The nature of First Amendment protection for speech in the context of the doctor-patient relationship has been the subject of inquiry for several decades. The Supreme Court has only addressed this issue three times — and each instance involved the regulation of speech regarding reproductive care. Unfortunately, the Court been less than clear about the role of the First Amendment in this context. In Planned Parenthood of Southeastern Penn. v. Casey, it held that mandated speech with respect to doctors performing abortions is consistent with the First Amendment.1 Twenty-six years later, it held in Nat’l Inst. of Family Life & Life Advocates (NIFLA) v. Becerra that mandated speech for crisis pregnancy centers that try to discourage women from seeking abortions is not.2 To achieve these divergent outcomes, the Court has had to thread the needle by making fine distinctions between speech in very similar contexts.
While some argue that the Court has carved a path that will make it difficult to uphold further state regulations of speech concerning abortion,Reference Gottesdiener3 others suggest these holdings reflect a form of “constitutional gerrymandering against abortion rights” by twisting First Amendment jurisprudence to achieve a desired outcome.Reference Chemerinsky and Goodwin4 This piece examines what these Supreme Court cases mean for regulations of speech in reproductive care. Specifically, it explores whether states can prohibit doctors from providing certain information obtained through prenatal testing or preimplantation testing of embryos created through in vitro fertilization (IVF).
While that scenario is currently hypothetical, it does not seem far-fetched. For decades, legislatures have been whittling away at reproductive rights through abortion regulations5 and limits on access to contraception.6 Often the measures are wrapped in the guise of uncontroversial goals, such as protecting maternal health or improving informed consent. But the ulterior motive is clear: to restrict reproductive rights.
One area of growing focus is reason-based abortion (RBA) bans — prohibitions of abortions based on particular reasons — first sex,7 then race,Reference Susan8 and more recently, Down syndrome and other genetic anomalies.9 As of July, 2021, nearly every state has proposed, and 17 states have enacted, such bans.10 Like other laws intended to chip away at reproductive rights, these laws draw on values that transcend the anti-choice movement, in this case, concerns about equality, disability rights, and preventing “eugenics.” Indeed, when the Court denied certiorari after the Seventh Circuit invalidated an Indiana law banning reason-based abortions, Justice Thomas wrote an impassioned concurrence describing such laws as remedies to the scourge of eugenics.11 Because Justice Thomas has long adamantly opposed constitutional protections of abortion, it is easy to dismiss his diatribe as simply rooted in animosity toward abortion rights.
While some argue that the Court has carved a path that will make it difficult to uphold further state regulations of speech with respect to abortion, others suggest these holdings reflect a form of “constitutional gerrymandering against abortion rights” by twisting First Amendment jurisprudence to achieve a desired outcome. This piece examines what these Supreme Court cases mean for regulations of speech in reproductive care. Specifically, it explores whether states can prohibit doctors from providing certain information obtained through prenatal testing or preimplantation testing of embryos created through in vitro fertilization.
But concerns about equality, disability rights, and eugenics exist on both sides of the political spectrum, not just among the antichoice camps. They are frequently cited in discussions about the societal and ethical implications of the expansion of reproductive testing. It therefore seems plausible, as some scholars have suggested, that reason-based abortion (RBA) bans could be just the first step down a path toward prohibiting the disclosure of some or all information from prenatal testing during pregnancyReference Miller and Berkman12 and preimplantation testing of embryos. Such laws would clearly raise First Amendment issues. Whether they would survive First Amendment challenges is the subject of this piece.
Part I describes current and future forms of reproductive testing. It then briefly delineates the concerns these technologies raise and how legislatures might use them to justify prohibiting disclosures of certain types of information from prenatal testing and preimplantation testing. Part II turns to the confusing Supreme Court jurisprudence on speech in health care as well as the lower courts’ struggles to develop coherent First Amendment principles in this area. Part III attempts to make sense of the contradictory case law to describe the level of scrutiny that should apply to regulations of speech in health care. Finally, Part IV analyzes how potential future laws prohibiting disclosure of reproductive testing results would fare under those approaches. It concludes that they would not survive First Amendment attacks.
I. Reproductive Testing
To set the stage for a discussion of the concerns that might prompt legislative action in this arena, I begin with a brief overview of prenatal and preimplantation genetic testing today and the direction it might go in the future.
A. Reproductive Technologies and Testing
Many forms of prenatal testing, including ultrasound, amniocentesis, chorionic villus sampling (CVS), and the increasingly routine non-invasive prenatal testing (NIPT), are available to pregnant people today. Amniocentesis and CVS involve obtaining fetal or placental cells for genetic analysis to determine whether the fetus has a genetic condition, like cystic fibrosis, or a chromosome anomaly, like Down syndrome (trisomy 21).Reference Farrell13 While some individuals seek prenatal testing to avoid having a child with a particular genetic or chromosomal disorder, to prepare for their future child, or to help physicians plan for possible complications during delivery, others are unsure how they might use the information.Reference Suter14
The newest prenatal test is NIPT, which analyzes fragments of cell-free fetal DNA circulating in maternal blood. It can identify chromosomal abnormalities like trisomy 21, 13, and 18, as well as fetal sex.Reference Mehlman, Rothstein and Suter15 NIPT is not yet truly diagnostic. If, as some anticipate, this noninvasive test ultimately provides the same information as amniocentesis and CVS, but without their risks, it could significantly increase the interest in prenatal testing.16
Reproductive testing can also occur through preimplantation genetic testing (PGT), which involves genetic analysis of an embryo created through in vitro fertilization (IVF). Like amniocentesis and CVS, PGT provides information about chromosomal anomalies and single-gene disorders.17 The results can be used to select embryos to avoid having a child with a genetic disease. While abortion would not be involved, it might result in embryo destruction.
