Despite a common perception to the contrary, the scientific domain of medicine is as subject to bias as any other domain. Indeed, much of the emphasis in clinical research in recent years has been directed toward standardizing study design to remove bias in clinical investigation. But there are certain sources of systematic bias in the medical community that influence the ability to pose moral questions fairly and have far—reaching consequences for all who come in contact with the medical establishment. One particular source of bias arises from merging the legal and political dimensions of the abortion debate into medical judgment and decision-making.
My practice and clinical research are in a subspeciality of obstetrics and gynecology, maternal fetal medicine, which concerns pregnancies complicated by maternal or fetal disease. There are few situations more daunting to those who advocate a consistent ethic of life than the circumstance in which the life of the mother is threatened by the continuation of the pregnancy. Although I do not acknowledge this conflict as justifying abortion, even the most dedicated of advocates for the life of the unborn are awed by this dilemma. Indeed, the power of this image has been one of the principal forces advancing the abortion movement in the United States and elsewhere. What do we know about it objectively?
Certain conditions that can be diagnosed in advance are associated with risk of maternal mortality greater than 20 percent: pulmonary hypertension (primary or Eisenmenger's syndrome), Marfan's syndrome with aortic root involvement, complicated coarctation of the aorta, and, possibly, peripartum cardiomyapthy with residual dysfunction. Taken altogether, abortions performed for these conditions make up a barely calculable fraction of the total abortions performed in the United States, but they are extremely important because they have been used to validate the idea of abortion as a whole. They stand as a sign that abortion is in some cases unavoidable—that it can be the fulfillment of the good and natural desire of the mother to live.
It should be emphasized how rare these conditions are. Our obstetric service in the Los Angeles area has been the largest in the United States for most of the last fifteen years, averaging fifteen thousand to sixteen thousand births per year. Our institution serves a catchment for all high-risk deliveries in an area with thirty thousand deliveries per year. Excluding cases that have been diagnosed late in pregnancy, we do not see more than one or two cases per year that pose this degree of risk of maternal mortality; these are exceedingly rare conditions. This rarity does not diminish the tragic dimension of such cases, but the cases are seen in perspective when their numbers are compared to the total number of abortions performed.
If we examine other conditions associated with lesser though still significant risk of maternal mortality (conditions for which abortion is often recommended), we find that in many cases the prognoses are changing, both because of a better understanding of the natural history of the disease and because of advances in therapy. Here is the paradox, however. As the actual risks to the mother diminish because of medical advances, concern about maternal and fetal risks from complications of pregnancy is still offered as a justification for many abortions. From the case histories that follow, the distorted milieu of medical practice into which most pregnant women now enter can be shown.
A twenty-one-year-old woman in her nineteenth week of pregnancy was referred for “immediate abortion.” She had complained of shortness of breath and a full evaluation revealed a complex maternal congenital heart lesion, tetrology of Fallot. This is a lesion frequently listed as a contraindication to pregnancy because of increased risk of maternal mortality. The senior house officer coordinating the abortion asked for a second opinion from our high risk clinic because the patient was distraught over the recommendation. Despite having been told that she had a significantly increased chance of dying if she remained pregnant, she had not reconciled herself to the abortion.
Repeat evaluation confirmed the diagnosis but showed that the particular manifestations of the condition in this patient were such that she could be expected to tolerate pregnancy without difficulty. She was a “pink tet,” tetrology of Fallot in which the patient is still receiving adequate oxygen. This could have been determined by the referring physician. As it was, the patient very nearly underwent an abortion for a non-indication. The patient delivered without significant complications following induction of labor in her thirty-eighth week of pregnancy.
A twenty-five-year-old woman in the twelfth week of her first pregnancy had shortness of breath and was found to have severe narrowing of one of the valves of the heart, mitral stenosis. Her physician recommended abortion and asked for confirmation of his decision in a phone conversation. We suggested that the patient be offered the opportunity to discuss balloon valve repair during pregnancy with a cardiologist skilled in that technique. We provided references showing that this could be accomplished safely in pregnancy. Her physician expressed concern about his liability if there were any abnormality of maternal or fetal outcome. “She's young,” he said. “She can have the valve repaired and try again.” The patient was never referred for formal consultation.
