INFANTS, MEDICAL ASPECTS AND ISSUES IN THE CARE OF

Overall, there are three relatively distinct groups of babies who are admitted to neonatal intensive care units (NICUs). These groups are different in ways that are ethically relevant. The first group consists of full-term or near-term babies with acute illnesses such as pneumonia or sepsis, as well as babies with surgically correctable anatomic abnormalities. The second group comprises babies with congenital anomalies, including chromosomal anomalies, which are not correctable at the present time. Many of these babies have problems that can be ameliorated but not corrected with surgical or medical treatment. The final group of babies includes those born prematurely who are otherwise physically normal— that is, they have no acute illness or congenital anomaly except their prematurity. The first two groups raise problems that are essentially similar to those that arise in other patient groups. They are different from other patients only in that they are babies. The third group is entirely unique. There are no other clinical populations in which the primary clinical problem arises from an arrested developmental process, and in which the clinical problem will correct itself if development is allowed to continue. This entry describes, in separate sections, the three groups and the different ethical issues associated with them.

Acute Illnesses

Acute illnesses in full-term babies are usually the least morally controversial of the clinical problems that arise in neonatology. Most acute illnesses can be treated if they are accurately diagnosed. The problems that arise in babies are similar to the problems in other high-risk populations. Diagnosis must be made quickly and treatment initiated expeditiously. The babies generally either get better quickly or they die quickly.

Rarely, treatment is only partially successful, and the babies survive but with severe long-term complications of their acute illness. For example, term babies might develop meningitis. The disease can be diagnosed and treated, but some babies are left with severe neurological impairment. In situations where treatment is only partially successful, the babies become similar to babies with uncorrectable congenital anomalies. The only difference is that, in these cases, the anomaly was acquired after birth rather than before. The process of decision-making, however, will be similar to that outlined below.

In general, the issues that arise in this group of NICU patients are not unique to NICU patients. The primary concerns are accurate diagnosis and appropriate treatment. If there is a treatment that works, it should be provided. There is rarely disagreement between doctors and parents in these cases. The ethical issues are driven almost entirely by the medical indications for treatment. Ethics simply dictates that doctors should be competent and should communicate well with parents.

Congenital Anomalies

Congenital anomalies were the primary focus of legal and moral controversy in the 1970s regarding treatment decisions for newborns (Lantos). The key cases focused on syndromes such as trisomy 21 (Down syndrome), spina bifida, and cases of multiple congenital malformations. Generally, the issue that arose was whether or not to attempt surgical treatment to correct some but not all of the congenital anomalies. Thus, for babies with Down syndrome, the issue was whether to correct an associated intestinal or cardiac malformation. In babies with spina bifida, the issue was whether to treat the hydrocephalus by placing a shunt in the brain, or whether to close an open lesion of the spinal canal.

It is understood that, in these situations, the underlying syndromes themselves cannot be treated. Babies with Down syndrome will still have Down syndrome, even if their intestinal or cardiac anomalies are repaired. Babies with spina bifida will still have the long-term neurological problems associated with the spinal cord injury, even if their hydrocephalus is treated. Thus, the issues that drive these decisions are fundamentally different from those in cases of acute illness. The primary focus in these cases is on long-term quality of life for survivors.

Congenital anomalies focus discussion on anticipated quality of life as opposed to prognosis for survival. The discussions often involve the ethical implications of active intervention as opposed to "letting nature take its course." In many of the discussions, there is a sense of fatalism in the face of what are seen as mistakes of nature. Perhaps, the thought goes, these babies were somehow meant to die, and interventions are both unnatural and inhumane.

Many congenital malformations that were once thought to be incompatible with life can now be treated. Given the capabilities of modern intensive care, there are very few congenital anomalies that are truly incompatible with life. Babies with severe congenital heart disease can have open-heart surgery, babies with no intestines can be given total parenteral nutrition, and babies with minimal brain function can be kept alive on ventilators. Unlike cases of acute illness, decisions in these cases are not driven primarily by the medical indications for treatment. Those are usually straightforward. Instead, the decisions are driven by judgments about whether the results of successful treatment will be acceptable. In other words, will the consequential quality of life be sufficient to make the life worth living? In analyzing such decisions, it is necessary to have a nuanced understanding of the different components of "quality of life."

