Late-Term Abortion and Medical Necessity: A Failure of Science

Abstract

Roe V. Wade (1973) placed the concept of medical necessity at the center of the public discourse on abortion. Nearly a half century later, 2 laws dealing with late-term abortion, 1 passed in New York and 1 set aside in Virginia, are an indication that the medical necessity argument regarding abortion has been rendered irrelevant. More importantly for this discussion, these laws are an indication of the failure of the US scientific and medical communities to inform this consequential topic with transparency, logical coherence, and evidence-based objectivity.

Roe V. Wade (1973) placed the concept of medical necessity at the center of the public discourse on abortion. Nearly a half century later, 2 laws dealing with late-term abortion, 1 passed in New York and 1 set aside in Virginia, are an indication that the medical necessity argument regarding abortion has been rendered irrelevant. More importantly for this discussion, these laws are an indication of the failure of the US scientific and medical communities to inform this consequential topic with transparency, logical coherence, and evidence-based objectivity.

Essentially, Roe allowed abortion without any regulation in the first trimester of pregnancy, but made abortions in the second and third trimesters contingent upon demonstrated threats to the pregnant mother’s health. Along with rape and incest, therefore, medical necessity became the pathway to unrestricted abortion access. It should be noted that varying definitions of medical necessity for abortion have ricocheted along a continuum with consideration of a “broad range of physical, emotional, psychological, demographic, and familial factors relevant to a woman’s well-being” at one extreme and “conditions which place a woman in danger of death” at the other. ^1,2 However, while the occasional politician or news reporter will still indicate that late-term abortions are most often performed in the case of “severe fetal anomalies” or to “save the woman’s life,” the trajectory of the peer-reviewed research literature has been obvious for decades: most late-term abortions are elective, done on healthy women with healthy fetuses, and for the same reasons given by women experiencing first trimester abortions. The Guttmacher Institute has provided a number of reports over 2 decades which have identified the reasons why women choose abortion, and they have consistently reported that childbearing would interfere with their education, work, and ability to care for existing dependents; would be a financial burden; and would disrupt partner relationships. ³ A more recent Guttmacher study focused on abortion after 20 weeks of gestation and similarly concluded that women seeking late-term abortions were not doing so for reasons of fetal anomaly or life endangerment. The study further concluded that late-term abortion seekers were younger and more likely to be unemployed than those seeking earlier abortions. ⁴ It is estimated that about 1% of all abortions in the United States are performed after 20 weeks, or approximately 10 000 to 15 000 annually. Since the Roe framework essentially medicalized abortion decisions beyond the first trimester, and since abortions in the United States are now performed on demand and only rarely for medical reasons which could end the life of the mother, what can we conclude about the value and impact of medical necessity determination in the case of induced abortion? A prescient proabortion author predicted today’s events with remarkable foresight when he concluded that the “rhetoric of medical necessity” is a mistaken strategy because “it is not the empirical evidence of what is or is not medically necessary which is important,” but rather “who possesses the ability to interpret necessity within key political contexts.” ⁵ When viewed from this perspective, it is possible to see the recent New York and Virginia legislation as a signal that politics, not science, is the most powerful influence on abortion issues and legislation.

Our medical, public health, and science organizations have become politicized. The independent guidance we might reasonably expect from the science community on these consequential health policy conundrums is absent—evidence replaced by advocacy. To illustrate this science chasm on abortion within the medical and science establishments, note the starkly different responses to the new Complex Family Planning subspecialty recently approved by the American Board of Obstetrics and Gynecology (ABOG) to treat women who have pregnancies that are abnormal, unintended, and/or unwanted. In its request for approval of the new subspecialty, the ABOG was careful to provide as a rationale both “the increasing emphasis on healthy child spacing” and “potentially life-threatening medical conditions”—a not very subtle endorsement of abortion on demand uninhibited by considerations of medical necessity for each and every abortion. ⁶ The American Association of Pro-Life Obstetricians and Gynecologists, by contrast, maintains that late-term abortions are “never necessary” and that the new subspecialty is merely an “attempt to increase the number of board-certified OB-GYNs trained to perform second and third trimester abortions.” ⁷

