Adriana Smith: Brain Death, Pregnancy, and Bodies the Law Has No Box For
“The decision should have been left to us” — a bioethics case study
- Who: Adriana Smith (1994–2025), a thirty-year-old registered nurse in Atlanta. Her mother, April Newkirk, became her decision-maker. Her partner, Adrian Harden, was the baby’s father.
- What: In February 2025, Smith went to a hospital with severe headaches and was sent home. The next morning she was found gasping for air. Blood clots had flooded her brain. She was declared brain-dead — legally dead — at nine weeks pregnant. The hospital kept her body on machines for about four months, over the family’s objection, citing Georgia’s abortion law. Her son Chance was delivered prematurely in June. Smith’s body was removed from support four days later.
- Where / When: Emory University Hospital Midtown, Atlanta, Georgia, February–June 2025. A custody fight over Chance ran into December 2025.
- Why it matters: This looks like an abortion case but is not one. It forces three questions a clinician can meet on any shift:
- When a patient is already dead, who decides what happens to her body — and to the pregnancy inside it?
- Why does this situation fit none of medicine’s familiar categories — organ donation, withdrawal of care, or research?
- When the law is unclear, whose body ends up absorbing the doubt?
- Pair with: Autonomy, Paternalism, and Reproduction (Part A — Paternalism and Its Limits)
The Case
Adriana Smith was thirty, a nurse, and nine weeks pregnant. In early February 2025 she went to Northside Hospital in Atlanta with a severe headache. She was given medicine and sent home. No scan of her brain was done. The next morning her partner, Adrian Harden, woke to the sound of her gasping for air and called 911. At Emory, a scan showed several blood clots in her brain. Her mother gave permission for a procedure to relieve the pressure, but doctors said it could not help. The injury was too far gone. Within a short time, Smith met the criteria for brain death.
What brain death is. A person is brain-dead when the entire brain, including the brainstem, has permanently and completely stopped working. Under the Uniform Determination of Death Act — the model law in Georgia and almost every other state — brain death is legal death. It is the same, in law and in medicine, as the heart stopping for good. A brain-dead patient is not in a coma, not unconscious, and not “on life support” in the usual sense. The patient has died. A ventilator and drugs can keep blood moving through the organs of a body whose owner is gone.
So the hospital was not keeping Adriana Smith alive. She was dead. The machines kept her organs working for the sake of the nine-week fetus she carried. Her mother, April Newkirk, later told reporters what the staff had said: Georgia law required them to continue support because the fetus had a detectable heartbeat. The family was given no choice. The plan was to maintain Smith’s body until roughly thirty-two weeks, when the baby would have a fair chance outside the womb. “We didn’t have a choice or a say about it,” Newkirk said. “We want the baby. That’s a part of my daughter. But the decision should have been left to us — not the state” (Lewis et al. 2025).
The law the hospital pointed to was the LIFE Act.
The LIFE Act (O.C.G.A. § 16-12-141). Georgia’s 2019 statute bans most abortions once a “human heartbeat” can be detected — about six weeks. It defines an “unborn child” as a member of the human species carried in the womb, and it treats that child as a legal person for more than abortion. Under Georgia law, a fetus with a detectable heartbeat can be claimed as a tax dependent and can be the subject of a wrongful-death lawsuit. Legal scholars describe the LIFE Act as, at its core, a fetal-personhood law, with the abortion ban as one branch of it (State of Georgia 2019).
For months the machines ran while the fetus grew. Reports later said scans had found possible problems with the fetus. The family, by their account, was never offered a way to stop. In mid-May, with the pregnancy at about five months, Newkirk went to the local press. The story was national within two days.
It drew a coalition you would expect — abortion-rights groups, reproductive-justice organizers, civil-liberties lawyers. It also drew one you might not. Several abortion opponents said the law did not require what Emory was doing. The state’s Republican Attorney General, Chris Carr, put it in writing.
Carr’s statement, May 2025. “There is nothing in the LIFE Act that requires medical professionals to keep a woman on life support after brain death. Removing life support is not an action ‘with the purpose to terminate a pregnancy.’” Georgia House Republican leaders agreed, calling the law “completely irrelevant” to Smith’s case (Office of the Attorney General of Georgia 2025).
Emory did not change course. The hospital said only that it used clinical experts, the medical literature, and legal guidance to make decisions “in compliance with Georgia’s abortion laws and all other applicable laws” (Lewis et al. 2025). In other words: the Attorney General’s opinion was not the hospital lawyers’ opinion, and the lawyers were the ones in charge.