The scope of information available through prenatal testing and PGT will likely expand as our understanding of genetics grows. Scientists are identifying ever more genetic variants associated with several complex diseases and even nonmedical traits. While many such variants may have limited predictive value because they have only a small effect on disease or traits, analysis of the aggregate effect of several variants can be calculated to determine polygenic risk scores for particular diseases or traits.Reference Choi, Mak and O’Reilly18 Polygenic risk scores are not yet part of prenatal testing. So far, only one company, Genomic Predictions, offers analysis of several hundred thousand genetic variants to help parents “prioritize embryos for transfer” based on risks for several polygenic diseases, including diabetes, coronary artery disease, some cancers, and schizophrenia.19 Some, however, question whether the science is good enough at this point to offer meaningful polygenic scores.Reference Regalado20
Polygenic risk scores could also potentially be used in reproductive testing for nonmedical traits, like skin, eye, and hair color; height; or maybe even intelligence.Reference Sulem21 Fertility Institutes in California, for example, advertises itself as “the first and only genetics-based fertility program … anywhere worldwide” that can “offer high level genetic screening of parents seeking to have a voice in determining the eye color of planned children.”22 In addition, two of the founders of Genomic Prediction are searching for genetic variants associated with intelligence and height. One of them recently suggested that “[a]ccurate IQ predictors will be possible” within five to ten years to help couples select the “smartest” embryo for implantation.Reference Devlin23 One study, however, showed PRS has limited predictive value for complex traits like height and intelligence.Reference Karavani24
It is crucial to emphasize that most diseases and traits are the result of a complex combination of genes and environmental factors; that is, genes are often not fully determinative. Even so, some traits, like height, have a strong genetic component, even though environment (diet, health, activity, etc.) can impact their expression.25 Whether PRS will ever meaningfully predict complex traits is uncertain. But as we better understand the relationship between genes and traits, it may well become a part of PGT, even if the results are only probabilistic.
Finally, genome sequencing (identifying all of the base pairs of the genome) might be used with reproductive testing in the future.26 Although not yet a routine part of clinical care, its decreasing costs might change that.Reference Suter27 Interpreting the sequence is the real challenge, however, because our understanding of the entire genome is still incomplete. Nevertheless, with more time and research, our knowledge will undoubtedly grow.
For all these reasons, future reproductive testing will provide more information about the fetus and embryo, potentially even information about minor medical traits, like myopia, and nonmedical traits beyond sex, like height, athleticism, intelligence, etc.
B. Concerns about Selecting Offspring
For some, the concerns about prenatal testing and PGT are rooted in anti-choice views that oppose pregnancy termination or embryo destruction regardless of the reason. But various other concerns have been raised about this technology. One is its potential harm to people with disabilities. Some fear it can lead to fewer births of people with disabilities as well as reduced social support for, negative societal attitudes towards, and heightened discrimination against them.Reference Saxton28
Testing for nonmedical traits raises additional concerns. Sex selection, which is currently possible through prenatal testing and PGT, has altered the normal male to female birth ratio in some countries, although not in the United States.Reference Chao29 Scholars and professional societies worry it reflects “prejudice against female children” or might take us down “a ‘slippery slope’ toward selection of many other traits” that some find “ethically problematic.”30
Several arguments against nonmedical sex selection apply to other forms of nonmedical trait selection. Some fear it denies children a “right to an open future” by imposing expectations associated with a particular trait, which might cause psychological harm or disrupt the parent-child relationship.31 A related fear is that reproductive selection commodifies reproduction, challenging parents’ ability to “appreciate children as gifts” and “not as objects of our design or products of our will.”Reference Sandel32 Finally, some worry that such reproductive selection will exacerbate societal inequities because insurance is unlikely to cover testing for nonmedical traits. They fear the wealthy will be most likely to select for traits, such as height or intelligence, that will increase the societal advantages of their children.Reference Mehlman33
C. Legislation that Goes Beyond Reason-Based Abortion (RBA) Legislation
While counter-arguments can be made in response to the concerns described above, including their speculative nature,Reference Suter34 legislatures might draw upon those concerns to justify limiting patient access to information from prenatal testing and PGT. With the growing focus on RBA bans, anti-choice efforts might commandeer these concerns, as they have done with respect to other values, to limit reproductive rights.Reference Colker35 Legislatures could prohibit the disclosure of information accessible through prenatal testing entirely or before the point of viability to discourage reason-based abortions, particularly because it is difficult to establish a person’s reasons for an abortion.Reference Rebouché36 Laws might also proscribe disclosure of information from PGT to discourage destruction of embryos.
Another possibility might be to ban disclosure of certain information, like information about nonmedical traits or minor medical conditions. States might adopt approaches used by countries like Germany, Austria, France, and Italy, which only allow PGT to prevent serious diseases, or the UK, which bans nonmedical sex selection.37 Such laws would align with public attitudes. While polls find majority support for using PGT to select against lethal, early childhood diseases (72.9%) or diseases that cause life-long disability (66.7%), only a minority support using it to select for sex (21.1%); traits like intelligence (18.9%); characteristics like height, eye color, or athleticism (14.5%); traits like intelligence (18.9%); and sexual orientation (13.3%).Reference Winkelman38
Legislatures could not be faulted for viewing the Supreme Court’s First Amendment jurisprudence in the reproductive realm as condoning such laws. After all, it has shown deference to speech restrictions disfavoring abortion, while applying strict scrutiny to those that don’t. As Part IV argues, however, such restrictions would violate the First Amendment. But before we turn to that analysis, we must first review the judicial landscape in this area.
II. Judicial Treatment of Speech in the Health-Care Context
Courts have tried to decipher the appropriate degree of First Amendment protection for speech in health care given the tension between the states’ power to regulate health care and the First Amendment interests of providers. The Supreme Court has done little to unravel these conceptual knots. In fact, it has spawned continued confusion for the lower courts regarding the extent to which the state may restrict speech in this context.
