could justify a pregnancy restriction setting aside clear wishes contained in a patient’s advance directive. Yet many pregnancy restrictions give primacy to the state’s interests throughout the pregnancy, permitting the patient’s interests to outweigh those of the state only when continued treatment would cause pain.220

As with the abortion cases, however, while this argument may be theoretically compelling, it has yet to be successful in overturning pregnancy restrictions. One reason is the timing: in In re A.C., the appellate court did not issue its opinion vacating the lower court order until three years after the events—long after the C-section, long after the deaths of A.C. and her daughter, and long after the pregnancy would have ended even had it come to term.221 While the Court of Appeals could make clear its views for future cases, the final opinion was of little to no use to A.C. and her husband. Unfortunately, this is a common posture for cases alleging violations of a pregnant woman’s decisional autonomy: a court-ordered intervention is overturned on appeal years later, if indeed the court is even willing to hear a case that might be considered moot.222 Moreover, if both mother and child are healthy, it is difficult to establish anything more than “dignitary” damages for ignoring the patient’s wishes.223 In short, the litigation simply comes too late to help pregnant patients in these disputes.

Moreover, Dobbs has called this analysis into question. The Dobbs majority took pains to clarify that the decision applied only to the issue of abortion: “And to ensure that our decision is not misunderstood or mischaracterized, we emphasize that our decision concerns the constitutional right to abortion and no other right. Nothing in this opinion should be understood to cast doubt on precedents that do not concern abortion.”224 Writing separately, Justice Kavanaugh “emphasize[d] what the Court . . . state[d]: Overruling Roe does not mean the overruling of those precedents, and does not threaten or cast doubt on those precedents.”225 But not all of the Justices agreed with that interpretation: Justice Thomas flatly stated that “in future cases, we should reconsider all of this Court’s substantive due process precedents, including Griswold, Lawrence, and Obergefell.”226Id. at 2301 (Thomas, J., concurring). The dissent also raised alarms about the assurances, noting that the majority’s analysis would similarly undermine other decisions involving intimacy and self-determination, and that there could be no guarantee that future Justices would adhere to the majority’s self-restrictions.227 In fact, the majority’s crucial distinction that only abortion involves the “critical moral question posed by abortion”—the destruction of “potential life” or an “unborn human being”228Id. at 2258 (majority opinion).—is precisely the issue when a pregnant patient seeks to exercise the right to autonomy and self-determination by refusing or terminating life support for herself and, most likely, for the fetus. As the dissent noted, “we cannot understand how anyone can be confident that today’s opinion will be the last of its kind.”229 In short, the law of medical self-determination may not be strong enough to support challenges to pregnancy restrictions after Dobbs.

3. Spillover Effects on Broader Autonomy Concerns

One of the most significant problems with pregnancy restrictions is that the limitations on self-determination cannot easily be cabined to the end-of-life scenario. Rather, the restrictions reflect a broader skepticism about the ability of pregnant persons—specifically women, in this context—to make decisions, medical and otherwise, that might affect a fetus. That skepticism is both reinforced by, and in turn reinforces, other efforts to restrict the actions of women solely because they are pregnant.

Commentators from a variety of analytical perspectives have warned of the dangers of limiting the autonomy of pregnant women. As Dr. Anne Lyerly explains, “[p]regnancy restrictions marginalize a woman’s interests and devalue the importance of her role as narrator of her own life.”230 Joan Mahoney defines the issue as one of basic dignity, noting that “[b]eing pregnant in itself seems to lead to a reduction in one’s dignity,” and argues that the restrictions “treat pregnant women as if they do not have the same right to a death with dignity possessed by all other people.”231 Even those who would accept restrictions in the context of brain death recognize the harm that might “come[] to the violence we would do to the respect for autonomy in general” if such restrictions are allowed to preempt a living patient’s express wishes.232Manninen, supra note 14, at 409.

Katherine Taylor addresses the issue through the lens of subordination, arguing that “the states conscript women’s bodies . . . to serve as a sort of medical device necessary for the prolonged gestation, or forced rescue, of the fetus. A policy that so radically objectifies the incompetent pregnant woman renders her interests almost invisible.”233 In her view, these restrictions are part of an effort to “mandate pregnancy,” which “not only enforce[s] women’s degraded status as the moral proletariat, but render[s] them ‘secondary citizens’ in the eyes of the state” as compared to men.234Id. at 163. Taylor also cautions against using the so-called feminist “ethic of care” to support these restrictions, warning that “no matter how entrenched (or supposedly desirable) the stereotype of women as ‘selfless mothers’ may be in our culture, the state must protect against the legal imposition of that role, lest women become second-class citizens under law.”235 Among other things, these efforts seek to impose a broader duty on pregnant patients to “rescue” the fetus than would apply to others in similarly dire contexts.236

