Between a Rock and a Hard Place

Medical ethicists have traditionally referred to beneficence, nonmaleficence, respect for autonomy, and justice as fundamental principles of medical ethics.¹ In some cases, however, adherence to these principles produces competing ethical obligations. Such is the dilemma when a patient with early Huntington disease (HD) forbids her physician to disclose this diagnosis to her daughter. The daughter is a patient of the same physician; she has a 50% chance of inheriting the HD mutation, which inevitably causes the disease.

The principle of respect for autonomy suggests that the physician defer to the mother's wish and not disclose her diagnosis. On the other hand, the principle of beneficence is most relevant to the physician's relationship with the daughter.² The beneficent act would be to inform the daughter of her risk for HD so that she can decide whether to receive genetic testing and potentially prevent transmitting the mutation to her children.³ Although some authors have suggested that patients have a "right not to know" about their risk for serious genetic diseases that lack effective treatments,⁴ informing the daughter that she is at risk would preserve her right not to know whether she has the HD mutation while offering her the opportunity for testing.

Unfortunately, no inherent feature of the 4 principles provides a method for untangling such conflicting obligations.⁵ They describe the physician's obligations to each patient individually but do not shed much light on how to balance incompatible duties. Thus, the physician must look elsewhere for guidance when basic principles alone are inadequate to resolve an ethical conflict. What method should the physician use in resolving this dilemma? Some ethicists would suggest that the physician ought to reason inductively on the basis of similarity to previous cases (a "bottom-up" approach), while others would favor a "top-down" approach, in which the physician would select general principles that describe a way to sort out competing obligations, and then apply these to the specific case. In practice, however, the choice of problem-solving strategy may be irrelevant, as different approaches may nonetheless yield the same solutions.⁶ Of course, there are cases in which rational people disagree about both action and rationale; sorting out these differences is indeed a major project of ethicists. The relevant point, however, is that the physician might avoid the gridlock of conflicting principles by applying consensus-based ethical guidelines. This approach would circumvent the dilemma of selecting strictly a top-down or bottom-up approach, and instead incorporate the strengths of both.

The physician could refer to the Institute of Medicine (IOM) Committee on Assessing Genetic Risks guidelines.⁷ These guidelines are congruent with those of other governmental commissions on bioethics, including the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavior Research⁸ and the British Nuffield Council on Bioethics.⁹ This congruence is significant because it broadens the population across which consensus was reached; assessing both US and British recommendations may help in reducing the influence of sociocultural bias and popular opinion on consensus. The IOM guidelines, furthermore, are consistent with the "duty to warn" set forth by the Tarasoff case, which established that clinicians have a duty to warn a third party that a patient had threatened harm to that person, even if this means breaching confidentiality.²

According to IOM guidelines,⁷ several criteria must be met before confidentiality can be broken. First, all attempts to elicit voluntary disclosure must have failed. In this case, the physician should not only encourage the mother but also ensure that she understands the implications of her decision. The mother, however, is adamant in her refusal to disclose.

The next 2 criteria, which are most easily considered together, are whether there is a high probability of irreversible harm to the daughter and whether the harm would be prevented by disclosing the mother's diagnosis to her. Certainly, developing HD would be harmful to the daughter. However, as there is no treatment for asymptomatic persons with the HD mutation,¹⁰ informing the daughter of her risk would not improve her outcome medically. Beyond physical harm from HD itself, the daughter could experience psychological harm were she to pass the HD mutation to her child or if she were eventually diagnosed with HD and wished that she had been tested for the mutation earlier.

Finally, what would be the consequences if the daughter were unknowingly to pass the HD mutation to her child? Without knowing the daughter's genotype, her child would have a 25% risk of having the HD mutation; a knowledge of her genotype would establish the risk as either 50% or 0%, depending on whether she carried the mutation. This risk of inheriting the HD mutation represents a significant possibility for irreversible, serious harm to the child. However, the issue of whether a child is harmed by being born with a genetic disease vs never being born is quite murky.¹¹ Currently the daughter is only considering starting a family, so there is no immediate risk of harm to a potential child. Further discussions with the daughter could determine how soon she might try to conceive or if she is particularly concerned about any risks she might have of genetic diseases.

Based on IOM guidelines, the physician is not justified in disclosing the mother's diagnosis. Yet how can one be confident that merely applying abstract guidelines will yield a tenable solution to this dilemma? The physician might also use a casuistic approach, using case-based analysis or analogy to compare the reasoning applied to this case with similar ones in the past.

In the 1995 Pate v Threkel case,¹² the highest state court in Florida unanimously held that a physician has a duty to warn a third party about a genetically inherited disease. The plaintiff, who had medullary thyroid carcinoma, sued the physician who had previously treated her mother for the same disease, alleging that the physician was negligent in not disclosing to the plaintiff her risk of the disease. The court stated that the physician had a duty to warn the plaintiff about the risks of genetic transmission, but in light of state confidentiality laws, this duty was satisfied by warning the mother—not the plaintiff directly—of the possibility of genetic transmission. Despite the fact that the plaintiff was directly harmed by the lack of knowledge about her mother's condition (medullary thyroid cancer carries a better prognosis with earlier detection), the court's decision emphasizes respect for confidentiality over acting beneficently toward relatives. The reasoning presented here—that unless there is a high probability of serious harm to a third party, confidentiality ought to be respected—is therefore also supported by case law.

What, then, has become of the physician's obligation to act beneficently toward the daughter? It leaves behind what Nozick¹³ termed a moral trace—that is, a residual attitude or obligation that remains after a prima facie principle has been overridden. The physician's residual obligation is to strongly encourage the mother to inform the daughter herself by exploring the mother's reasons for not wanting to "burden" her daughter and perhaps by using motivational interviewing techniques to overcome her resistance. The physician should emphasize to the mother that fully involving her family in her care, particularly as the disease progresses, could result in a better quality of life.¹⁴ Additionally, although genetic counseling is typically nondirective, it may be appropriate in this case for a genetic counselor to advise the mother to disclose her diagnosis. In the end, however, the decision to disclose the diagnosis ought to be the mother's, and solely hers.

How much confidence ought the physician to have in this decision? Despite the reasoning process described herein, it is still possible that cultural or social factors influence the analysis. For instance, popular attitudes toward paternalism and patient autonomy have changed dramatically over the past century. However, it is more important to ask whether cultural or social bias necessarily has a deleterious impact on ethical decision making. Indeed, one might argue that an ethical position must be commensurate with widely held cultural values. Ultimately, the physician must simply trust that appropriate decision making techniques will yield ethically tenable solutions to difficult problems.