A. The Supreme Court
The Supreme Court has only examined First Amendment issues related to the speech of health-care professionals in three instances.39 The first was Rust v. Sullivan,40 which addressed federal regulations restricting recipients of Title X family planning funds from offering abortion counseling, referrals, or advocacy of abortion as a method of family planning. The Court rejected the Title X recipients’ First Amendment challenge of the regulations. It first observed that the government may make “‘a value judgment favoring childbirth over abortion, and … implement that judgment by the allocation of public funds.’”41 Because the speech restriction was tied to Title X’s goals to “encourage family planning,” this was not government suppression of “‘a dangerous idea.’” Instead, the government was simply prohibiting grantees from “engaging in activities outside of the project’s scope.”42 Because the regulation did not prohibit health care providers from engaging in abortion counseling or referrals through other programs “separate and independent” from Title-X funded programs, it found no First Amendment issue.43
Despite hinting that speech regulations in the doctor-patient relationship may be unique,44 the Court refused to address the argument that speech within those relationships “should enjoy projection under the First Amendment,” even when the government subsidizes those relationships. Instead, it unpersuasively asserted that “the regulations did not “significantly impinge upon the doctor-patient relationship”45 because physicians were not required to represent views they did not hold. Moreover, the Title X doctor-patient relationship was limited to preconception care, so patients would not expect “comprehensive medical advice.” Because the provider could explain that advice about abortion “is simply beyond the scope of the program,” clients could not interpret “silence” as an indication the physician “does not consider abortion an appropriate option.”46 This was the Court’s first hint that speech restrictions intended to promote childbirth and discourage abortion could survive First Amendment challenges.
Just a year later, in Planned Parenthood v. Casey, the Court addressed several abortion regulations, including an informed consent requirement that physicians performing abortions describe the risks of the procedure, the risks of childbirth, and the probable age of the fetus, and inform patients of the availability of state-provided information about adoption and child support.47 In upholding the informed consent mandate, the Casey plurality focused primarily on whether the regulation violated the Due Process Clause. Relying on its newly crafted undue-burden test, it found the law did not impose a substantial obstacle because the mandated language was “truthful and not misleading.”48
While acknowledging that the statute raised First Amendment issues, the plurality devoted only a paragraph to conclude that the mandated disclosures presented “no constitutional infirmity.” The heart of its argument can be found in one sentence and two citations:49
To be sure, the physician’s First Amendment rights not to speak are implicated, see Wooley v. Maynard, 430 U.S. 705 (1977), but only as part of the practice of medicine, subject to reasonable licensing and regulation by the State, cf. Whalen v. Roe, 429 U.S. 589, 603 (1977).50
In failing to explain the applicable standard of review, this brief and opaque discussion functions like a Rorschach test.Reference Suter51 Some read it as using a rational basis test, and certainly not strict scrutiny.Reference Coleman52 Others see it as employing some kind of intermediate or heightened scrutiny.Reference Dunlap53
Many hoped the Court would clarify its position on speech in health care in NIFLA v. Becerra.54 In that 2018 case, the Court considered the constitutionality of a California statute requiring licensed clinics that offer services to pregnant people to provide specific notices about the availability of “free or low-cost access to comprehensive family planning services … prenatal care, and abortion.”55 The National Institute of Family and Life Advocates (NIFLA), an organization of crisis pregnancy centers, challenged the notice requirements as violating their First Amendment rights to free speech (and free exercise of religion). The Ninth Circuit affirmed the denial of a motion for a preliminary injunction,56 concluding that the notice was a form of professional speech subject to, and likely to survive, intermediate scrutiny.57
Describing the law as intended to regulate crisis pregnancy centers, which “‘aim to discourage and prevent women from seeking abortions’” and are commonly associated with groups that oppose abortion,58 the Supreme Court disagreed. It first noted that content-based regulations of speech are generally “presumptively unconstitutional” and subject to strict scrutiny.59 And it insisted that the mere fact that speech “is uttered by ‘professionals’” does not mean it is not protected.60 In fact, it emphasized that the Court’s precedents have not recognized “a category for ‘professional speech.’”61
Writing for the majority, Justice Thomas had to acknowledge that the Court had, in fact, “afforded less protection for professional speech” before. One exception he pointed to involved the mandated disclosure upheld in Casey. 62 Despite the strong parallels between the speech regulations in Casey and NIFLA, both of which compelled statements about reproductive options, Justice Thomas made a tortured attempt to treat them as distinct. Casey, he stated, involved State regulation of “professional conduct” that “incidentally involves speech” and was consistent with “firmly entrenched” informed consent requirements for operations.63 Allegedly in contrast, NIFLA’s notice requirement applied “whether a medical procedure [was] ever sought, offered, or performed.”64 Thus, it did not “facilitate informed consent to a medical procedure.”65
As Justice Breyer noted in his dissent, this distinction “lacks moral, practical, and legal force.”66 While abortion is “a medical procedure that involves certain health risks,” he emphasized, “carrying a child to term and giving birth” also poses risks. Thus, health “considerations do not favor disclosure of alternatives and risks associated with the latter but not those associated with the former.”67 Further, even if the majority believes that “speech about abortion is special” because it involves “views based on deeply held religious and moral beliefs about the nature of the practice,” Justice Breyer argued, the Court should treat “like cases alike,” especially given the “strong, and differing, views” American hold regarding abortion.68
Justice Thomas did not disguise the Court’s view that speech concerning abortion is special when he distinguished NIFLA from Zauderer v. Office of Disciplinary Counsel of Supreme Court of Ohio, 69 the other instance where he noted the Court applied lesser scrutiny to a law regulating the speech of professionals.70 Thomas observed that Zauderer upheld the state’s discipline of an attorney for failing to disclose the terms of contingent fees in his advertising71 because it involved “‘purely factual and uncontroversial information” about payment for services, and it was not “‘unjustified or unduly burdensome.’”72 Even though the NIFLA notice was also purely factual, the Court remarkably found the compelled disclosures distinct because it mentioned abortion, which the Court described as “anything but …‘uncontroversial.’”73
The Court implied that requiring mention of the “controversial” word abortion raised the “the danger of content-based regulations ‘in the fields of medicine and public health, where information can save lives.’”74 It devoted a full 41.8% of the opinion’s words to describe the threats of regulating speech in health care “to increase state power and suppress minorities,”75 to suppress “‘unpopular ideas or information,’”76 and to “‘maniuplate[] the content of doctor-patient discourse’ to advance … iniquitous interests.’”77 But the Court never explained why compelled disclosures of reproductive options, including prenatal care, contraception, and abortion, impose such a threat, while compelled disclosures of nonmedical options, such as adoption and child support, do not. The implication is that compelled speech that treats abortion as acceptable is dangerous, whereas compelled speech that discourages it is not.