This issue is not only of theoretical interest, but it also has troubling practical implications. Courts have ordered women to undergo C-sections not only when the mother is terminally ill, as in In re A.C., but also when the mother is awake and competent but there is concern for fetal health.237 Lynn Paltrow and Jeanne Flavin identified at least thirty such instances between 1973 and 2005, and the number has only grown.238 Women who have refused C-sections, and even some who have unsuccessfully attempted suicide, have been prosecuted for homicide when their babies have not survived.239 While, as noted above, the women may win in the (rare) event that they appeal, that remedy once again comes too late.240Villarreal, supra note 8, at 1067–68. Indeed, Linda Fentiman has argued that mothers are increasingly considered not a source of nurturance for their children but rather a source of risk, particularly when they decline to follow advice about appropriate behavior during pregnancy:

[T]hese efforts are not simply isolated recommendations made by the media, family members, or health care professionals. Instead, they provide the foundation for a variety of legal enforcement actions. . . . [P]regnant women who consume both illegal and legal drugs may be criminally prosecuted. In addition, many states have enacted laws authorizing the civil commitment of drug-using pregnant women. Further, in a number of celebrated cases, when women have refused to give informed consent to proposed medical interventions, physicians have obtained court orders for the women to be hospitalized and to undergo Caesarian sections or other medical treatment, all in the name of protecting fetal health.241

While some commentators have criticized pregnancy restrictions as encroachments on incompetent pregnant women’s decisions that clearly would be unacceptable if applied to competent pregnant women,242 the reality is far less clear. It may be inconceivable to many observers that states would try to mandate prenatal behavior, but some clearly have. Insisting that such efforts are illegal—in the face of legislation and case law supporting them—is not only a losing argument but also risks diverting attention away from these very real threats to the autonomy of pregnant women, of which pregnancy restrictions are merely a part.

B. Case Law

Despite these arguments, legal challenges to pregnancy restrictions have been difficult to maintain. Procedurally, most suits have faltered due to lack of standing, ripeness, or a justiciable controversy. In DiNino v. State ex rel. Gorton, for example, JoAnn DiNino executed a directive under Washington state law that altered the model form language by making clear that she intended her wishes to be followed in case of pregnancy, and asked her physician to include the directive in her medical record.243 When he refused, citing potential liability for failure to follow the statute, both he and his patient sued the State seeking a declaration that the directive was valid. The trial court granted partial summary judgment, finding the pregnancy restriction to be an unconstitutional violation of DiNino’s right to privacy because it interfered with her ability to control reproduction, but denying her motion for a declaration that the directive was valid.244Id. On appeal the Washington Supreme Court reversed, holding that there was no justiciable controversy. Because DiNino was “neither pregnant nor suffering from a terminal condition,” the dispute was not “ripe for review” and “present[ed] a purely hypothetical and speculative controversy.”245Id. at 1300. A dissenting justice countered that DiNino had a “present, existing interest” in being able to draft a valid directive.246

Plaintiffs have also been unsuccessful in mounting facial challenges to the constitutionality of the statutes. In Almerico v. Denney, a group of women—all of childbearing age, all of whom had been pregnant and had children, and two of whom were pregnant at the time—brought suit alleging that the pregnancy restriction contained in Idaho’s model form and on the state’s website violated the Due Process and Equal Protection Clauses.247 The district court held that the facial challenge was governed by United States v. Salerno, which required a “showing that ‘no set of circumstances exist[] under which the [a]ct would be valid.’”248 The court rejected that argument, noting: “Simply put, there are circumstances under which a pregnant woman’s right to autonomy in her health care decisions can be circumscribed by the state’s right to protect a third party. Those situations may very well be rare, but they exist.”249Id. at 927–28. While denying the facial challenge, the judge left open the possibility of an as-applied challenge.

The DiNino dissent had highlighted the difficulty of bringing such a suit, accusing the majority of functionally precluding any meaningful challenge to the state’s restrictions:

By the majority’s reasoning, a woman must be pregnant and terminally ill before the issue is ripe for determination. . . . [T]he woman whose directive will then be “justiciable” will never benefit from a ruling on the matter. In fact, the case would run a very real danger of being declared moot . . . .250

This criticism has proven prescient. Limiting challenges to situations involving a pregnant, incapacitated woman with clear evidence of her wishes to forgo life-sustaining treatment would require her family members and other surrogates not only to have the wherewithal to access the court system, but to do so at a time of unimaginable tragedy. Such challenges require not only financial but also emotional resources that few are likely to have under the circumstances.

The one case that has come closest to satisfying these conditions is the case of Marlise Munoz, a Texas woman who was fourteen weeks pregnant when she collapsed at home and was rushed to the hospital, where she was declared brain-dead.251See Goodwyn, supra note 46. Her husband, Erick, and her parents told doctors that Marlise “had made it clear to everyone she didn’t want to be kept alive by machines under any circumstances.”252Id. Indeed, both Marlise and Erick were paramedics, leaving little doubt that they understood the consequences of that decision. Yet the hospital, citing the Texas pregnancy restriction, refused to remove life support.253 Erick sued, requesting a declaratory judgment that the statute did not apply to patients who were brain-dead; in the alternative, he asked the court to declare the statute an unconstitutional infringement on the rights to privacy and equal protection under the Fourteenth Amendment.254 Two weeks later, the court held that the statute did not apply because Marlise was dead under Texas law, a determination that made the constitutional analysis unnecessary.255 Although the case seemed well-positioned to force the court to grapple squarely with an as-applied challenge to the pregnancy restriction, the fact that Marlise was brain-dead, rather than alive but in an irreversible condition, allowed the judge to avoid addressing the constitutional arguments head-on.