QUALITY OF LIFE COMPONENTS. Quality of life can be broken down into a number of ethically relevant components, each of which must be considered in these cases. These components include the anticipated cognitive or neurological function, the anticipated physical disabilities, the pain and suffering that is associated with the disease itself, and the burdens of the treatments that will be necessary in the future.

Most people today hold that a certain minimal level of cognitive or neurological function is essential for a life to be considered worth living. This was one of the rare areas of consensus in the 1980s controversy about federal regulation of nontreatment decisions for newborns. The agreement, in principle, that cognitive function is an important consideration begs the question of appropriate thresholds. Babies with no cortical function at all, such as babies with anencephaly or babies with prolonged cortical unresponsiveness as a result of anoxic (oxygen-deficiency) injuries, define one extreme. Babies with syndromes such as Down syndrome that lead to mild mental retardation are at the other end of the spectrum. In between are babies with other chromosomal or genetic anomalies, babies with intraventricular hemorrhages, or babies with neurological damage as an aftereffect of treatment for an acute illness.

The process for decision making in such cases requires recognition of a real but constantly shifting boundary or threshold that has clearly defined extremes and a well-recognized "gray zone" in the middle. Today, as a matter of societal consensus, the quality of life in Down syndrome is considered to be above the threshold, so these babies, and babies like them, must be treated. The quality of life in anencephaly is considered to fall below the threshold, so babies with this syndrome generally ought not be treated. There are, occasionally, exceptional cases of anencephaly in which treatment is provided at the parents' insistence, but they are noteworthy because they are exceedingly rare. Exceptions to the rule do not undermine the validity of the rule, they simply highlight the difficulty of imposing universal compliance with the rule. Cases in between these extremes are still difficult and controversial.

The physical disabilities associated with a condition must be addressed separately from the cognitive or neurological disabilities. Often, babies have an intact brain but have other physical disabilities. In severe spina bifida, for example, the spinal cord damage may make it impossible for a person to move about independently. Generally, physical disabilities, by themselves, cannot justify a decision to withhold life-sustaining treatment. It is clear from studies of adults with spinal cord injuries that it is possible for a person with severe physical disabilities to lead a rich and satisfying life. Thus, in such cases, the focus of discussion is usually on developing an adequate support system and insuring access to rehabilitation services so that function can be maximized.

A third part of any assessment of quality of life has to do with the pain and suffering associated with the disease. Some diseases lead to unrelenting pain and suffering. For example, severe epidermolysis bullosa is a disease that causes blistering of the skin over the entire body, including the oral cavity and intestinal tract. Swallowing is impossible. Scarring of the skin leads to contractures (permanent shortenings) of all the joints. Even comfort care is difficult because merely handling babies with this syndrome causes pain and exacerbates the condition. In such a case, an attempt to prolong life inevitably prolongs the suffering. It is appropriate in such cases, or in cases like them (though there are not many other syndromes that are relevantly similar to epidermolysis bullosa), to withhold life-sustaining treatment based solely on the pain and suffering associated with the disease.

Another component of quality of life has to do not with the pain and suffering of the underlying condition but with the pain and other burdens associated with the necessary treatments. Babies with short gut syndrome, for example, can survive, but only with indwelling venous catheters placed into large veins in the chest or neck. These central lines often become infected and must be replaced. When they become infected, patients must be admitted to the hospital for intravenous antibiotics. Parenteral nutrition often causes secondary problems such as liver failure. In extreme cases, patients are frequently hospitalized to deal with the complications of the treatment, and further treatment predictably exacerbates these complications in ways that cannot be prevented. Another example of excessively burdensome treatment is the provision of mechanical ventilation for babies with progressive and degenerative motor neuron disease. Some such babies are unable to eat, breathe, or talk, but their cerebral cortex is intact, so that they can think. Prolonged mechanical ventilation can prolong life for such babies, but the burdens of the treatment are thought to be high enough that a decision not to initiate mechanical ventilation, or to discontinue it once started, is usually considered acceptable. In such cases, the burdens of treatment drive the decision.

