Toward an Ethically Sensitive Implementation of Noninvasive Prenatal Screening in the Global Context

Noninvasive prenatal screening using cell‐free DNA, which analyzes placental DNA circulating in maternal blood to provide information about fetal chromosomal disorders early in pregnancy and without risk to the fetus, has been hailed as a potential “paradigm shift” in prenatal genetic screening. Commercial provision of cell‐free DNA screening has contributed to a rapid expansion of the tests included in the screening panels. The tests can include screening for sex chromosome anomalies, rare subchromosomal microdeletions and aneuploidies, and most recently, the entire fetal genome. The benefits of this screening tool are generally framed, by both providers and commercial laboratories, as enhancing reproductive autonomy and choice by providing an earlier, simpler, and more accurate screening while potentially reducing the need for invasive follow‐up testing. The majority of the literature has explored these issues empirically or conceptually from a European or North American vantage point, one that assumes normative priorities such as individual reproductive autonomy and the clinical availability of maternal health care or prenatal screening programs within which cell‐free DNA screening is offered. While its implementation has raised both challenges and opportunities, very little is known about real‐world experiences and the implications of the rapid introduction of cell‐free DNA screening outside of North America and Europe, especially in low‐ and middle‐income countries. To begin addressing this gap in knowledge, we organized a four‐day international workshop to explore the ethical, legal, social, economic, clinical, and practical implications of the global expansion of cell‐free DNA screening. We describe eight key insights that arose from the workshop.     

Stemming the Standard‐of‐Care Sprawl

The “best interests of the patient” standard—a complex balance between the principles of beneficence and autonomy—is the driving force of ethical clinical care. Clinicians’ fear of litigation is a challenge to that ethical paradigm. But is it ever ethically appropriate for clinicians to undertake a procedure with the primary goal of protecting themselves from potential legal action? Complicating that question is the fact that tort liability is adjudicated based on what most clinicians are doing, not the scientific basis of whether they should be doing it in the first place. In a court of law, clinicians are generally judged based on the “reasonably prudent” standard: what a reasonably prudent practitioner in a similar situation would do. But this legal standard can have the effect of shifting the medical standard of care—enabling a standard‐of‐care sprawl where actions undertaken for the primary purpose of avoiding liability reset the standard of care against which clinicians will be adjudicated. While this problem has been recognized in the legal literature, neither current ethical models of care nor legal theory offer workable solutions . One of the best examples of the conflict between evidence‐based medicine and common clinical practice is the use of electronic fetal monitoring. Despite strong evidence and professional guidelines that argue against the use of EFM for healthy pregnancies, the practice persists. One of the main reasons for this is often assumed to be physicians’ concerns about liability .     

Physician‐Assisted Death and Severe,Treatment‐Resistant Depression

Should people suffering from untreatable psychiatric conditions be eligible for physician‐assisted death? This is possible in Belgium and the Netherlands, where PAD for psychiatric conditions is permitted, though rare, so long as the criteria of due care are met. Those opposed to all instances of PAD point to Belgium and the Netherlands as a dark warning that once PAD is legalized, restricting it will prove impossible because safeguards, such as the requirement that a patient be terminally ill, will inevitably be eroded or discarded. However, some supporters respond that limiting PAD to those suffering from terminal illness, or physical illnesses generally, is arbitrary and illogical. In addition, precisely because such patients are not terminally ill, their suffering may last for years, even the rest of their lives. Finally, severe depression may not be treatable. If PAD is justifiable under some conditions—as I shall assume in this article—then why wouldn't it be justifiable for these patients? Why shouldn't psychiatrists who have nothing else to offer their suffering patients be able to help them to die, if that is what they want?      

The Real‐World Ethics of Adaptive‐Design Clinical Trials

From the earliest application of modern randomized controlled trials in medical research, scientists and observers have deliberated the ethics of randomly allocating study participants to trial control arms. Adaptive RCT designs have been promoted as ethically advantageous over conventional RCTs because they reduce the allocation of subjects to what appear to be inferior treatments. Critical assessment of this claim is important, as adaptive designs are changing medical research, with the potential to significantly shift how clinical trials are conducted. Policy‐makers are swiftly moving to encourage greater use of adaptive designs. In 2016, the newly enacted 21st Century Cures Act instructed the Food and Drug Administration to help product sponsors incorporate adaptive methods into proposed clinical trial protocols and applications for investigational drugs and also biological products. In this article, we review the ethical justifications commonly offered for adaptive designs, explore these arguments in the context of actual trials, and contend that clinical equipoise is a useful standard for adaptive‐trial ethics. We distinguish between theoretical and clinical equipoise and explain why ethical arguments related to adaptive trials tend to focus on the former. Yet we contend that theoretical equipoise can be an unreliable standard for adaptive ethics. While we contend that clinical equipoise is the most critical principle for the primary ethical concerns posed by adaptive trials, we suggest ethical approaches to deal with some additional concerns unique to adaptive designs.     