A subsequent decision in Almerico offers some hope for as-applied challenges by women who are not in such dire situations. While initially rejecting the facial challenge, the trial judge highlighted a potential alternative theory:

[T]he constitutional injury could, alternatively, be described as occurring at the moment the state limits a woman’s right to prospectively dictate the healthcare she receives in the event she becomes incapacitated. Under this view a woman who has executed a healthcare directive which the state has indicated it will not permit to be enforced, suffers an immediate constitutional injury which can be redressed in an as-applied challenge to the statute.256

The plaintiffs filed an amended complaint making those arguments, and the court held that they had standing both because two plaintiffs were pregnant when they filed suit, and because “[w]omen are injured for standing purposes when they draft a directive without a pregnancy exclusion because they face an immediate credible threat that their directives will be ignored and that they will receive end-of-life medical treatment to which they did not consent.”257

The court interpreted the Idaho statute as not requiring a directive to incorporate the pregnancy restriction and found that the state’s fifteen-year practice of promoting a contrary interpretation on its website required the court to reach the constitutional merits. Citing Cruzan, the court held that forcing a pregnant woman to “have life support forced upon her until her baby could be delivered,” despite her directive to the contrary, would violate “the constitutional right of a competent person to refuse unwanted lifesaving medical treatment.”258 While acknowledging that abortion jurisprudence limits the ability to terminate a pregnancy, the court held that Idaho’s interpretation “completely denies the choices of women, regardless of the viability of the fetus,” going beyond what the law permits.259

The Almerico decision represents an important procedural and substantive victory for opponents of pregnancy restrictions. Procedurally, by recognizing that women suffer harm when they face a credible threat that their directives will be ignored, the court opened the door to challenges by healthy nonpregnant women rather than only those in tragic end-of-life circumstances. Substantively, the decision offers crucial support to the argument that Cruzan and its progeny bar such forced medical treatment. Yet whether the decision will stand, and whether other courts will choose to follow it, remains an open question. The decision itself is currently on appeal to the Ninth Circuit. Moreover, the unique posture of the Idaho pregnancy restriction—a restriction found in the model form and state guidance but not in the statute itself—makes it unclear whether this reasoning would apply to a different restriction explicitly incorporated into the text of the statute, such as the clear-cut Texas prohibition.260 Finally, while courts remain free to ground such analysis in the relevant state constitution, Dobbs raises significant doubt about whether the federal Constitution can continue to be interpreted to support such privacy and liberty rights.261 Regardless, the decision represents the first real litigation progress on the issue.

IV. A Necessary Alternative: A Pregnancy Advance Directive

As the previous Sections make clear, while there may be strong arguments that pregnancy restrictions should be held unconstitutional, litigation and academic criticism have proven to be unsatisfactory methods for addressing concerns. Although the recent Almerico opinion offers hope, it thus far has been difficult for affected individuals to satisfy the procedural barriers to as-applied challenges—suggesting that litigation is not a realistic solution for patients who want to ensure their directives will be followed when they are pregnant.262Almerico, 532 F. Supp. 3d 993. Litigation also fails to address genuine concerns about whether pregnancy is a circumstance not contemplated at the time directives are created; there is no evidence that litigation is coupled with educational efforts designed to improve decision-making. Finally, from perhaps the most important perspective—the clinical perspective—the litigation process fails to provide timely guidance to family members and treating physicians. While litigation may eventually define the permissibility of pregnancy restrictions, there is a critical need for a method of resolving these disputes in the short term, or better yet for avoiding them entirely. For that, I suggest we look to the way similar concerns have been addressed in advance directives created for Alzheimer’s patients, and I ultimately propose the creation of a “Pregnancy Advance Directive.”

A. Alzheimer’s Advance Directives

Alzheimer’s disease is a form of dementia “that slowly destroys memory and thinking skills, and, eventually, the ability to carry out the simplest tasks.”263 Affecting an estimated six million Americans, symptoms begin with cognitive difficulties “such as word-finding, vision/spatial issues, and impaired reasoning or judgment,” and worsen as damage to “areas of the brain that control language, reasoning, conscious thought, and sensory processing” occurs; eventually, “[m]emory loss and confusion grow worse, and people begin to have problems recognizing family and friends. . . . [and] may have hallucinations, delusions, and paranoia and may behave impulsively.”264Id. As the National Institute on Aging explains, “[u]ltimately, . . . brain tissue shrinks significantly. People with severe Alzheimer’s cannot communicate and are completely dependent on others for their care.”265Id. The relentless progression of the disease poses obvious problems from the perspective of advance care planning. At the time the patient will need to make crucial choices—including not only decisions about life-sustaining care, but more basic decisions about when to stop driving, how to manage financial affairs, and whether to receive in-home assistance or enter a long-term care facility—the patient likely will not have the mental capacity to do so.