Any adequate discussion of quality of life must separate the components. Nevertheless, in most cases, a combination of these components exists. Generally, the task of moral reasoning about any unique individual case requires doctors and parents to analogize the case with better known paradigm cases. For doctors and parents, the question may be whether a particular case is more like Down syndrome than it is like anencephaly, or whether a burdensome treatment is more like lifetime mechanical ventilation than it is like lifetime dependence on insulin.

Extreme Prematurity

Babies with extreme prematurity comprise the third group of babies admitted to the NICU. The moral considerations involved with these babies include not only all of the considerations in the other two groups but also an important new one—long-term prognostic uncertainty.

Prematurity is both an acute crisis and a chronic condition. The acute crisis requires an emergency response driven by medical indications, just as in the cases of full-term babies with acute medical problems. At the time treatment is initiated, however, the baby's prognosis is usually uncertain in a different way than in the other two situations. With acute pneumonia, treatment usually either succeeds, in which case the problem is completed resolved, or it fails, in which case the baby dies. There is almost no middle ground. With congenital anomalies and syndromes, treatment cannot cure the underlying disease. So a baby with Down syndrome will still have Down syndrome, even if the congenital heart disease is repaired. The long-term prognosis for survivors is clearly predictable. Again, there is almost no middle ground of uncertainty. With extremely premature babies, by contrast, the prognosis is radically uncertain. It ranges from early death through later death to survival with severe disabilities, moderate disabilities, or no disabilities. The disabilities can be cognitive, pulmonary, or involve virtually any other organ system. When treatment must be initiated, nearly all babies are in a prognostic gray zone. The outcome for any particular baby simply cannot be known, and it can range across the entire spectrum of possibilities from the very best to the very worst. This raises a whole different set of ethical considerations.

PROGNOSIS FOR SURVIVAL AND THE BIRTHWEIGHT FACTOR. At the time of birth, it is difficult to say whether a particular baby will live or die. It is difficult to predict how long life can be prolonged in cases where death will ultimately ensue. And it is difficult to predict whether survivors will have mild, moderate, or severe chronic problems or no problems. Obviously, an accurate prognosis for a particular baby would be essential to making the best ethical decision for that baby. If survival is impossible, then treatment should not be provided. If intact survival is likely, then treatment is morally obligatory. To a certain extent, clinical research in neonatal intensive care since the early 1980s has helped bring about a greater understanding of these issues and helped to refine, though not perfect, doctors' prognostic abilities.

The goal of this research has been to develop a method to precisely predict the anticipated outcome for each premature baby. The most powerful prognostic measure has always been birthweight. Overall, bigger babies do much better than smaller babies. Almost no babies who weigh less than 500 grams at birth survive, whereas nearly all babies who weigh more than 875 grams at birth survive. The zone of controversy is in between these two birthweights. The weights correspond, roughly, to the time between about twenty-three weeks of gestation and twenty-six weeks of gestation, or between the fifth and sixth months of pregnancy.

One plausible response to these data would be to suggest that only babies over 850 grams should be treated. If this course were taken, however, there would be many babies in the 500- to 850-gram birthweight range who might have survived but who would be allowed to die. Another response might be to treat all babies over 500 grams. With this option, however, treatment is provided to many babies whose death is likely. One way to refine the prognostic estimates is to look a little more closely at the clinical course of these babies.

It turns out that most premature babies who are going to die do so in the first few days of life. The sickest babies are very sick. Because the sickest babies die quickly, the babies who survive for even three days are, by definition, much more likely to survive than other babies of the same birthweight. In fact, by seventy-two hours of age, birthweight virtually disappears as a relevant predictor of survival. The 600-gram babies who survive do just as well as the 1,000-gram babies who survive (Meadow, Reimshisel, and Lantos).

These clinical epidemiological facts have shaped the moral responses of NICU professionals. Prior to the 1980s, discussions about the appropriateness of decisions involving whether or not to treat premature babies presumed that the decision should be made at birth and in the delivery room. It was seen as a one-time, either/or decision. The newly understood clinical realities show why that did not make sense. Among all babies born at less than 750 grams today, half can be saved and half cannot. At the time of birth, however, it is almost impossible to tell which baby will be in which group. The only way to separate the two groups is to initiate treatment on all of them. The sickest babies then declare themselves by getting sicker in spite of medical treatment, and the healthier babies declare themselves by improving.