On June 13, 2025, surgeons delivered Chance by emergency cesarean. He was born at about six months of pregnancy — extremely premature by World Health Organization terms — weighing one pound, thirteen ounces. (Accounts of his exact gestational age vary.) Four days later, Emory removed Adriana Smith’s body from support. A funeral followed that month.
The question her case raises is not whether anyone broke the law. The Attorney General said the hospital was not forced to do this. The question is harder: what kind of medical act was Emory performing for four months, and what rules should govern it?
Eight People in One Room
Before taking sides, it helps to slow down and ask a plain question that nurses and ethics committees ask all the time: who has a stake here, and what does each one want? This is called a stakeholder analysis. You list everyone the decision touches, name their interest, and note where those interests collide. It does not tell you the answer. It keeps you from missing someone — which is exactly what the family says happened here.
The stakeholders in Adriana Smith’s case:
- Adriana Smith herself. She was dead, but the dead still have interests we honor — in dignity, and in not having their bodies used in ways they would have refused. She left no advance directive, so no one knew her wishes.
- The fetus, later the baby Chance. An interest in being born and in a decent life. But scans suggested possible harm, and his odds were uncertain.
- April Newkirk, the mother. Adriana’s legal decision-maker. She wanted the baby — but wanted the choice to be the family’s. She called the months of waiting “torture.”
- Adrian Harden, the father. Deeply invested, present from the first 911 call. He had almost no legal standing the whole time.
- The bedside nurses and doctors. Many were Smith’s own former colleagues. They gave months of intensive care to a body that could not get better, with no clear goal and real moral distress.
- The hospital as an institution. Worried about prison time for its doctors and uncapped lawsuits under a personhood statute.
- The state of Georgia. A stated interest in protecting fetal life — voiced loudly by some officials and waved off by others.
- The public. Watching a case that would shape how the next family is treated.
The list points to a recurring tension in the case. The person with the strongest legal voice — the hospital’s lawyer — had the least personal stake. The people with the deepest personal stake — the mother and father — had little or no legal voice.
Why This Is Not an Abortion
The first thing to get straight is the one most headlines got wrong. This was not an abortion case, and that is not a technicality. It turns on a distinction that runs through all of ethics: the difference between what you intend and what you merely foresee.
The LIFE Act defines abortion as an act done “with the purpose to terminate a pregnancy.” That word, purpose, is the hinge. When doctors stop machines on a brain-dead body, what is their purpose? Standard medical practice says: to stop futile treatment on a patient who has died. The death of the fetus is a foreseen side effect of that — sad, but not the aim. On this reading, the act is simply not an abortion, because the purpose is different (Watson 2025).
This is a real-world use of a principle called the Doctrine of Double Effect (DDE).
The Doctrine of Double Effect. An old idea, often traced to Thomas Aquinas, that an act with two effects — one good, one bad — can be permissible even if the bad effect is certain, provided you aim at the good one and merely tolerate the bad one. A doctor may give a dying patient enough morphine to ease pain even if it shortens life, because the aim is comfort, not death. Same outcome, different intention, different moral (and legal) verdict.
DDE explains the strange coalition. It let abortion opponents say the ban did not apply without giving up the belief that the fetus is a person. You can hold that the fetus is a full person and still say: stopping a dead woman’s machines is not an act aimed at ending the pregnancy. That is why even some pro-life voices agreed disconnection was permissible.
The “Not an Abortion” Argument (the Attorney General’s position)
- The LIFE Act bans only acts done with the purpose to terminate a pregnancy.
- Stopping support for a brain-dead patient aims at ending futile treatment on a dead body; the fetus’s death is foreseen, not intended.
- Therefore the LIFE Act does not forbid — and does not even speak to — stopping support, and the family’s wishes should govern.
Here is the catch, and it is the heart of the hospital’s fear. DDE is controversial, and “intent” is slippery. Critics of the doctrine have long argued that the line between intending a harm and foreseeing it is easy to blur and easy to game. From the outside, the two acts look identical: in both, machines come off and the fetus dies. The only difference lives inside someone’s head — and a courtroom is exactly the place where someone else gets to second-guess what was really in your head.
