Breaching Confidentiality in Genetic and Non-Genetic Cases: Two Problematic Distinctions

Abstract: Respect for patient confidentiality is widely regarded as a fundamental principle of medical practice. Yet, ethical questions about the limits of confidentiality arise when patients refuse to share genetic risk information with their relatives. In such cases, healthcare providers may struggle with the permissibility of breaching confidentiality to warn their patients’ at-risk relatives. A number of authors exploring this issue have converged around the idea that genetic cases differ from non-genetic cases (e.g., involving a threat of violence or the spread of an infectious disease) along two related dimensions: (1) In genetic cases, the risk of harm is already present in an at-risk third party, whereas in non-genetic cases, it is not; and (2) In genetic cases, the patient does not create a risk of harm to a third party, whereas in non-genetic cases, the patient does. I argue that these distinctions fail to differentiate between genetic and non-genetic cases and should not bear on the permissibility of breaching confidentiality. Instead, such decisions should be based on the following neutral considerations: (A) the at-risk third party is identifiable; (B) the information is actionable; (C) the risk of harm is serious; (D) the information can be disclosed without endangering the patient; (E) the potential harm is imminent; (F) the third party is unlikely to learn about their risk elsewhere; and (G) the risk poses a public health threat. Based on these criteria, I conclude that breaching confidentiality to warn patients’ at-risk relatives will usually be impermissible, though important exceptions exist.