NEUROLOGICAL OUTCOMES. Determining the prognosis for neurological outcome is even more difficult than determining the prognosis for survival. Clearly, premature babies have worse neurological outcomes than full-term babies. Numerous studies have shown a higher incidence of cerebral palsy, seizures, chronic lung disease, and educational problems among premature babies than among their full-term peers. Nicholas S. Wood and colleagues summarized outcomes for tiny babies in a 2000 issue of the New England Journal of Medicine (see Figure 1).

These statistics, however, are like the statistics showing poor survival for tiny babies. The relevant question for clinicians and parents is not whether, overall, a group of babies has a worse prognosis but whether for any particular baby the likely outcome can be predicted. There are predictors of bad outcomes, but they are imperfect.

As they reported in 2002, Carl T. D'Angio and colleagues studied long-term neurologic, cognitive, and educational outcomes for babies born at less then twenty-nine weeks of gestation. They showed that the only predictors of bad neurologic outcomes were neonatal intraventricular hemorrhage, severe lung disease, and low socioeconomic status. Betty R. Vohr and colleagues found similar results in a 2000 study. Interestingly, the first two factors are physiological while the third is social, but each is independently associated with bad outcomes. At the very least, this suggests a complex interplay between physiological and sociological factors. Importantly, neither birthweight nor gestational age is, by itself, associated with poor neurological outcomes for these babies. This again suggests that, in the clinical setting, a simple criterion for treatment or nontreatment based on birthweight alone is likely to be relatively inaccurate in tailoring treatment decisions in an ethically appropriate way.

FIGURE 1

Implications of Clinical Knowledge for Ethical Decision Making

These epidemiological facts help define the zone of parental discretion. In order for a decision to withhold life-sustaining treatment to even be considered, doctors must first determine that a baby has an appropriately severe condition. Thus, doctors initiate most discussions of treatment withdrawal. Sometimes, parents will initiate the discussions but when they do, the doctor's task is the same—deciding whether or not the baby fits into one of the categories in which treatment withdrawal is permissible. If not, the doctor must rebuff the parents' request. If so, the doctor should facilitate the process in a way similar to the way she would if she had initiated it herself.

Over the years, different schools of thought have evolved about the proper tone and structure for such discussions. These might be characterized as the objective information approach and the broad shoulders approach.

In the objective approach, doctors see it as their responsibility to give parents information in the most nondirective way. They simply provide the facts and try to empower parents to understand those facts and to come to a decision that reflects the parents' personal moral or spiritual values. In this approach, the doctor does not make a recommendation about the appropriate course of treatment. If they are asked what they would do, they refuse to answer. The moral psychology of this approach is based upon a fear of being coercive. It views doctors as inappropriately empowered and parents as problematically vulnerable to being overpowered. Given that sociological background, doctors have a moral obligation to restrain their own implicit dominating impulses. Sociologists, who examine the power structures of human communities, often see this sort of pattern of interaction. Some philosophers, especially those for whom individual autonomy is a paramount moral principle, are the most articulate defenders of this approach.

The broad shoulders approach takes a different tack. By this view, parents' particular vulnerabilities require doctors to take some of the burden of decision making upon themselves. Instead of simply giving parents the facts, doctors are obligated to make a recommendation. Advocates of this approach point out that the circumstances of serious illness are circumstances of personal moral and psychological crisis in which ordinary moral principles may not be applicable. Individuals may not be capable of the same sort of autonomy in such situations as they are in other situations. They may need subtle and often implicit assistance to understand their own wants, needs, and values, and they may have trouble owning the decision that flows from these values.

In spite of these radical differences in understandings of the moral underpinnings of the conversations that lead to decisions, in practice, the structure of conversations between doctors and parents look similar in both. The first discussion is one of facts and possibilities. The clinical facts are explained. The possibilities for treatment or nontreatment or presented. Questions are answered. Usually, this first discussion is then adjourned, and parents are allowed time to think. In most cases, they seek outside support—from extended family, from clergy, or from mental health professionals.