Finally, in an unusual series of decisions that were vacated and replaced by new ones, the Eleventh Circuit addressed a Florida law prohibiting physicians from asking patients whether anyone in their family owned firearms or ammunition. After three decisions upheld the law under different approaches, the Eleventh Circuit en banc invalidated it. Despite the state’s “substantial interest in regulating professions like medicine,” it found the state does not have “carte blanche to restrict the speech of doctors and medical professionals.”
Despite the Court’s strong rhetoric that professional speech should be treated like all other speech for First Amendment purposes, it did not “foreclose the possibility” that there may be a “persuasive reason for treating professional speech as a unique category, exempt from ordinary First Amendment principles.”78 What constitutes a “persuasive reason” is the million-dollar question, making NIFLA the Court’s latest Rorschach test in this area.
B. Lower Courts
The Supreme Court’s opaqueness has left the lower courts struggling to discern the proper standard of review for speech regulations in the doctor-patient relationship. Not surprisingly, “nothing even approaching judicial consensus” exists among the circuit courts.79 Further confusing matters, most of lower court decisions arose before NIFLA, creating uncertainty about the reach of their holdings.Reference Halberstam80 We turn first to decisions concerning prohibitions of speech, and then to compelled speech.
1. prohibitions of speech
The Ninth Circuit was the first to address prohibitions of speech in health care, which it handled differently in two cases by drawing on a speech/conduct distinction. In Conant v. Walters, it found unconstitutional the government’s threat to revoke controlled substance registrations from physicians who recommended marijuana use for medical purposes.81 A decade later, in Pickup v. Brown, however, it upheld a law banning mental health care providers from using sexual orientation change efforts (SOCE) to try to alter minors’ sexual orientations.82 The rationale for more “deferential review” in the latter case was that, unlike the regulations in Conant, which targeted “doctor-patient communications about medical treatment,”83 the restriction in Pickup was a “regulation of the practice of medicine” because it applied to a form of therapy that was not, itself, “an act of communication.”84
The Third Circuit rejected this speech/conduct distinction when it upheld a ban on SOCE in King v. Governor of New Jersey.85 Although it considered SOCE “‘speech’ for purposes of the First Amendment,”86 it reasoned that “speech that occurs as part of the practice of a licensed profession” is not “fully protected by the First Amendment.”87 The court also emphasized that patients “have no choice but to place their trust in” these highly trained and educated professionals.88 To strike a balance between allowing legislatures to prohibit “harmful or ineffective professional services” and preventing legislatures from “too easily suppress[ing] disfavored ideas under the guise of professional regulation,”89 it applied intermediate scrutiny. Because SOCE could harm patients, the ban survived such scrutiny.90
Finally, in an unusual series of decisions that were vacated and replaced by new ones, the Eleventh Circuit addressed a Florida law prohibiting physicians from asking patients whether anyone in their family owned firearms or ammunition.91 After three decisions upheld the law under different approaches,92 the Eleventh Circuit en banc invalidated it.93 Despite the state’s “substantial interest in regulating professions like medicine,” it found the state does not have “carte blanche to restrict the speech of doctors and medical professionals.”94
Most recently, in a post-NIFLA decision, Otto v. City of Boca Raton, the Eleventh Circuit found unconstitutional an ordinance prohibiting therapists from practicing SOCE on minors.95 Unlike the Ninth and Third Circuits, it applied strict scrutiny. Pointing to the speech/conduct used in NIFLA,96 it concluded that the banned therapy is not conduct or a procedure because it is based entirely on speech.97 Further the regulation was a content-based restriction of speech, prohibiting therapists “from communicating a particular message.”98 Quoting NIFLA, it spoke of the “‘inherent risk that the Government seeks not to advance a legitimate regulatory goal, but to suppress unpopular ideas or information.’”99 Finally, it emphasized NIFLA’s and its own refusal “to recognize professional speech as a new speech category deserving less protection.”100
The court found that the law didn’t survive strict scrutiny. First, the state’s compelling interest in protecting minors did not allow it to “‘restrict the ideas to which children may be exposed.’”101 Second, unpersuaded by the government’s assertions of the risks of SOCE, it found the law was not narrowly tailored. Demonstrating a remarkable lack of deference to voluminous research, it concluded that relying on “professional societies’ opposition to speech,” would simply allow “majority preference” to justify speech restrictions.102 Finally, it argued, if the SOCE ban could stand, so could laws prohibiting therapists from validating clients’ same-sex attraction or gender identity,103 completely ignoring that such laws would deviate wildly from professional standards.
The dissent argued for intermediate scrutiny, but believed the law would survive strict scrutiny. Not only did it find compelling the state interests in “protecting minors from harmful professional practices” and regulating the practice of medicine, but it also found the law was narrowly tailored to that goal based on the “mountain of rigorous evidence” that SOCE was harmful and inefficacious in changing sexual orientation.104 Notable for our purposes is the test it advocated for the standard of review. Rejecting the speech/conduct distinction, it proposed that lesser scrutiny should apply when a speech restriction is “auxiliary to” the practice of medicine.105
2. compelled speech
The inconsistency courts have shown regarding speech restrictions applies equally to compelled speech, including laws requiring ultrasounds to be performed on and displayed to people seeking abortions. The Fifth Circuit, in Tex. Med. Providers Performing Abortion Servs. v. Lakey, found such compelled expression “more graphic” than that in Casey, but not “different in kind” because both provide truthful and not misleading information.106 Describing Casey’s standard of review as the “antithesis of strict scrutiny,” the court seemed to apply rational basis in easily finding the law constitutional.