The abortion science infrastructure, defined by the availability of valid data and research funding, is woefully inadequate. The abortion reporting system in the United States is demonstrably limited. The Centers for Disease Control and Prevention (CDC) Abortion Surveillance System is voluntary. Three states (California, Maryland, and New Hampshire) do not report at all, and they account for 20% of the total US abortions. There is no uniformity in the data provided by reporting states so that major variables of interest such as race are available for only subsets of states and for intermittent time periods. Data are provided in aggregate tables and no individual event-level information is available, seriously constraining its analytical power. As a result, CDC abortion reporting has very limited value as a research resource. The Guttmacher Institute provider and patient surveys are likely a more complete source of abortion incidence data because they survey all states. However, Guttmacher does not do a survey every year, nor is its data openly available to all investigators. These inadequacies in abortion reporting also erode the quality of science in related areas such as maternal mortality. ⁸ Research funding for abortion studies is also inadequate. According to the National Institutes of Health Research Condition and Disease Categorization System, developed to provide transparency in the reporting of funded research, abortion is essentially nonexistent as a subject for research funding. The system annually reports 282 different, presumably researchable conditions and categories including climate change, adolescent sexual activity, eczema, and food allergies. However, there is no category for abortion. A query of the system using abortion as the search term returns the following response: “No estimate of funding information is found.” So, the available data are compromised and minimal. Available funding is sparse. In the nation which has revolutionized the application of data and analytics, no one can say with certainty how many induced abortions are performed, what are the characteristics of the affected population of women, nor the characteristics, volumes, and outcomes of the providers who have performed them. There is no vibrant transparent exchange of data, findings, and policy interpretations occurring in the US peer-reviewed professional and scientific journals.

If an induced abortion is healthcare, still a widely debated question, then the procedure must meet the requirements of being medically necessary. Exempting abortion from the test of medical necessity essentially relinquishes any claim that it is health care. While the concept of medical necessity has been defined in myriad ways, a few key elements present in all of the definitions across a range of medical specialties are especially relevant in the context of induced abortion:

The service must be required to prevent, diagnose, or treat an illness, injury, or disease. Pregnancy is neither an illness nor a disease and, following conception, is no longer preventable. Therefore, the treatment (abortion) must target another specified illness, injury, or disease.

The service must be clinically appropriate and considered effective for the individual illness, injury, or disease. This requirement implies that credible, evidence-based peer-reviewed literature exists that the abortion procedure will produce a positive result on specified outcomes related to the pregnant woman’s illness, injury, or disease. In many states, the official language of the medical necessity determination form is too vague to allow such treatment-to-outcome specificity. In New Jersey, for example, physicians may consider “physical, emotional, and psychological factors” in determining whether a termination of pregnancy is medically necessary. There are specific clinical criteria available for determining the medical necessity for psychiatric treatment: a diagnosed disorder; which can be improved by the treatment based on accepted medical standards; presence of the illness documented by Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition) codes assigned; and determination made by a licensed mental health professional. ⁹ Too often, these assessments are neglected or superficially completed using inappropriate documentation and by persons without appropriate credentials and experience. ¹⁰

The service is not primarily for the convenience of the individual, the individual’s health-care provider, or other health-care providers.

The specific steps which would allow the determination of medical necessity for all second and third trimester–induced abortions are easily evident. Every abortion must be reported by an appropriately trained medical professional(s); the illness or disease condition presumably ameliorated by the abortion should be clinically defined and the patient history appropriately documented; the expected effect of the abortion on the illness or disease should be specified in terms of measurable outcomes as well as necessary follow-up care; and a statement of the preferred cost-effectiveness of abortion to alternative treatments should be rendered. Of course, one would also advocate that there would be uniform necessity criteria and required universal reporting across all states. Event-level data, comprehensively and uniformly reported as described, would provide researchers with the ability to inform the public discourse of the determinants, correlates, outcomes, and prevention of late-term abortion. It will never be possible, nor is it likely desirable, to remove politics and ideology from the discussion of abortion-related public policy. However, it is possible for science to fulfill its unique and essential role in returning valid, objective, evidentiary findings to the consequential public dialogue on abortion.