Traditional advance directives do little to address these problems. Given the focus on life-sustaining care, the major question has been whether Alzheimer’s is a triggering condition that permits the directive to go into effect. Some states have amended their statutes to explicitly include advanced dementia as a triggering condition, but other states are less clear.266 Moreover, while questions do arise regarding medical treatment for Alzheimer’s patients, including life-sustaining treatment, the more pressing decisions may be far more mundane, regarding finances and living situations—issues not addressed by traditional directives.267 As elder law expert Lisa Brodoff explains, “[o]ther than living wills, which typically involve planning for end-of-life decisions and the refusal of treatment that prolongs the dying process, people with Alzheimer’s disease have had almost no ability to plan for or decide in advance on their care.”268

Because of these shortcomings, some Alzheimer’s patients have turned to a different type of planning document: a psychiatric or mental health advance directive (PAD or MHAD).

A MHAD is a legal planning document typically made by people with mental illnesses . . . to state their treatment preferences and to appoint a substitute decision maker. The idea is that, while in a period of capacity and stability, people could state their wishes and instructions in advance to be implemented during a period of compromised capacity. In this way, psychiatric patients who have episodic illnesses that impair insight can meaningfully participate in their own treatment decisions at a time when they would otherwise be incapable of making those decisions.269

The decisions covered by an MHAD may include, for example, “electroconvulsive treatment[,] . . . treatment of mental illness with psychotropic medication, and admission to and retention in a facility for care or treatment of mental illness.”270 The benefits of involving patients in such advance decision-making include “increase[d] motivation for treatment [and] improve[d] crisis intervention,” making patients “feel less coerced into treatment and more like collaborators in their care,” and reduced worry and stress for both patients and their surrogate decisionmakers.271Brodoff, supra note 266, at 250.

Even more so than mental illness, timing is a major concern for advance planning in Alzheimer’s disease. Unlike episodic mental illness, which may be marked by “periods of relative stability followed by decompensation,” Alzheimer’s is marked by progressive deterioration.272Id. at 241. Some advocates warn that dementia patients have “a very narrow window to complete an advance directive that likely cannot be changed before being deemed incompetent.”273Padama, supra note 12, at 192. Because Alzheimer’s disease “has a fairly predictable and lengthy course of progression from diagnosis to death,” however, there often is time to plan for future decisions before the patient becomes severely impaired.274Brodoff, supra note 266, at 243. Professor Brodoff suggests completion of an MHAD tailored to the “predictable decision points that Alzheimer’s disease patients are likely to face and about which they may be able to make an informed advance judgment,” including an introductory values statement that can be used to guide decisionmakers when unanticipated issues arise.275Id. at 257. It is crucial not only that the directive be completed when the patient has capacity, but also that it be completed after diagnosis: “an advance directive created after the patient is diagnosed and has experience with the disease is the most complete expression of his preferences and provides the most autonomy for a patient experiencing a cognitive decline.”276

Some advocates have gone a step further, suggesting the creation of a specific Alzheimer’s/Dementia Directive. The most prominent example is the Living with Dementia Mental Health Advance Directive created by Professor Brodoff and Robb Miller for End of Life Washington.277 Based on the Washington state MHAD, the directive includes a personal history and values statement and instructions about care and treatment, including site of care (in-home or institutional), financial planning, and preferences for issues such as aggressive behavior or potential new intimate relationships.278Brodoff & Miller, supra note 11. Although the directive “may not be legal or honored outside of Washington,” patients “can still use it to document [their] wishes and provide a guide for [their] family, health care providers, long-term care providers, and others.”279Id. at 1. While acknowledging concerns over the generally low prevalence of advance directives,280 Brodoff believes the document could prove more popular for Alzheimer’s patients:

Unlike the completion of living wills, where most people who are considering writing them are being asked to plan for the unlikely hypothetical situation of being in a permanent coma or having a terminal illness, here the person doing the MHAD planning may have already been diagnosed and may be looking to plan for the certainty of care needs in the years ahead.281

Alzheimer’s directives—or at least the process used to create them—can also serve an educative function. The stress of the diagnosis itself, as well as initial cognitive difficulties, make this a particularly difficult time. And while the disease follows a general progression, the timing and nuances vary greatly from patient to patient. As Brodoff notes:

One danger . . . is that there are many decision points . . . that are unlikely to happen to most people with Alzheimer’s disease . . . , [and] can paint an extremely scary, needlessly foreboding, and in most cases erroneous picture of life in the future with Alzheimer’s disease.282

Patients thus need to be educated not only about possible future events, but also their likelihood. Brodoff further notes that these types of sensitive discussions already occur in the hospice context, demonstrating that, as difficult as the conversation may be, “if done with skill and sensitivity, it can be a huge relief for both the patient and family to deal with these important issues up front.”283Brodoff, supra note 266, at 297. Advance planning for Alzheimer’s care thus requires both sensitivity and accurate information, with providers compassionately leading the patient and caregivers through the decisional process.