The Fourth Circuit applied decidedly more stringent scrutiny in Stuart v. Camnitz. 107 While finding the ultrasound information to be “the epitome of truthful [and] nonmisleading,”108 it found the regulation “ideological.”109 By requiring the disclosure of “facts that all fall on one side of the abortion debate,” it essentially compelled a “pro-life message.”110 Moreover, the law deviated from informed consent by forcing the display and description of the ultrasound to a pregnant person when she was “most vulnerable.”111 The patient could only avoid it by covering her eyes and ears, thus threatening her psychological wellbeing and undermining her trust in her doctor by making the physician “the mouthpiece of the state.”112
In a post-NIFLA decision, the Sixth Circuit, in EMW Women’s Surgical Center, P.S.C. v. Beshear, upheld a Kentucky mandatory speech-and-display ultrasound law.113 Reasoning much like the Eleventh Circuit, it found no material difference between the mandates in Casey and the Kentucky law. Drawing heavily from NIFLA, the Sixth Circuit concluded that heightened scrutiny should not apply because the ultrasound law regulated “professional conduct that only incidentally burdens professional speech.”114 Moreover, it rejected a “‘sliding scale’ test” for professional speech, which, it noted, NIFLA expressly refused to adopt.115 It also rejected the Fourth Circuit’s pre-NIFLA argument that heightened scrutiny was appropriate because the compelled message was ideological.116
Like the Eleventh Circuit, the Sixth Circuit ignored professional norms in evaluating the speech restrictions. It reasoned that informed consent “may be created by law, as opposed to merely medical custom,”117 and it observed that both Casey and Gonzales v. Carhart 118 upheld medical requirements “directly contrary to alleged medical-professional custom.”119 Thus, the key inquiry was whether the law provided “truthful, non-misleading, and relevant information aimed at informing a patient about her decision to abort unborn life,”120 not “necessarily whether the law is consistent with medical-profession custom or views of certain medical groups.”121 It found that in offering more specific information about the pregnancy than the Casey disclosures, the Kentucky law was “the epitome of ensuring informed consent.”122
The dissent countered that the guiding First Amendment principle in Casey was that the “‘physician’s First Amendment rights not to speak are implicated, but only as part of the practice of medicine, subject to reasonable licensing and regulation by the State.’”123 Thus, together, Casey and NIFLA established “that reasonable regulations that facilitate informed consent to a medical procedure are excepted from heightened scrutiny.”124
In the spirit of the Otto dissent, the EMW dissent pointed to overwhelming evidence that the mandate requires “physicians to violate their professional and ethical obligations” by imposing a “one-size-fits all approach” for informed consent,125 potentially harming patients by forcing them to see images that could cause distress.126 The dissent feared that upholding the mandate would “open floodgates … to manipulate doctor-patient discourse solely for ideological reasons,127 which, it reasoned, violates NIFLA’s admonition that “the state ‘cannot co-opt [physicians] to deliver its message for it.’”128
As we have seen, the courts have offered contradictory and inconsistent approaches to First Amendment doctrine in the context of health care both with respect to prohibited and compelled speech regulations. We turn now to scholarly interpretations of the doctrine.
III. Making Sense of the Cases
A. Scholarly Views of Speech in Health Care
Given the confusion created by the Supreme Court and lower courts regarding the level of scrutiny for regulations of doctor-patient communications, it is not surprising that scholars interpret this body of law in various ways. Some suggest strict scrutiny should apply,Reference Berg129 which seems consistent with NIFLA’s reluctance to afford less First Amendment protection to professional speech. Indeed, the Court cited one of these scholars to argue that strict scrutiny applies to virtually all content-based restrictions, even in healthcare.130 But that standard cannot be the uniform rule given NIFLA’s recognition of the Casey exception.
Many scholars carve out areas of speech in health care that deserve more or less First Amendment protection. Claudia Haupt, for example, argues for “robust First Amendment protection” for speech that would be “acceptable as good professional advice,” whereas speech that falls outside that” acceptable range” should have no First Amendment protection.”Reference Haupt131 Others argue that compelled “ideological” messages should be subject to “rigorous and almost certainly fatal First Amendment scrutiny”Reference Orentlicher132 or that laws regulating professional-client communications about constitutional rights, like abortion rights or the right to bear arms, should be subject to strict scrutiny because they “are, and ought generally to be treated as, regulations of political expression based on content.”Reference Zick133 I have argued that heightened scrutiny should apply to laws regulating informed consent because such speech is central to helping patients exercise their autonomy in making informed medical decisions.134
It is difficult, however, to square these positions with NIFLA’s interpretation of Casey as deferential to an informed consent mandate that 1) dealt with abortion rights; 2) was ideological in discouraging abortion, as the Casey Court itself acknowledged; and 3) required disclosure of nonmedical information, which deviates from typical informed consent doctrine.
Writing after NIFLA, Carl Coleman suggests the level of scrutiny depends on the governmental purpose of the law. If the restriction is “substantially related to … professional quality,”135 it should survive intermediate scrutiny. But if the justification is based on “other governmental interests,” strict scrutiny applies.136 He emphasizes that “the primary justification for regulating professional speech is to counterbalance the inherent knowledge disparity between professional and clients, which makes individuals vulnerable to exploitation by incompetent or unscrupulous practitioners.”137 He does not find Casey inconsistent because it dealt with “factual information that a reasonable patient would arguably want to know.”138 Moreover, although regulations must be “informed by those who have specialized knowledge and experience that laypersons lack,”139 “nontechnical dimensions, including materiality of information to patients,” are also relevant in assessing professional quality.140
Coleman’s view raises questions about what to do when states justify speech regulations in the guise of protecting professional quality. Even more challenging, it does not help courts decide how much laws can deviate from professional norms in regulating professional quality and how much deference should be accorded such norms. It risks inviting the kind of blithe dismissal of professional standards that the Sixth and the Eleventh Circuits demonstrated in dismissing comprehensive research and established medical customs.