That is precisely what Emory’s lawyers seem to have worried about. It is one thing for the Attorney General to announce that the purpose is innocent. It is another to bet a doctor’s freedom on a future prosecutor — or a wrongful-death jury — agreeing about what that doctor “really meant.” Under the LIFE Act, a physician can face up to ten years in prison, and the institution faces lawsuits with no dollar cap. So the soft spot in DDE — that intent is hard to prove — was not a seminar puzzle for Emory. It was the legal risk. The hospital read every ambiguity in the most defensive direction, because the doctrine that should have protected it was one a hostile reader could attack.
The contrast with Kate Cox’s case in Texas is instructive. There, a single physician, Damla Karsan, put her own name on a lawsuit and accepted the personal legal risk. Emory, by contrast, was an institution weighing liability rather than an individual acting on conscience. A similar kind of statute produced opposite choices.
Almost, But Not Quite
If this is not an abortion, what is it? The honest answer is that it does not fit any box medicine already has. It looks like several familiar things and turns out to be none of them. That is what makes it genuinely hard.
It is almost like organ donation — but the analogy breaks at both ends. We routinely use a dead person’s body to help others. The reason that works, ethically and legally, is that organs are treated almost like property: things that can be given, with consent, under gift law. We have a whole system of donor cards, registries, and family authorization. But a fetus is not property. Under the LIFE Act it is a person, with standing to sue and to be claimed on taxes. You cannot “donate” it, and you cannot discard it, the way you would a kidney. So the consent-and-property machinery that settles organ cases has nothing to grab onto. And it breaks from the other side too: organ donation needs the dead person’s prior yes. Adriana Smith never gave one. No one used her like a registered donor; they used her without any donation at all.
It is almost like withdrawing care from a dying patient — but she was not dying. She was dead. Normally a family can refuse life-sustaining treatment, and the law backs them. Georgia even has an Advance Directive Act for exactly this. But that law only applies to a patient with “a terminal condition or state of permanent unconsciousness.” Brain death is neither. It is death. A dead person cannot have treatment “withdrawn,” and an advance directive does not apply to her, because directives are instructions from the living (Watson 2025). The familiar end-of-life framework simply switches off.
It is almost like research — and here the experts openly disagree. The bioethicist Arthur Caplan argued that keeping a brain-dead body pregnant for months has almost no evidence base, no settled protocol, and no proven track record. A medical act like that, he said, is not treatment at all — it is closer to an experiment. And experiments on human subjects require informed consent, which no one here could give and no one sought.
The Experimentation Argument (Caplan)
- A medical act with essentially no evidence base, no established protocol, and no prior outcomes is research, not ordinary care.
- Research on a human subject requires informed consent — from the subject or a legal surrogate.
- Long-term support of a brain-dead pregnant body meets the test for research, and no such consent was sought for Smith.
- Therefore Emory performed unconsented human experimentation, which research ethics forbids no matter what an abortion law says (Caplan 2025).
But this claim is disputed. A 2025 review in The American Journal of Bioethics by Ariane Lewis and colleagues pushes back hard. They count at least forty-seven reported cases of brain death in pregnancy since 1982, and they point to detailed international guidance — the World Brain Death Project even publishes a management checklist for it. On their account, this is “not novel or experimental.” It is rare and difficult, but known (Lewis et al. 2025).
So who is right? The two sides may not fully contradict each other. Caplan could mean that the four-month length, and the post-Dobbs legal pressure, were unprecedented even if brain death in pregnancy is not. Lewis could be right that the basic practice has decades of cases behind it. Underneath the disagreement sits a deeper question: what exactly makes a medical act count as “experimental”? It could be any of three things:
- That it has never been done before.
- That it has no settled protocol.
- That there is no good evidence it works.
These come apart. Many accepted treatments are short on evidence, yet no one calls them experiments. Where the line falls decides whether Caplan’s powerful argument lands.
Even declaring brain death is done differently in pregnancy. One more sign this case sits outside the normal categories: the usual brain-death exam includes an “apnea test,” which briefly stops the ventilator to see if the patient breathes. Recent work suggests this test may be risky for a fetus, so some experts say it should be skipped in pregnancy in favor of blood-flow scans. Some hospital policies have even barred declaring brain death in a pregnant patient at all until the fetus is delivered (Lewis et al. 2025). Here, the pregnancy affects how death itself is determined.
When the Baby Was Born, the Problem Moved
The trouble did not end when Chance was born. It just changed shape — and exposed a second gap in the same law.
