The American Journal of Bioethics

Doukas and Berg (2001) present a model for avoiding ethical conflicts which can occur when one family member's genetic test results could alert another family member to eventual, if not imminent, harm. In this model, the family members enter into a covenant in which they agree to become the unit for healthcare with individuals treated in the context of family. The physician facilitates resolution of conflicts that arise with respect to disclosure of information about one family member that is pertinent to the healthcare of other family members, especially in regard to results of genetic tests. This model avoids a more paternalistic approach that would simply include a waiver of autonomy as part of the informed consent, a process that does not necessarily include negotiation among family members. [End Page 25] Unlike a one-time informed consent, this model lays the groundwork for future sharing of information pertinent to all family members.

One limitation of this model is the assumption that physicians often treat adult members of the same family other than husbands and wives or that physicians will have the opportunity to work with extended families as a unit. Although in the era of the general practitioner extended families frequently used the same physician, little evidence exists that this is the case in the era of the family practitioner. The U.S. population is relatively mobile, with adult children and siblings often living in different cities or regions (Gober 1993). Doukas and Berg cite evidence that a significant percentage of family members see the same primary care physician (Medalie et al. 1998; Murata and Kane 1989), but their evidence refers to husbands, wives, and their minor children, not to husbands, wives, and adult siblings, or adult siblings and their nuclear families. The model does not seem appropriate for a nuclear family with minor children.

Another issue of concern is the time investment required of primary care physicians to become competent to deal with genetic issues and to implement such a model (Fetters, Doukas, and Phan 1999; Doukas 1993). Physicians may appreciate the need to recognize when to refer their patients for genetic services or even to provide counseling for genetic testing. However, they may not yet find the investment of time in genetics education cost-effective given the low prevalence of gene mutations that put families at high risk for disease. Add the low prevalence of high-risk mutations to the paucity of opportunity to treat extended families, and the time investment becomes even less cost-effective for practicing physicians.

In spite of its limitations, the model would be more feasible if doctors had the information needed to address medical genetics issues with family members. As medical curricula include genetics education for primary care physicians (American Society of Human Genetics 1995), this model becomes more reasonable. Specialists in genetics and genetic counseling provide the ideal venue for counseling, testing, and risk communication with respect to high-risk mutations. In the future, molecular genetic information probably will be important in primary care, just as family history is important (Collins 1997). The distinction between the role of genetic predisposition conferred by high-risk mutations to rare forms of disease and genetic predisposition conferred by common gene variants to sporadic forms of disease will be elucidated in the coming decades. The latter may well become part of risk assessment in primary care but would not have the same implication for family members as do high-risk mutations.