The Haves,the Have‐nots,and the Will‐nots

What if science enabled us to live an extended lifespan? Well, not us, but people in the future, and perhaps not everybody in the future, at least not at first. Should we allow and encourage science to develop this capability, or should we try to prevent or inhibit it? John Davis's book New Methuselahs: The Ethics of Life Extension is a thorough exploration of these questions. He presents the arguments for and against developing this capacity, and he considers three perspectives: those of individuals who will be able to extend their lives (the Haves), individuals who will not because they can't afford to (the Have‐nots), and individuals who will not extend their lives because they are opposed to doing so (the Will‐nots). Davis is a philosopher, and the bulk of this book is for the cognoscenti.     

Dispelling Stem‐Cell Ideology

Week‐old embryos are considered the richest source of stem cells usable in medical treatments. Because the embryos are destroyed when the stem cells are removed, the debate over the embryo's legal, moral, political, and scientific status has exploded. In this debate, Sheldon Krimsky's Stem Cell Dialogues: A Philosophical and Scientific Inquiry into Medical Frontiers (Columbia UP, 2015) is the single best book. Evenhanded, eminently readable, up to date, educational, scientifically precise, powerfully researched, and very entertaining, Krimsky's slim volume is one that no scientist, policy‐maker, ethicist, or intelligent reader should miss.     

Welcoming the “Intel‐ethicist”

In this issue of the Hastings Center Report, Mélanie Terrasse, Moti Gorin, and Dominic Sisti, urge ethicists to devote scholarly attention to a wave of troubling artificial intelligence applications affecting health consumers’ rights and the quality of their care. I very much agree. We already have neuroethicists, business ethicists, and genetics ethicists; AI‐related systems in health care present more than enough warrant to herald the appearance of a new ethics specialist—the “intel‐ethicist,” let’s say. Nonetheless, Terrasse and colleagues may have exaggerated some of the potential moral problems of AI. Examining only social media and e‐health programs, the authors produce an impressive array of questions to consider, but I will argue that two of them are not nearly as worrisome as might be supposed.     

Trump's Abortion‐Promoting Aid Policy

On the fourth day of his presidency, Donald Trump reinstated and greatly expanded the “Mexico City policy,” which imposes antiabortion restrictions on U.S. foreign health aid. In general, the policy has prohibited U.S. funding of any family‐planning groups that use even non‐U.S. funds to perform abortions; prohibited aid recipients from lobbying (again, even with non‐U.S. money) for liberalization of abortion laws; prohibited nongovernment organizations from creating educational materials on abortion as a family‐planning method; and prohibited health workers from referring patients for legal abortions in any cases other than rape, incest, or to save the life of the mother. The policy's prohibition on giving aid to any organization that performs abortions is aimed at limiting alleged indirect funding of abortions. The argument is that if U.S. money is used to fund nonabortion programs of an abortion‐providing NGO, then the NGO can simply shift the money thus saved into its abortion budget. Outside the context of abortion, we do not reason this way. And the policy's remaining three prohibitions are deeply troubling.     

Access to Health‐Related Goods

There are many good reasons for a merger between bioethics and human rights. First, though, significant philosophical groundwork must be done to clarify what a human right to health would be and—if we accept that it exists—exactly how it might influence the practical decisions we face about who gets what in very different contexts.     

Incidental Findings in Low‐Resource Settings

Much new global genetic research employs whole genome sequencing, which provides researchers with large amounts of data. The quantity of data has led to the generation and discovery of more incidental or secondary findings and to vigorous theoretical discussions about the ethical obligations that follow from these incidental findings. After a decade of debate in the genetic research community, there is a growing consensus that researchers should, at the very least, offer to return incidental findings that provide high‐impact, medically relevant information, when it is not unduly burdensome to the research enterprise to do so. Much as genetic research has been limited to U.S. and European settings, the incidental findings debate has primarily focused on research conducted in high‐income countries. In a 2015 paper, Alberto Ortiz‐Osorno, Linda Ehler, and Judith Brooks note salient differences between the circumstances of research participants in low‐ and high‐resource settings that alter the analysis of when and why incidental findings should be offered to research participants. In this article, we expand on their analysis and present a framework for thinking about how investigators’ obligations to return genomic data might change in low‐resource settings, particularly in settings where participants do not have access to the medical care needed to treat, assess, or monitor incidental findings that are actionable in settings with plentiful resources.     