A thirty-eight-year-old woman was referred to us by her pastor. She was eleven weeks pregnant and was found to have breast cancer with spread to the regional lymph nodes. She was told that, for the best chance of long-term survival, she should undergo chemotherapy, but that the pregnancy should be terminated first. She was told that her prognosis would be worse if she remained pregnant and that the chemotherapy would definitely harm her baby. Her abortion had been scheduled.
We discussed with her published evidence that breast cancer is not affected by pregnancy and that the chemotherapy regimen required for her condition is apparently well-tolerated by the fetus. The experience with any given chemotherapy regimen is limited, and we were frank with the patient that there were open questions about long-term effects. When her physician was informed of the patient's desire to undergo chemotherapy and continue the pregnancy, he suggested that we take care of her and accept the liability.
The patient underwent chemotherapy (Adriamycin and Cytoxan) and delivered a baby boy who appeared entirely normal at birth. That many chemotherapy regimens can be continued without apparent ill-effect in pregnancy is information readily available to any interested physician, but the patient was not informed.
A twenty-year-old woman in the eighteenth week of her first pregnancy arrived with severe renal disease that appeared to be due to new onset of systemic lupus erythematosus. The first consultant recommended abortion for her own health and out of concern that the medication required to control her disease might injure her fetus. The patient was anxious not to abort. We were able to tell her that although the chance of successful pregnancy outcome was low, abortion would not predictably affect the course of the disease. We discussed the considerable experience available with the medications she would require (principally steroids) and the fact that there were no apparent serious fetal effects related to this type of treatment. The patient elected to continue her pregnancy.
Subsequently, a kidney consultant recommended kidney biopsy as part of efforts to confirm the diagnosis—but not until after an abortion. We presented data showing that renal biopsy can be accomplished safely in pregnancy and that the need for this test should not be considered an indication for abortion.
Finally, the patient required a lengthy procedure under X-ray fluoroscopy. The radiologist recommended abortion because of significant X-ray exposure. After consultation with the radiation physicist, however, it was clear that the actual X-ray exposure to the fetus in this case posed no significant risk. The patient ultimately delivered a premature infant at twenty-seven weeks gestation. The child did well for one week until dying, suddenly, of an overwhelming infection.
A thirty-two-year-old nurse had herself tested for cytomegalovirus in the seventh week of her pregnancy. This is a type of virus known to be capable of crossing the placenta and infecting the fetus, sometimes resulting in retardation and multi-organ system disease, especially if the infection has occurred for the first time in pregnancy. The results of her testing profile suggested a recent infection, indicating that her baby could be infected. She was advised to terminate the pregnancy and had made plans to do so, although with great regret. Her doctor stated that he had confirmed his recommendations with a “high-risk pregnancy specialist.”
She was referred to us by a neighbor who was a physician. On initial review of the tests, it did appear that there had been an acute infection during pregnancy. We presented the patient with data that the likelihood of her child being seriously affected was only four in a hundred, with half of these suffering only isolated hearing loss. She was stunned and relieved to learn that the risk was no greater. As it turned out, a more specific indicator of infection (which we recommended be checked before any decision be made) revealed that there had been no infection at all. She delivered a healthy boy at term. She refers frequently to her “miracle baby,” a pathetic reflection of the circumstances that nearly took the baby from her.
All this patient received from us was an accurate assessment of the risk to her child. And that was enough for her to continue the pregnancy, even before she learned that there had been no infection. It might not have been for the next woman. In fact, the same woman could have had an abortion for any reason at all the next day. But it would not have been under the pretense of a medical indication. One goal of our clinic is simply to restore a rational medical assessment to these issues. A strict focus on the justification for abortion draws attention to the way in which medical judgment has been vitiated.
The recommendations for abortion in this and the earlier cases were partly the result of ignorance of the data, but also of something else—a belief that it is better to err on the side of abortion if there are doubts about the effect of the pregnancy on maternal or fetal outcome.
These cases are just some that we have seen in the last year. They include only women who, usually because of their own convictions, could not easily accept the recommendation for abortion and sought more information. There is no doubt that many others have received such recommendations and proceeded to abort simply on the basis of the doctor's authority.