A second discussion, during which a decision is reached, usually follows within a few days. Three sorts of conclusions can be reached. In the first, parents decide that they do not want to stop treatment and do not want to reconsider their decision in the future. They want everything done to keep their baby alive. Generally, this leads to a discussion of the ambiguity of the term everything done in today's medical environment. The second sort of conclusion that can be reached is for a time-limited trial of continued treatment. By this approach, doctors agree to continue treatment for a defined period and to set certain parameters or endpoints that they might then look for to see if the treatment is leading to anticipated goals. For example, doctors might offer to continue mechanical ventilation for another week and if, at that point, the ventilator can be safely discontinued, then it will be. If not, however, it will be discontinued anyway in a manner that will likely lead to the death of the baby. The final sort of conclusion that can be reached is a decision to withdraw life-sustaining treatment immediately. In these situations, a standard set of rituals, familiar to the staff in most NICUs, ensue. The baby is moved to a separate room, the parents are called in, the ventilator or the intravenous fluid pumps are removed, and the parents are allowed time alone to hold their dying baby.

These approaches reflect the inherent uncertainty of the process. When death is inevitable, there is no moral decision to be made. In those circumstances, heroic efforts are often made to prolong life and those efforts fail. A moral statement has been made, a moral commitment fulfilled. Moral decisions arise only when there is ambiguity or uncertainty about the prognosis and about the efficacy of treatment. As has been shown above, however, there is almost always uncertainty. Uncertainty creates the necessity for moral, as opposed to simply clinical, decision making.

Conclusion

The current state of ethical decision making for infants in neonatal intensive care units involves several tasks. The first task is to correctly categorize the clinical indication for intensive care treatment. The second is to determine, as accurately as possible, the baby's prognosis for survival. Finally, doctors must estimate the prognosis for long-term outcome among survivors in terms of neurologic disability and quality of life. These facts ground discussions of the proper ethical course of action. In most cases, they lead quickly to a consensus about the proper course of action.

john d. lantos

kathryn l. moseley (1995)

revised by john d. lantos

william meadow

SEE ALSO: Abuse, Interpersonal: Child Abuse; AIDS; Care: Contemporary Ethics of; Children; Clinical Ethics; Compassionate Love; Embryo and Fetus; Family and Family Medicine; Feminism; Genetic Testing and Screening; Infants, Ethical Issues with; Life, Quality of: Quality of Life in Clinical Decisions; Maternal-Fetal Relationship; Medicaid; Moral Status; Pediatrics; Research Policy: Risk and Vulnerable Groups; Sexism; Trust

BIBLIOGRAPHY

D'Angio, Carl T.; Sinkin, Robert A.; Stevens, Timothy P., et al. 2002. "Longitudinal, Fifteen-Year Follow-Up of Children Born at Less Than Twenty-Nine Weeks' Gestation after Introduction of Surfactant Therapy into a Region: Neurologic, Cognitive, and Educational Outcomes." Pediatrics 110(6): 1094–1102.

Lantos, John D. 1987. "Baby Doe Five Years Later: Implications for Child Health." New England Journal of Medicine 317(7): 444–447.

Meadow, William,; Reimshisel, Tom; and Lantos, John. 1996. "Birth Weight–Specific Mortality for Extremely Low Birth Weight Infants Vanishes by Four Days of Life: Epidemiology and Ethics in the Neonatal Intensive Care Unit." Pediatrics 97(5): 636–643.

Vohr, Betty R.; Wright, Linda L.; Dusick, Anna M., et al. 2000. "Neurodevelopmental and Functional Outcomes of Extremely Low Birth Weight Infants in the National Institute of Child Health and Human Development Neonatal Research Network, 1993–1994." Pediatrics 105(6): 1216–1226.

Wood, Nicholas S.; Marlow, Neil; Costeloe, Kate; et al. 2000. "Neurologic and Developmental Disability after Extremely Preterm Birth." New England Journal of Medicine 343(6): 378–384.