How to Rethink the Fourteen‐Day Rule

Recently, attention has been drawn to the basic principles governing the use of human embryos in research: specifically, the so‐called fourteen‐day rule. This rule stipulates that human embryos should not be allowed to grow in vitro past fourteen days of development. For years, the fourteen‐day limit was largely theoretical, since culture techniques were not sufficient to maintain embryos up to this point. Yet in the past year, research has suggested that growing embryos beyond fourteen days might be feasible and scientifically valuable. At the same time, work with pluripotent stem cells, including human PSCs, has shown that under certain conditions, they can form structures that recapitulate developmental features of the postimplantation embryo. This raises the possibility that PSCs could generate embryo‐like structures in vitro, even “synthetic embryos,” that might provoke moral concern but would not fall under most current embryo research policies. In countries that permit embryo research, the fourteen‐day rule has long been the linchpin of an effective policy compromise between what remain deeply divided moral positions on the human embryo's status. It has also, particularly in the United Kingdom, been influential in establishing a bioethics public‐policy process. Any moves to change the rule must consider not just the implications for the use of embryos but also the potential impact of this model of bioethical governance of science.     

On Patient Well‐being and Professional Authority

Two papers in this issue address the limits of surrogates’ authority when making life‐and‐death decisions for dying family members or friends. Using palliative sedation as an example, Jeffrey Berger offers a conceptual argument for bounding surrogate authority. Since freedom from pain is an essential interest, when imminently dying, cognitively incapacitated patients are in duress and their symptoms are not manageable in any other way, clinicians should be free to offer palliative sedation without surrogate consent, although assent should be sought and every effort made to work with surrogates as harmoniously as possible. Ellen Robinson and her colleagues report on the implementation of a policy at Massachusetts General Hospital that supports do‐not‐resuscitate orders when cardiopulmonary resuscitation is likely to be ineffective or harmful, even if surrogates disagree. The “Doing No Harm” policy at MGH allows for what MGH calls a “medically indicated DNR” and what in some other places is called “a unilateral DNR”—the writing of an order not to provide cardiopulmonary resuscitation, regardless of surrogate disapproval. These kinds of DNR policies have emerged in some hospitals across the country and for much the same reason that Berger provides in his argument regarding palliative sedation. I support the reasoning and the policies in both the Berger and Robinson papers. However, as the authors would most likely agree, the problems they aim to remedy are not simply about the scope of surrogate and professional authority. They are also symptoms of inattention to professional obligations and system failures.     

What's Wrong with Evidence‐Based Medicine?

Medicine in the last decades of the twentieth century was ripe for a data sweep that would bring systematic analysis to treatment strategies that seemingly had stood the test of time but were actually unvalidated. Coalescing under the banner of evidence‐based medicine, this process has helped to standardize care, minimize error, and promote patient safety. But with this advancement, something of the art of medicine has been lost.     

The Intimate Responsibility of Surrogate Decision‐Making

Daniel Brudney's clear‐headed analysis, in this issue of the Hastings Center Report, of the difference between a patient's and a surrogate's right to make medical treatment decisions contributes to a longstanding conversation in bioethics. Brudney offers an epistemological and a moral argument for the patient's and the surrogate's right to decide. The epistemological argument is the same for both parties: the (competent) patient has a right to decide because she is presumed to know her own interests better than anyone else, and the surrogate is entitled to make decisions because she knows the patient better than anyone else. However, argues Brudney, the moral arguments are not parallel. The patient's moral right to decide for herself is an exercise of autonomy, but the only ground for the responsibility held by the surrogate, says Brudney, is that she knows the patient better than the health care professionals do. If in fact that's not the case, then she forfeits her right to be the patient's surrogate. For all the clarity and force of his argument, I think Brudney may be barking up the wrong tree. In cases of proxy decision‐making, it's intimacy, not knowledge, that does the heavy moral lifting.     

On Legalizing Physician‐Assisted Death for Dementia

Last November, soon after Colorado became the latest state to authorize physician‐assisted suicide, National Public Radio's The Diane Rehm Show devoted a segment to legalization of “physician assistance in dying,” a label that refers to both physician‐assisted suicide and voluntary active euthanasia. Although the segment initially focused on PAD in the context of terminal illness in general, it wasn't long before PAD's potential application to dementia patients came up. A caller said that her mother had Alzheimer's disease and was being cared for at great expense. Suspecting that she will suffer the same fate, the caller reported that she had included in her will “my specific request that if I should be diagnosed, and it is legal to do so, I would like to opt out of a life and many years of suffering.” A few countries, such as the Netherlands and Belgium, already allow PAD for dementia in certain circumstances. It wouldn't be surprising to see a U.S. legalization effort in the coming years.