Alzheimer’s directives are not without their detractors, however. Professor Brodoff acknowledges that they may be subject to some of the same criticism levied against advance directives in general, including both that “it may be difficult for a person to predict accurately what type of care he or she would want at a future time” and that “it is questionable whether or not the [document] will be read, let alone followed” by the relevant decisionmakers.284 Others raise a deeper philosophical concern, questioning whether the brain changes caused by Alzheimer’s mean that a patient in the later stages of the disease becomes, in essence, a different person from the one who made the earlier directive—and whether the wishes of the prior individual should no longer be binding. As Rebecca Dresser explains:

Personal identity theory, which addresses criteria for the persistence of a particular person over time, provides another basis for questioning precedent autonomy’s proper moral and legal authority. . . . On this view, substantial memory loss and other psychological changes may produce a new person, whose connection to the earlier one . . . could be no stronger than that between you and me.285Dresser, supra note 79, at 35.

Dresser disagrees with Ronald Dworkin, who argues that we should honor dementia patients’ exercise of their “precedent autonomy” and permit prior wishes to override contemporaneous best interests.286 Dresser remains “unconvinced that an individual’s former wish to avoid such a state should always take priority over her current interests in continuing a life that she appears to value.”287

These concerns may well be overstated: living will statutes, in particular, make it notoriously easy for patients to revoke their earlier instructions even after they are incapacitated, including by nonverbal means. In Texas, for example, the patient “may revoke a directive at any time without regard to the [patient’s] mental state or competency,” including by “canceling, defacing, obliterating, burning, tearing, or otherwise destroying the directive.”288 It is plausible that a patient’s eager acceptance of spoon feeding, for example, could be interpreted as a nonverbal revocation of prior wishes to prohibit feeding assistance. Moreover, patients who create these directives only after they have been diagnosed with Alzheimer’s—rather than decades earlier, as may be the case with typical advance planning documents—are closer to and better educated regarding potential decisions, which may assuage some concerns regarding the disconnect between the wishes of a young, healthy person and the one who currently suffers from the disease. Nonetheless, these arguments illustrate the deep discomfort that may arise when a prior directive requests the termination of a treatment (such as feeding assistance) that appears, as best we can tell, to bring enjoyment to the patient in the current circumstances. Yet, for all its shortcomings, the Alzheimer’s advance directive offers an important option for patients to make their preferences known.

B. A Pregnancy Advance Directive?

Although the situations differ in significant ways, advance directives for early-stage Alzheimer’s patients may offer options for addressing questions of life-sustaining care during pregnancy without having to rely on litigation challenging pregnancy restrictions. It may be time to consider the creation of a “Pregnancy Advance Directive”: a targeted medical form addressing a patient’s wishes in the case of decisional incapacity during pregnancy, which could be completed only after the patient has become pregnant. Similar to the Alzheimer’s advance directive, a Pregnancy Advance Directive could offer the opportunity to make wishes known at a point when the patient clearly does have capacity, while perhaps addressing some of the concerns that have led to the creation of pregnancy restrictions in the first place.

1. Analogies Between Alzheimer’s and Pregnancy

At first blush, it might seem that an Alzheimer’s patient and an incompetent pregnant patient on life support have little in common beyond their lack of mental capacity and the fact that, with time, their underlying conditions will be life-ending. Upon deeper reflection, however, the situations are analogous in key ways, including the progression from capacity to incapacity; the early opportunity to make an informed choice; and, perhaps most importantly, concerns about whether earlier preferences can adequately anticipate the patient’s changed circumstances. Those analogies suggest that a targeted advance directive might also be helpful in the pregnancy context.

Although very different, Alzheimer’s and pregnancy are both progressive conditions. Alzheimer’s disease progresses from initial capacity deficits, through more significant damage, and ultimately to death.289See supra notes 263–265. Pregnancy is also progressive, albeit in a more positive way: from fertilization of the egg, to implantation of the embryo, to fetal development, and ultimately, we hope, to a healthy birth.290 Despite being progressive, both conditions vary significantly from patient to patient, with some experiencing difficult complications—including, for pregnant patients, the need for life-sustaining care—that others avoid entirely.291 Thus, in both cases, patients can be offered good information about the progression of the condition, although the need for specific interventions remains unpredictable. If directives can address this level of uncertainty for Alzheimer’s patients, there is no reason to believe similar documents could not be written regarding pregnancy.

That uncertainty, in turn, illustrates the crucial need for patient education and planning in both Alzheimer’s and pregnancy. As Professor Dresser explains:

The concern for education is one that applies to advance treatment directives generally, but one that is not widely recognized or addressed . . . . People complete advance directives in private, perhaps after discussion with relatives, physicians, or attorneys, but often with little understanding of the meaning or implications of their decisions.292Dresser, supra note 79, at 34–35.

In Alzheimer’s disease, creation of an MHAD provides a mechanism for “important conversation and education . . . . about the real possibilities in the future, issues that may not have been addressed by traditional medical providers because they are too sensitive and there may not be enough time with the patient to engage in a lengthy discussion.”293Brodoff, supra note 266, at 294. Similar concerns arise in the pregnancy context, where “[p]atients often wish to avoid thinking about [withholding or withdrawing life-sustaining treatment]—a completely understandable response.”294Shepherd, supra note 25, at 1727. Yet if we are reluctant to discuss advance directives with healthy pregnant women, we miss a crucial opportunity to obtain important information. As awkward as those discussions may be, having them—ideally, in advance of labor—is an important way to ensure that health care providers and decisionmakers will not be stranded without crucial preference information if an emergency materializes.295