Miller and Berkman, also writing after NIFLA, argue that physician speech is “instrumentally high value” speech because of its role in achieving “good medicine,”141 therefore “rational basis is wholly inappropriate.”142 They criticize NIFLA for trying to distinguish physician speech based on whether or not it is tied to a medical procedure. Instead, they suggest, physician speech, should be treated as high value speech or “professional speech — not medical conduct — when it promotes patient safety, occurs within the confines of a doctor-patient relationship, and is supported by evidence-based medicine.”143 While entirely sensible, this doesn’t accord with NIFLA’s implication that the speech in Casey should be accorded less First Amendment protection. Informed consent, after all, falls within their category of high value speech144 — it “promotes patient-safety, occurs within the confines of a doctor-patient relationship,” and is largely supported by evidence-based medicine (to determine material risks).
In fact, the gerrymandering in NIFLA and the two post-NIFLA cases goes beyond reproductive rights and is intertwined with concerns about religious liberties. Thus, strict scrutiny applies to restrictions of speech that conflict with a group’s religious beliefs (such as mandating statements with the word “abortion” or prohibiting SOCE), even if they are informed by and consistent with professional standards. Yet laws that promote a particular perspective, such as an anti-abortion stance, are subject to less scrutiny, potentially even rational basis, no matter how much they deviate from medical customs and professional norms. The Court, it seems, is using the First Amendment to protect and promote certain perspectives, which as Chermerinsky and Goodwin argue, is unconstitutional.
Another view is that NIFLA did not concern professional speech in the context of a doctor-patient relationship. As Haupt points out, someone could enter a clinic, receive the mandated notice, and leave before any relationship was created between a patient and health care provider.Reference Haupt145 The Court, however, never made such a distinction; in fact, it took great pains to emphasize why the speech of professionals generally should not be less protected than other speech. It also spent a great deal of time discussing the dangers of the state controlling communications between doctor and patient,146 suggesting that any regulation of physician speech in the doctor-patient relationship should be subject to strict scrutiny, as long as it is not incidental to regulation of conduct. The Eleventh Circuit certainly adopted that view in applying strict scrutiny to and overturning the SOCE ban.147
Finally, Professors Chemerinsky and Goodwin simply avoid seeking doctrinal consistency between Casey and NIFLA. Instead, they attribute the Court’s different treatment of speech in the context of abortion and speech within a licensed pregnancy clinic to a form of “constitutional gerrymandering against abortion rights” that twists First Amendment jurisprudence to achieve a desired outcome.Reference Chemerinsky and Goodwin148 Under this view, speech regulations aimed at discouraging abortion are subject to deferential review, whereas most other speech regulations are subject to strict scrutiny. Thus, a statute that goes against informed consent norms in mandating the disclosure of nonmedical information to discourage abortion can stand. But mandated disclosures about access to reproductive options (including abortion) that are consistent with medical norms cannot.
As Chemerinsky and Goodwin point out, this inconsistency reflects “not simply a content-based restriction on speech,” but a viewpoint restriction, which is “never allowed.”149 In fact, the gerrymandering in NIFLA and the two post-NIFLA cases goes beyond reproductive rights and is intertwined with concerns about religious liberties. Thus, strict scrutiny applies to restrictions of speech that conflict with a group’s religious beliefs (such as mandating statements with the word “abortion” or prohibiting SOCE), even if they are informed by and consistent with professional standards. Yet laws that promote a particular perspective, such as an anti-abortion stance, are subject to less scrutiny, potentially even rational basis, no matter how much they deviate from medical customs and professional norms. The Court, it seems, is using the First Amendment to protect and promote certain perspectives, which as Chermerinsky and Goodwin argue, is unconstitutional.
B. Attempting to Reconcile the Various Views
In many ways, I am most sympathetic to Chermerinsky and Goodwin. Yet I fear that succumbing to their position allows for continued gerrymandering that could undermine reproductive and other rights and run roughshod over professional custom and medical expertise, as occurred in EMW and Otto. Instead, we must try to find a coherent doctrinal position that balances the First Amendment interests of physicians and patients with the state interest in regulating health care, without allowing the state to use regulations to promote ideologies contrary to medical practice. Trying to thread that needle, given NIFLA, is challenging.
On the one hand, speech within the doctor-patient relationship is of great importance. It is “instrumentally high value” speech because it promotes health,150 and its “[u]ndue regulation” could undermine the “well-being of patients.”151 But the State also has authority to regulate the “practice of medicine,” as Coleman suggests, to protect patients who are “vulnerable to exploitation by incompetent or unscrupulous practitioners” given the knowledge disparity.152 Thus, some form of intermediate scrutiny should apply.
The line NIFLA draws between speech and conduct, however, challenges this view as do the two post-NIFLA appellate decisions — one, applying strict scrutiny to speech qua speech,153 and the other, applying something like rational basis to speech incidental to a procedure.154 The NIFLA Court, however, never offered a persuasive rationale to explain why the level of scrutiny depends on whether the regulated speech is tied to a medical procedure.155 It pointed to precedents that “have long drawn” a “line between speech and conduct,” even though none of them distinguished speech incidental to conduct and speech as speech in health care. It also noted the state’s authority to regulate professional conduct, as exemplified by malpractice torts.156 But of course, the state’s interest in regulating speech in the doctor-patient relationship exists whether or not a procedure is involved.157
The dissenting opinions in Otto and EMW offer a potential framework for deciding when a less stringent standard of review applies to speech restrictions like that in Casey. The EMW dissent identified the “‘practice of medicine’” as “the driving term” in Casey. 158 Building on this idea, the Otto dissent looks to whether the “affected speech is ‘auxiliary to’ or ‘inconsistent with’ the practice of medicine.”159 When a law regulates speech “auxiliary to a medical practice,” it should receive more deferential review, but when it is “inconsistent with the practice of medicine,” heightened scrutiny should apply.160
While the speech mandate in Casey is not consistent with informed consent norms in requiring disclosure of nonmedical information intended to discourage abortion, it only required the doctor to mention the availability of a state-created document with nonmedical information about adoption and child support. It did not require physicians to actually make statements inconsistent with informed consent or to conduct procedures that were not medically indicated. Thus, under the law, physicians could speak in a manner consistent with the professional and ethical norms of informed consent. Reading the law as mandating mention of the availability of such information, but not as mandating actual disclosure of non-medical information, offers a way to find it consistent with informed consent norms.
