The intense politicization of abortion in U.S. public discourse obscures its status as a health and health care issue. Medical centers may therefore not be doing the careful preparation needed to manage the health system–wide impact of abortion’s criminalization. What follows is a framework for preparation in a state where abortion will become illegal.

At the University of Michigan, we’ve been actively preparing for the loss of abortion care since the December oral arguments in Dobbs v. Jackson Women’s Health Organization made explicit the Supreme Court’s eagerness to overturn Roe v. Wade. In Michigan, a 1931 law criminalizing abortion will come into effect if Roe is overturned. It’s among the strictest laws in the country, permitting abortion only to “preserve the life” of a pregnant person.

Our preparations to date raise far more questions than they answer. Every new voice joining our efforts brings new questions; thoughtful planning, in this and in everything, requires a diverse set of people at the table.

In Michigan, we’ll be able to continue providing “life preserving” abortion care. When my family planning colleagues and I perform abortions in critically ill patients in the intensive care unit (ICU), it’s reasonably clear that we’re working to “preserve the life” of a pregnant patient. Pregnancy demands intense work from all organ systems, which the bodies of critically ill people often cannot accommodate. Ending a pregnancy is an effort to save them. These patients may have severe exacerbations of underlying conditions, such as heart failure or lupus. Or they may have pregnancy-related illnesses in the first or second trimester, such as eclampsia or chorioamnionitis with sepsis.

Beyond such cases, however, it’s unclear what, precisely, “lifesaving” means. What does the risk of death have to be, and how imminent must it be? Might abortion be permissible in a patient with pulmonary hypertension, for whom we cite a 30-to-50% chance of dying with ongoing pregnancy? Or must it be 100%? When we diagnose a new cancer during pregnancy, some patients decide to end their pregnancy to permit immediate surgery, radiation, or chemotherapy, treatments that can cause significant fetal injury. Will abortion be permissible in these cases, or will patients have to delay treatment until after delivery? These patients’ increased risk of death may not manifest for years, when they have a recurrence that would have been averted by immediate cancer treatment. We’ve identified countless similar questions.

Of course, patients facing such risks are the minority of those currently receiving abortion care. Most will not “qualify” for abortion under Michigan’s law and will have only three options: leave the state for care, self-manage an abortion, or give birth. All three have enormous effects on health systems, including primary care and subspecialist care across adult and pediatric settings. All three affect insurers and medical education programs. And since more than half the medical workforce can become pregnant, health systems’ human resources will also be affected.

People with the necessary resources and support will seek care out of state. Many, especially teens, won’t have those resources. According to the Guttmacher Institute, half of Americans seeking abortion care live on incomes under the federal poverty level and another 25% live on incomes one to two times that level, which may make travel out of reach. In Michigan, the average travel distance for care will increase from less than 20 miles to more than 260 miles. Most patients seeking an abortion are already parents, so travel is complicated by child care needs. Many cannot afford to lose wages or will be fired if they take time away.

Nevertheless, health systems in states where abortion is illegal will play vital roles for patients who can leave their state. Clinicians in states with bans should (but may not be allowed to) offer referrals or pretravel “teeing up.” This may include ultrasound, blood work, and perhaps “fast track” subspecialist consultation to ensure that patients with underlying illnesses can safely receive care on arrival at an out-of-state outpatient facility. Hospital systems will need to determine whether neighboring states will have capacity to see their patients who require hospital-level care and will need to develop plans for transferring already hospitalized patients across state lines. Insurers will need to decide whether to cover out-of-state abortion care and associated travel expenses.

The second option is self-managed abortion. Over 20 years’ experience shows that pregnant patients can safely have an abortion at home after receiving the Food and Drug Administration–approved regimen of mifepristone and misoprostol in a medical office. Increasing evidence from around the world shows that self-sourced mifepristone and misoprostol, and misoprostol alone, are safe and effective as well.1 Patients without access to these medications — those without Internet access or a credit card, or with no private way to receive mail, or those who simply don’t know about it — might use methods that are ineffective or, worse, dangerous. These may include insertion of implements, objects, or caustic substances into the cervix or vagina; ingestion of poisons; or intentional trauma. Many care providers, out of compassion and conscience, will want to steer people away from potentially life-threatening methods and toward safe ones, though some states may try to prohibit even such guidance. Health systems should consider what a “harm reduction” approach to abortion would look like in their setting.2