Another similarity between pregnancy and Alzheimer’s disease is the fact that there is usually ample time between diagnosis and any decision point for the patient to make preferences known. After diagnosis, we no longer inhabit the zone of purely hypothetical concerns that alarm the critics of advance planning: the difficulty of “conjur[ing] up preferences for an unspecifiable future confronted with unidentifiable maladies with unpredictable treatments.”296 For those who worry that “the patient’s interests when incompetent—viewed from her current perspective—are no longer informed by the interests and values she had when competent,” a diagnosis alone may not be sufficient.297 Yet while an Alzheimer’s or pregnancy diagnosis may not make that future entirely clear, certainly some previously hypothetical risks become far more salient—be they the types of values the dementia patient would want to guide future decisions, or the way in which pregnancy may cause the patient to balance fetal and maternal health concerns. At the very least, the patient is now far closer to that decision point than at any prior time. That particularly holds true for pregnancy, which has a nine-month duration requiring that previously theoretical decisions about epidurals and delivery settings be made within a period of months rather than years.298

Of course, diagnosis does not cure all decisional ills, nor does it make the unknowable known. As Fagerlin and Schneider note, “[e]ven patients making contemporary decisions about contemporary illnesses are regularly daunted by the decisions’ difficulty. They are human.”299 The stressful experience of being diagnosed with Alzheimer’s—or even with far happier pregnancy news—does not make these decisions any less complex. Yet as discussed with regard to Alzheimer’s disease, a crucial aspect of these targeted directives is their educative function: the opportunity to discuss these issues with a medical professional, especially one with whom the patient has a long-term relationship.300

The obstetrician-gynecologist is in an ideal position to have ongoing discussions with healthy patients about their values and wishes regarding future care and to encourage them to complete an advance directive . . . . Ideally, the physician’s office is the preferred place for such a discussion, rather than in the hospital at a time of crisis, and discussion early in the course of prenatal care or during well-woman treatment is recommended.301

In these discussions, “[p]hysicians can identify and correct misunderstandings about the medical situation, options, and outcomes,” which means that it is the discussion itself, rather than any resulting form, that is “at the core of informed advance care planning.”302

2. Advantages and Drawbacks of a Pregnancy Advance Directive

An advance directive specifically targeting wishes regarding life-sustaining care during pregnancy would offer several advantages over the current situation in many states. Most importantly, it would not only permit patients to make their wishes known, but would also address the concerns of those who fear prior advance directives may no longer reflect the new circumstances of pregnancy. Such a directive would also have the salutary effect of encouraging physicians to discuss these scenarios with their pregnant patients, leading to deeper consideration of these issues and establishing a more detailed record in the event a difficult decision must be made. However, it would not answer those critics who prioritize the interests of the fetus above all else.

The chief advantage of a Pregnancy Advance Directive would be the opportunity for patients to make their wishes known should they become incapacitated and in need of life-sustaining care during pregnancy. In the majority of states, advance directive forms do not even pose the question.303 Most current statutes choose one of two options: assume that the patient’s wishes would not change during pregnancy, or impose the state’s preference that life-sustaining treatment may not be withheld or withdrawn in those circumstances.304See supra Section III.A. A patient who is aware of the issue may try to amend the form to include these instructions, but far too many individuals lack even the most rudimentary appreciation of the existence of pregnancy restrictions in the first place.305 For documents that are supposed to memorialize the patient’s wishes, the failure to even raise the question—to invite the patient to reflect on their preferences—is an unforgivable oversight. Although it may not be a perfect answer, a Pregnancy Advance Directive “preserves choice while also prompting women who may not have thought about the possibility that they could become incapacitated while pregnant to ensure that their wishes remain unchanged.”306Villarreal, supra note 8, at 1075.

A Pregnancy Advance Directive would also be helpful in addressing concern that the patient’s prior advance directives—potentially created many years earlier—might not be an accurate reflection of wishes in the new circumstance of pregnancy. By limiting the directive to those who have already been diagnosed as pregnant, similar to directives created after a diagnosis of Alzheimer’s disease, the form can only be completed by a patient who is both temporally closer to a potential decision point and far more aware of the consequences that decision might entail. While some commentators may worry that it is never truly possible to know what a patient would want in such circumstances, limiting the form to pregnant patients gets us significantly closer to the potential decision point.307See supra notes 79–82.

Above all, a Pregnancy Advance Directive would encourage physicians to discuss these difficult issues with their patients. Pregnancy may well be a joyful time, but it is fraught with uncertainty: uncertainty over bodily changes, fetal development, medication use, diet, delivery options, breastfeeding, and childcare, just to name a few.308 Advocating that physicians steer the prenatal conversation toward unthinkable tragedy may seem to ask too much. And yet the failure to have that conversation eliminates perhaps the only opportunity to help the patient think through these issues. Healthy pregnant individuals, as a general population, tend to be young enough that they may not yet have created advance directives nor appreciate that pregnancy restrictions might apply.309See Villarreal, supra note 8, at 1063. For that reason, the American College of Obstetricians and Gynecologists strongly suggests that obstetrician-gynecologists have these discussions not only during pregnancy, but also periodically for patients of childbearing age, including assisting patients “in considering how a possible future pregnancy may influence their directives.”310 The discussion might lead to more careful consideration of the issues than otherwise might take place.