I readily concede that this distinction is somewhat forced, but it offers a way to understand the doctrine that does not allow states to profoundly distort medical practice “solely for ideological reasons.”161 Wholesale acceptance of legislative disregard for the standard of care when regulating speech is problematic, particularly when driven by ideological concerns. Under the approach I advocate, speech restrictions wholly inconsistent with medical practice should be subject to heightened scrutiny. This approach thus avoids the troubling lack of deference to professional expertise demonstrated in Otto and EMW.
With respect to regulations of speech auxiliary to the practice of medicine, intermediate scrutiny is appropriate. As discussed above, it accommodates the tension between important First Amendment values and the States’ interest in regulating medicine. Some have even suggested that NIFLA implied that intermediate scrutiny applies to regulations like those challenged in Casey when it emphasized that the notice requirement did not even survive such scrutiny.162 While one might argue that the Court’s terse support of the informed consent statute in Casey hinted at a rational basis test, it is also possible the plurality thought the law easily satisfied intermediate scrutiny, believing (as it seemed to) that the mandated disclosure was consistent with informed consent practices.163 Thus, intermediate scrutiny in this context can be reconciled with Casey and NIFLA.
Although the speech/conduct distinction articulated in NIFLA and two circuit courts is problematic, it may nevertheless become binding. Thus, I also consider how that distinction might play out with respect to laws banning disclosure of information from reproductive testing. But even if that line holds, one question remains. Are there only two relevant categories of speech in health care — speech incidental to conduct and speech qua speech — or is the second category really two categories — speech qua speech within the doctor-patient relationship and speech qua speech outside the doctor-patient relationship? If three categories exist, intermediate scrutiny would apply to speech incidental to conduct because the state has wider authority to regulate medical conduct than speech,164 heightened intermediate scrutiny would apply to speech as speech in the doctor-patient relationship, while strict scrutiny would apply to the last category. Under these different theories, therefore, speech regulations within the doctor-patient relationship should receive at least intermediate scrutiny.
IV. Evaluating Prohibitions of Disclosures of Prenatal and PGT Information
A. Standard of Review
As noted in Part I, legislatures might commandeer concerns about disability rights, equality, commodification, and eugenics to prohibit the disclosure of certain information that could be obtained through prenatal testing or PGT. To assess whether such laws would survive First Amendment challenges, the first question is the level of scrutiny.
If the NIFLA Court’s distinction between speech incidental to procedures and speech as speech is binding, the outcome is not immediately clear. Are laws that prohibit the disclosure of or analysis of certain types of genetic information restrictions on speech qua speech or speech incidental to a medical procedure? These potential speech regulations are clearly distinct from laws prohibiting SOCE therapy or restricting health care providers from making inquiries about gun ownership, neither of which centers around a particular medical procedure.
One might argue they are very much “tied to a procedure” because medical procedures — amniocentesis, CVS, drawing blood for NIPT, or retrieving eggs for IVF/PGT — are necessary to obtain the information. Viewed that way, intermediate scrutiny would apply. However, the information is not incidental to the procedure; it is derived from a procedure. Obtaining that information is the very purpose of the procedure. Thus, information gleaned from prenatal testing and PGT is speech qua speech, not speech incidental to a procedure. Under a theory that does not distinguish between speech within the doctor patient relationship and other speech, strict scrutiny should apply because these laws are content-based regulations. Even if NIFLA does allow some lesser protection of professional speech, as opposed to other speech, heightened scrutiny should apply because this is speech qua speech.
If a court were instead to determine the level of scrutiny by asking whether the regulation is auxiliary to the practice of medicine, à la the Otto and EMW dissents, the analysis would be different. Under this test, laws prohibiting access to information about genetic risks from prenatal testing or PGT would violate clearly established standards of care. The very purpose of these procedures is to provide information so individuals can make reproductive decisions — e.g., whether to terminate or continue the pregnancy or to select embryos to prevent disease in a future child. Prohibiting disclosure of this information would render the procedures worthless. Indeed, the standard of care is so clearly established here that most jurisdictions allow wrongful birth claims when providers negligently fail to deliver correct information from reproductive testing.165 Laws banning disclosure of this information would therefore be highly inconsistent with medical practice and should fail under First Amendment scrutiny.
But what about information that is not directly related to medical risks, such as sex or nonmedical traits?166 While information about fetal sex from prenatal testing is routinely disclosed, the purpose is not to facilitate decisions regarding prenatal care, whether to continue a pregnancy, or decisions regarding delivery.167 Instead, the information is provided because of its personal and social value to some parents. Failing to provide such information would likely not be the basis for a wrongful birth claim. After all, what would the damages be?168
In the context of PGT, only a slight majority of clinics offer nonmedical sex selection,Reference Capelouto169 and very few offer other kinds of nonmedical trait selection.170 Even though information about sex (and potentially other nonmedical traits) is far more likely to influence embryo selection than decisions about pregnancy termination, at least in the United States, no clear standard of care exists regarding nonmedical trait selection through PGT. Indeed, professional organizations seem ambivalent about the ethics of nonmedical sex selection.171 Thus bans on disclosure of nonmedical information about fetuses and embryos are not “inconsistent with the practice of medicine,” suggesting they should be subject only to intermediate scrutiny.Reference Weaver and Lively172
B. Applying the Standards
As we saw above, the level of scrutiny for these laws ranges from intermediate to strict. We begin by analyzing them under intermediate scrutiny, which requires a showing that a “statute directly advances a governmental interest and that the measure is drawn to achieve that interest.”173 The regulations need not be perfectly tailored to the important state interests, but if the restrictions do not sufficiently advance those interests, they cannot survive.174
To evaluate the laws, we must first articulate the state interests. The state would likely assert three, the first being the promotion of fetal and embryonic life.175 The Supreme Court has described the interest in fetal life as “legitimate and substantial.”176 While it focused on the state’s interest in life from the outset of pregnancy,177 courts would likely also find a legitimate and maybe even substantial interest in ex vivo embryos given their potentiality for life.