Health care providers, especially in emergency department and primary care settings, will need to become familiar with the normal course of self-managed abortion with medications and its rare complications, as well as complications of unsafe methods. Because medication-induced abortion is safe, many patients who seek follow-up care will require only confirmation that their abortion is complete or outpatient intervention if it’s incomplete. Ensuring that abortion is complete is also important because misoprostol can be associated with fetal birth defects in the rare cases in which it does not end the pregnancy. Patients who use unsafe methods, on the other hand, may require lifesaving critical care for sepsis, hemorrhage, pelvic-organ injury, or toxic exposures. Clinicians will need to intervene only rarely in cases of mifepristone–misoprostol use and simultaneously be ready for aggressive treatment when patients use dangerous methods.1

Because mifepristone and misoprostol are safe, the biggest risks to patients may be legal ones: threat of reporting, arrest, and detention. Pregnant patients who have bleeding in pregnancy or pregnancy loss may be vulnerable to reporting and criminal prosecution, whether they took measures to end the pregnancy or are having a miscarriage; spontaneous pregnancy loss and self-managed abortion with medications are virtually indistinguishable. Data show that health care providers are most likely to report Black pregnant patients and those living on low incomes to the authorities.3 Hospitals will need clear policies for all staff regarding this risk and regarding medical record documentation in this new climate. Currently, no state requires reporting of suspected self-managed abortion.1

People who can’t travel for care or manage their own abortion will give birth. Recent unpublished updates to older estimates — from economist Caitlin Myers and other researchers — are that 18 to 57% of women deciding to end a pregnancy in counties where travel distances for abortion care increase will give birth.4 This estimate translates to a 5-to-17% increase in births in Michigan, which already has maternity care deserts. Requiring labor and delivery units to work over capacity will affect all birthing people, not just those who would have ended their pregnancy. It will affect newborn care as well. Neonatal ICUs, and later pediatricians, will see more babies, including some with substantial medical needs in infancy and beyond, whose parents might have ended their pregnancy after receiving fetal anatomical or genetic diagnoses. It is not at all clear that medical and social safety nets for families of children with disabilities or complex medical needs will expand as need does.

Maternal mortality will increase because abortion is far safer than childbirth. Data from the Centers for Disease Control and Prevention show that the risk of dying from childbirth is 50 to 130 times greater than dying from an abortion. Demographers estimate that maternal mortality will increase by 21% under a ban — but, echoing existing disparities, 13% among White and 33% among Black birthing people.5 These estimates don’t account for additional likely increases in mortality from unsafe abortion. Undoing the systemic inequities and racism that lead to such disparities will become even more urgent than it already is.

The perinatal mental health needs of pregnant patients who are continuing undesired pregnancies, including those resulting from sexual assault, will undoubtedly intensify as well, further stressing an overtaxed mental health care system.

The implications of an abortion ban extend to additional dimensions of reproductive health care. Absent clear policies permitting it, doctors may hesitate to treat patients with ectopic pregnancy, inevitable miscarriage, or previability rupture of membranes when fetal cardiac activity remains. Hospital pharmacies, doctors, midwives, and advanced practice clinicians will need to consider whether they’ll continue to stock and offer the best evidence-based medication treatment for spontaneous abortion — mifepristone and misoprostol, the same medications used in abortion care — whose use could bring accusations of criminal activity. Infertility care practices may need to halt provision of selective reduction for multifetal pregnancies resulting from superovulation or in vitro fertilization (IVF); without multifetal reduction, loss of the entire pregnancy, premature delivery with concomitant risks of neonatal complications or death, and clinically significant maternal complications are likely. Some IVF practitioners may decline to provide treatment altogether, given the potential for embryo loss in IVF. And health care providers everywhere will need to ensure all forms of contraception are readily available, without barriers to access.

The overturn of Roe will also affect medical education. Abortion training opportunities are required for accreditation of obstetrics and gynecology residency and complex family planning training programs, and they are an integral part of some nursing, midwifery, advance practice clinician, family medicine, maternal–fetal medicine, and gynecologic oncology training. Programs will need to consider out-of-state training, with its accompanying licensure and logistic issues. Patients will feel the downstream effects if abortion training halts: if trainees can’t learn “non-lifesaving” abortion care, within a generation there may be no one left to perform “lifesaving” abortions. Routine care such as miscarriage management will also be affected; one of the best predictors of a physician’s providing the full range of miscarriage-management options is having had abortion care training as a resident.

Finally, our health system is disproportionately “manned” by women — including nurses, medical assistants, administrative assistants, inpatient-unit clerks, phlebotomists, x-ray technicians, and more, as well as physicians. It’s not clear how smoothly health systems will function when a larger fraction of the workforce is pregnant, on parental leave, or traveling for abortion care.

Health systems that view abortion exclusively as a political or partisan issue, perhaps one they’d like to avoid, will soon bear witness to the reality that abortion care, or lack thereof, is a health care and health equity issue. Avoiding the issue will not be possible, short of abandoning care and equity missions altogether. Thoughtful preparation is needed now.