The core advantage of this approach for the patient is not the binding nature of the document, but rather the opportunity it presents to discuss the issues with physicians and other important individuals, including potential surrogate decisionmakers. In both the MHAD and general informed consent contexts, involving patients in treatment decisions has been shown to have a number of benefits, including increasing motivation for and compliance with treatment.311 Even if the patient decides not to complete a pregnancy advance directive, or the directive is not recognized as binding under state law, the discussion alone might establish a more detailed record in the event such a decision must be made.312 Rather than physicians and surrogate decisionmakers having to fly blind, or trying to extrapolate from off-hand conversations years prior, a pregnancy-specific advance directive could provide a more accurate and detailed sense of the patient’s preferences.313 In the all-too-likely event that the directive does not legally apply, both the discussion and the written record can provide a valuable informational fail-safe: an explanation of the patient’s overall values and preferences, which may be helpful both in assessing the patient’s best interests and in reassuring caregivers that they are doing their best to honor the patient’s desires.314

While a Pregnancy Advance Directive could assuage concerns that patients may not presently contemplate how their wishes regarding life-sustaining treatment might (or might not) change if pregnant, the provisions would be of less use in countering the explicitly pronatalist rationale that prioritizes the interests of the fetus above all else. A confirmation that the pregnant patient truly does not desire life-sustaining treatment would be of little interest under Georgia’s “public policy favoring the maintenance of every reasonable possible chance for life,” for example.315 Nor would an explicit directive likely convince the National Conference of Catholic Bishops, which has accused states with similar statutory language of “ignoring the unborn child’s independent interest in life” and has argued that “the law should provide for continued treatment if it could benefit the child.”316 If fetal interests are paramount, it is irrelevant how accurately a directive reflects the patient’s wish not to be maintained on life-support; fetal interests will nearly always demand that such care be maintained, regardless. To a certain extent, then, advocates on the different sides of this issue simply are not speaking the same language. While there may yet be common ground, the Pregnancy Advance Directive likely would not be it.

3. What Might a Pregnancy Advance Directive Look Like?

There are several possibilities for what a Pregnancy Advance Directive might look like, which could be tailored to fit within both existing state law and the relevant political context. The key decisions are whether to create a separate directive form or incorporate the provision into existing forms; how to ensure the provision is applicable both to living wills and HCPOAs; and the importance of taking care in the wording of the directive to avoid unwanted influences on the patient’s decision. For maximum effect and to avoid confusion and provide transparency, the provision should be included in both the advance directive statutes and in any model forms; moreover, the statutory default procedures for decisions in the absence of a directive should make clear that pregnancy does not preclude a surrogate decisionmaker from considering all permissible treatment and nontreatment options.317 True protection of autonomy also demands that a Pregnancy Advance Directive be written inclusively, focusing on the condition of pregnancy rather than the gender identity of the pregnant patient.318See MacKinnon et al., supra note 16.

A Pregnancy Advance Directive could be a free-standing form, such as the MHAD, that applies only to individuals who have been confirmed to be pregnant at the time the form is signed. Particularly if physicians initiate the discussion during routine prenatal care, this might increase the chances that a form will be completed, or at least discussed.319 Moreover, a free-standing provision could be written to apply to both living wills and HCPOAs, avoiding inconsistencies between the various documents that can complicate decision-making.320 No state currently appears to follow this model, although Alaska uses a similar mechanism to achieve the opposite result: a single pregnancy restriction refusing to permit “an advance health care directive by a patient or a decision [to withhold or withdraw life-sustaining procedures] by the person then authorized to make health care decisions for a patient” if a live birth is probable.321 A freestanding provision thus has several advantages, but also one key disadvantage: given the relatively low uptake of advance directives in general, it may be unrealistic to expect patients to complete an additional planning document.322

The other option is to add a pregnancy-specific provision to existing advance directive provisions. This is the general approach taken by states that currently give patients the option to choose whether their wishes will differ during pregnancy. A good example is Maryland’s model form, which contains an “In Case of Pregnancy” section, applicable both to living wills and HCPOAs, that invites the patient to state: “If I am pregnant, my decision concerning life-sustaining procedures shall be modified as follows.”323 The form explicitly states that such instructions are optional, and the form remains valid if that section is blank.324Id. Virginia similarly offers the option of memorializing specific wishes in case of pregnancy: “If you wish to provide additional instructions or modifications to instructions you have already given regarding life-prolonging procedures that will apply if you are pregnant at the time your attending physician determines that you have a terminal condition, you may do so here.”325

To be consistent and comprehensive, a similar provision should be added to both the health care proxy and living will forms.326 Maryland’s “In Case of Pregnancy” section is found in both the living will and HCPOA provisions.327Md. Code Ann. Health–Gen. § 5-603. The New Jersey statute does not use identical provisions, but rather clarifies that the patient has broad authority to include in a directive “information as to what effect the advance directive shall have if she is pregnant” and allows the patient to limit (or not) the authority of the proxy in case of pregnancy.328 In both Virginia and Maryland, a single form includes both the living will and health care proxy provisions, making uniformity easier.329 This approach, however, should be viable even in states with different statutes governing different forms.330