Second, the state might assert interests in promoting social values and preventing “morally repugnant” acts,Reference Suter178 in this case, what Thomas calls “eugenic-like” practices.179 It might also assert an interest in preventing negative social effects, including discrimination based on sex, disability, or disfavored traits. A related goal might be preventing a reduction in the number of children born with disabilities or less desirable traits180 and the exacerbation of inequalities if those with more resources are better able to select for traits that confer social advantages.181 Finally, the State might want to discourage parents from treating their children as products whose quality must be controlled.182
Courts would likely view the state’s interests in addressing troubling social values and societal effects such as alleged eugenic uses, discrimination, prejudice, and commodification of reproduction as legitimate, and perhaps substantial, interests, particularly when considered in the aggregate. Even if those state interests are substantial, however, preventing physicians from disclosing information they would otherwise disclose to patients under the medical standard of care is not sufficiently related to these interests to satisfy a heightened or even intermediate standard of scrutiny for several reasons.
First, such legislation would be overly broad, at least with respect to prenatal testing. Not all people seeking such testing would terminate based on prenatal information. Patients may want information to prepare for having a child with certain traits (including gender) or a disability.183 In addition, some prenatal information is central to prenatal care and birthing.184 Thus, these laws would restrict access to information that can be of great personal and medical value to patients, without protecting fetal life when the patient wasn’t considering termination. In addition, sometimes such laws might result in fetal loss. A couple at risk for a serious genetic condition, for example, might terminate the pregnancy, rather than risk passing on a serious disease gene. Thus, they could potentially end a pregnancy with an unaffected fetus they would not have terminated if they had had access to the prenatal information.
While information from PGT almost always influences which embryo is implanted (that is, after all, why people seek PGT), a ban on disclosure of information may not actually spare many embryos. Because IVF often results in more embryos than can be implanted, information from PGT usually affects which embryos are implanted, but not how many; it is not likely to influence decisions about whether to destroy embryos or donate them to infertile couples.
Bans on disclosure of this reproductive information would also be too broad to address the state interest in social values and effects, particularly in the context of prenatal testing. Even if the information were used to decide whether to terminate a pregnancy, not all (and perhaps not most) choices to terminate pregnancies based on prenatal information are rooted in prejudice or commodification of children. A pregnant person may decide, for example, to terminate a pregnancy based on a condition like Tay Sachs, not because of prejudice or because she views her child as a product. Instead, she may want to prevent suffering or have concerns about her emotional and/or financial capacity to care for a child with a disability.
One might defend the laws by pointing to the dramatic decline of children born with Down syndrome in Scandinavian countries.185 Although not nearly so stark, the numbers in the United States are not insignificant.Reference de Graaf186 Even so, prohibiting disclosure of prenatal information normally disclosed as part of the standard of care is not a useful way to address these behaviors. First, under the approach I advocate, the law would be subject to strict scrutiny because it would deviate from the standard of care. Second, far less intrusive and more effective measures exist. States could educate the public about Down syndrome (or other disabilities) or provide relevant information about the condition when prenatal testing identifies it.187 Most important, they could offer adequate educational and other support for children with disabilities so that having such children would feel like a viable option to parents.
Nor do worries about the aggregate effect of embryo selection based on disease, sex, or other nonmedical traits support such bans. Given IVF’s high cost, PGT is not likely to become widespread. And although wealthier people could more easily access PGT, potentially exacerbating social inequities, the physical burdens of egg retrieval would likely discourage many of them from using PGT, especially for minor diseases or mere traits. Indeed, polls suggest only a minority support embryo selection for purposes other than avoiding serious disease. Moreover, societal inequities due to wealth disparities may be more profound than those based on genetics. Studies have shown that household income is far more predictive of future success than genetics.Reference Dam188 Thus, addressing income inequality through something like child tax credits would do far more to prevent exacerbation of inequities than banning information from PGT.Reference Pulliam and Reeves189
For all these reasons, even if the state interests motivating such bans are deemed substantial, these potential laws would not advance those interests in a meaningful way. Because the laws would struggle under intermediate scrutiny, they would surely fail under strict scrutiny. As a starting point, the state interests are not compelling. Preserving fetal life only becomes compelling at viability,190 but prenatal testing usually occurs before viability and PGT before a pregnancy is even established. Further, the Supreme Court has not expressly described a state interest in the social values and effects legislatures might point to, suggesting they too are not compelling. Finally, because the laws are not closely enough drawn to the state interests for intermediate scrutiny, they clearly are not narrowly tailored to those interests.
V. Conclusion
Given the increase in reason-based abortion bans, it seems entirely possible that some legislatures may restrict physicians from disclosing information obtained through prenatal testing and PGT based on concerns about eugenics, disability rights, commodification, and equality. While cognizant of the First Amendment doctrinal morass regarding speech in health care, I nevertheless attempt to offer a consistent and coherent interpretation of NIFLA and Casey. Under that approach, such laws would violate the First Amendment.
I end by noting a few key issues left unexplored in this piece, given space constraints. First, should the law treat compelled speech differently from restricted speech? Courts rarely raise this issue and the “Supreme Court has been deliberately noncommittal”191 about it, despite suggesting the distinction is not constitutionally significant.192 Second, how should legislatures and courts grapple with the challenges and normative elements of drawing lines between medical and nonmedical conditions — a line that informs the analysis? Nor do I fully grapple with the political elements that may shape understandings of the standard of care or what constitutes a medical condition. Finally, I do not address the variations of intermediate scrutiny that courts have deployed. I hope to address these issues in a future project that will propose a theory of First Amendment analysis for speech regulations in health care generally and that avoids potential constitutional gerrymandering of the First Amendment.
Acknowledgements
Many thanks to Catherine Ross, Radhika Rao, Rachel Rebouché, Barbara Billauer, Doron Dorfman, as well as participants in the GW Law Symposium “First Amendment Values in Health Care” and the Virtual Health Law Workshop. I am also grateful to Alexandra Marshall and Heather Skrabak for their excellent research assistance.
Note
The author has no conflicts to disclose.