The LIFE Act had treated the fetus as a person with powerful standing: enough to keep a dead woman on machines for months, enough for tax and wrongful-death purposes. But when Chance was actually born, his biological father, Adrian Harden, discovered he had no automatic legal rights to his own son at all. Because Harden and Smith had never married, Georgia law gave an unwed father no parental standing by default — even with a positive DNA test, and even though the baby’s mother was dead. His attorneys warned that Chance could be placed in foster care without a court order (Dukes 2025).
So Harden had to sue for custody of his own child. The fight stretched across most of the year:
- August 2025 — he filed an emergency petition for legitimation and custody.
- September 2025 — a court granted temporary custody, letting Chance go home with him.
- December 2, 2025 — a judge issued the final order granting sole custody.
That last date was nearly six months after his son was born, and ten months after Adriana collapsed.
These two outcomes come from different bodies of law. The LIFE Act is fetal-protection law; an unwed father’s rights are governed by Georgia’s family and legitimation statutes, which long predate it. There is no logical contradiction between them. But placed next to each other in a single family’s story, they show how unevenly the law assigns standing. The fetus had standing the moment a heartbeat was detected. The biological father, with a positive DNA test and the mother deceased, had none until a court granted it months later.
Whose Body Absorbs the Doubt
A final factor is harder to measure but well documented. Georgia’s LIFE Act is, on its face, race-neutral; it says nothing about race. Yet the way ambiguous cases actually get resolved often tracks who the patient is — something that, in principle, should make no difference to the outcome.
The pattern is documented, not merely suspected. Black women in the United States are three to four times more likely to die from pregnancy-related causes than white women, and studies find racial gaps in how quickly pregnant patients are even seen in emergency rooms (Lewis et al. 2025). Adriana Smith was sent home with a headache and no brain scan; whether bias played a role in that first visit is something her own mother raised.
The deeper worry is structural. When a law is ambiguous and the penalties are severe, someone has to absorb the doubt — and it is rarely distributed evenly. Hospitals with thin budgets and many low-income patients have the least room to risk a lawsuit, so their lawyers read every gray area in the most defensive way. Wealthier hospitals have more freedom to read the same gray area in the patient’s favor. The same statute can land very differently depending on whose body it lands on. The reproductive-justice group SisterSong, a longtime opponent of the LIFE Act, made Smith’s case a rallying point for exactly this reason (SisterSong Women of Color Reproductive Justice Collective 2025; Roberts 1997).
None of this decides whether Emory was right. It adds a further consideration: when the rules are unclear, the cost of that uncertainty is not shared equally. It tends to fall hardest on the hospitals — and the patients — with the least room to absorb legal risk.
Discussion Questions
Write out The “Not an Abortion” Argument in your own words, premises first. Which single premise was Emory’s legal team most afraid to rely on, and why?
The case turns on the gap between intending a death and foreseeing it. Describe one other medical situation where that gap matters. Is the distinction real, or just a way to feel better about the same act?
Caplan calls the four months “experimentation”; Lewis calls it rare-but-known care. Pick the one feature you think best decides whether a medical act is “experimental,” and use it to judge Smith’s case.
You are the night-shift charge nurse in a Georgia ICU. A patient is declared brain-dead at eleven weeks pregnant. Her sister, the legal decision-maker, asks that support be stopped; the hospital lawyer has not called back. What do you say in the meeting, and which idea from this case is doing the work?
The law gave a nine-week fetus more standing than it gave the baby’s living father. Propose one change to fix that mismatch — then name one thing your change would not fix.
Further Reading
- Ariane Lewis, Gwendolyn Quinn, and Kimberly Mutcherson, “Ethical Controversies in the Adriana Smith Case in Georgia,” The American Journal of Bioethics (2025). The clearest medical-ethical overview; the source for the claim that brain death in pregnancy is rare but not new.
- Arthur Caplan, “The Adriana Smith Case Unfolding in Atlanta Raises Many Questions,” Bioethics Today (May 2025). Short, sharp, and the origin of the “experimentation” framing.
- Katie Watson, “Brain Death in Pregnancy — Abortion, Advance-Directive, or End-of-Life Law?” New England Journal of Medicine 393:4 (2025). A three-page tour of why the case fits no legal box.
- Hilde Lindemann Nelson, “The Architect and the Bee: Some Reflections on Postmortem Pregnancy,” Bioethics 8:3 (1994). The philosophical roots of the worry about using the dead.
- The Wikipedia article “Adriana Smith pregnancy case” — unusually well-sourced for a factual timeline.