But the significance of these cases is not in the stories themselves, however disturbing they may be. The cases are insignificant compared with the total number of abortions, and the ethical dilemma in such cases is commonly understood in a context that accepts that abortion would indeed be appropriate were there a significant risk to the mother or were the fetus seriously malformed. The real significance of these cases is in what they reveal about the attitudes of the physicians these women first encountered.
Why are physicians not providing readily available information that could affect their patients' judgment regarding abortion? I believe that there are two related reasons for this phenomenon and that they go much deeper than simple ignorance of the facts. One is the transference of an ambivalent attitude toward the developing human (virtually codified in Roe v. Wade) into the medical arena. Since the fetus can be aborted for any reason, the physician may see fit to suggest or recommend abortion for almost any reason.
The basis for this attitude is closely linked to a second concern: the unbalanced legal burden of informed consent. When a mother has a major medical problem in pregnancy (or indeed any medical problem), the medical record must reflect the patient's informed consent to continue with the pregnancy despite the risks. If doctors fail to disclose those risks, they are negligent because the patient could have chosen abortion as a different course of treatment. To compound the problem for the physician, there are no clear legal guidelines to determine which risks do not warrant communication to the patient. A fact is considered to be material if a reasonable, prudent person in the position of the patient would attach significance to it in deciding whether or not to submit to the proposed treatment. The accurate assessment of the risks in a given case can be a tedious process. Should any untoward outcome result from the pregnancy, the record will be scrutinized intensely. No method of documentation is watertight.
The doctor's alternative—to suggest or recommend that abortion is the safest route—carries no such legal liability. There does not appear to be a legal precedent for a physician's liability in a case where abortion was recommended on supposed medical grounds that were subsequently found to be baseless or misrepresented.
With regard to fetal abnormalities, the burden is equally one-sided and even more clearly delineated in law. Physicians have legal and, some would say, ethical duties to inform pregnant women of prenatal tests that would affect their willingness to continue the pregnancy. The concept of “wrongful birth” in law establishes that failure to inform of tests that are widely accepted in the medical community as part of the standard of care could lead to legal liability. The related concept of “wrongful life,” although less commonly invoked legally, is instructive for discovering the idea behind the law. In such cases, the child brings suit, claiming that it would be better not to have been born than to have been born with defects.
The concept of informed consent, simple in theory, is almost impossible in practice. For many physicians it translates into simply recommending every possible test and erring on the side of suggesting abortion whenever there is a question of risk to the mother or child. There is a tremendous imbalance between the liability involved in not informing the mother of risks compared to the liability of suggesting the alternative of abortion. All pregnant women, no matter what their personal convictions, are subject to the effects of this imbalance.
Real or even perceived risk to the mother or child is actually relatively infrequent. Much more common in practice is the situation in which a presumably healthy mother and fetus are offered screening tests in the hopes of identifying congenital anomalies of the fetus. Maternal serum alphafetoprotein (MSAFP) screening, capable of identifying fetuses with neural tube defects or Down's syndrome, was introduced in the early 1980s. Its place in practice was virtually mandated by the 1985 liability alert from the American College of Obstetricians and Gynecologists, a direct response to the perceived liability from wrongful birth. Every physician must inform patients about these tests or be liable for the results. The patient may refuse the test, but usually must make a positive statement to that effect. The inescapable implication is that the woman who refuses the test is outside the norm.
More recently, various tests of maternal blood designed to identify fetal Down's syndrome have been introduced into clinical practice. The number of tests on the horizon that will allow identification of other fetal abnormalities appears limitless. The ineluctable logic of these legal precedents affects every pregnant woman and her child. No matter what the personal convictions of the mother, she must receive her care in a system in which every possible problem of maternal or fetal well-being is a test of whether the pregnancy will be allowed to continue. And in that balance, the developing human has little or no value. There is no counterweight to “wrongful birth.” There is no “wrongful abortion.”
Although few women may actually abort because of this bias, many will learn the lesson: that a new human being is accepted conditionally, one test at a time.
Thomas Murphy Goodwin, M.D., is Assistant Professor in the Department of Obsterics and Gynecology at the University of Southern California's Women's Hospital and is Director of Maternal-Fetal Medicine at the Hospital of the Good Samaritan in Los Angeles.