The goal of the Pregnancy Advance Directive is not to bind the patient to an earlier choice to reject life-sustaining care, nor to force a change in prior instructions in the case of pregnancy: it is simply to permit the patient to memorialize their own wishes should the situation arise.331See Villarreal, supra note 8, at 1053. However, scholars of behavioral economics have demonstrated how easily patients can be swayed by “framing effects” such as the default selection or the order in which options are presented.332Id. at 1071–72. Current pregnancy provisions differ significantly in whether the choice is crafted as an opt-in or opt-out; under some statutes the default is that the patient’s wishes to withhold or withdraw life-sustaining care will not be followed during pregnancy unless the patient provides instructions, while in others the general wishes will be followed during pregnancy unless the patient indicates they should not be. Oklahoma, for example, instructs that a “pregnant patient shall be provided with life-sustaining treatment and artificially administered hydration and nutrition, unless the patient has specifically authorized, in her own words” that care should be withheld or withdrawn.333 Maryland, by contrast, defaults to enforceability of the existing directive unless the declarant expressly opts out in case of pregnancy.334Md. Code Ann. Health-Gen. § 5-603.

If the goal of the Pregnancy Advance Directive is to memorialize the patient’s own wishes, rather than to nudge patients to make a particular choice, care must be taken in drafting the provisions. As Fagerlin and Schneider recognized in their withering critique of advance directives, “[a]n ocean of evidence affirms that answers are shaped by the way questions are asked.”335 This observation has been bolstered by the advance of behavioral economics, which “suggests that living will templates will solicit more accurate and unbiased responses if they are drafted with neutral language. Similarly, no model form should have a preselected default.”336 Oklahoma’s assumption that care would be desired during pregnancy clearly fails that test, but so does Maryland’s assumption that the patient would want prior wishes to govern during pregnancy.337 Villarreal concludes that an approach such as Connecticut’s is preferable because it asks patients “to make an active choice between applying their living will without modifications during pregnancy, accepting additional medical intervention, or something else that they specify.”338 Although these considerations may make the drafting process more difficult, they are by no means insurmountable barriers, and they would go a long way toward ensuring that documents accurately capture the patient’s wishes—whatever those wishes may be.

Conclusion

No one wants to contemplate the need for end-of-life decisions during pregnancy. The juxtaposition of one of the most joyful events in many lives with one of the most wrenching is difficult to grasp. But, however rare, these events do at times coexist. The tragedy may be compounded when family, friends, and other decisionmakers (including physicians) learn that in the majority of states, their ability to withhold or withdraw life-sustaining treatments is limited and the patient’s prior advance directives are invalid—often without the patient’s knowledge at the time the documents were drafted.

Pregnancy restrictions have been subjected to criticism in the scholarly literature and challenged in the courts. The literature has focused primarily on three contexts: prior abortion jurisprudence, the common law right to autonomy in medical decision-making, and broader autonomy concerns. Yet none of these arguments have succeeded in convincing legislators that the statutes are inappropriate restrictions on procreative liberty, and the chances of succeeding are far slimmer after Dobbs. Similarly, procedural issues such as lack of standing and ripeness have doomed the majority of the few cases challenging these state laws. Thus far, litigation has not been proven a realistic solution for patients who want to ensure their advance directives will be followed when they are pregnant, nor do these challenges adequately address concerns by proponents that pregnancy may not be a circumstance contemplated at the time most directives are written. And from the perhaps most important perspective—the clinical perspective—these challenges do little to ensure that family members, friends, and treating physicians have timely guidance regarding the patient’s wishes. In short, another approach clearly is needed.

Taking a cue from those who advocate for the creation of special advance directives for early-stage Alzheimer’s patients, this Article has argued that it is time to consider the creation of a Pregnancy Advance Directive: a targeted medical form addressing a patient’s wishes in the case of decisional incapacity during pregnancy, which could be completed only after the patient has become pregnant. Although it would not answer critics who prioritize the interests of the fetus above all, it would address the concerns of those who fear prior advance directives may no longer reflect the new circumstances of pregnancy. Most importantly, it would have the salutary effect of encouraging physicians to discuss these difficult decisions with their pregnant patients in advance, hopefully leading to deeper consideration of the competing concerns and establishing a more detailed record in the event—however rare—that such a difficult decision must be made.

* Dan K. Moore Distinguished Professor of Law, UNC School of Law; Professor (Secondary Appointment), Social Medicine, UNC School of Medicine; Adjunct Professor, Health Policy & Management, Gillings School of Global Public Health. I am grateful for the help of research assistants Andrea Liberatore, Grace Nelmes, Caroline Pope, and Sonam Shah, Reference Librarian Donna Nixon, and participants in the American Society of Law, Medicine & Ethics 2021 Health Law Professors Conference and the Birth Rights Bar Association 2021 Childbirth and the Law Conference. I thank Dr. Annie Lyerly for the conversation that sparked the idea for this article. I am indebted to the students in my Fall 2020 Women & Health Law Seminar, particularly Trey Ellis, for their thoughtful discussion of these issues. All errors are mine alone.