Discussion

The battlefield was set in the period 1990–1994, with eight separate publications regarding predictive genetic testing in minors. At the beginning of the ethical debate, authors promoted a conservative approach, arguing against testing for minors. Eight distinct arguments against predictive testing came to light in the very first year. Indeed, the case against testing was largely established by 1994, and these initial arguments against testing appeared repeatedly in the following years. By contrast, momentum in favour of testing evolved more slowly, with most arguments beginning to surface in the mid-to-late 1990s. Beyond this broad commentary on the ‘shape’ of the discourse thus far, there is no obvious pattern.

Certain arguments are more frequently cited in the literature than others (1A.1, 1B.1, 1C.1–1C.3, 1D.1, 1D.2, 2A.1 and 2C.1). There are a number of considerations that come to bear in examining the relative frequency of each of the arguments. The prominence of an argument, in terms of the number of times it appears, provides an element of insight into the level of importance attributed to it by commentators. Repetition in these terms, though, does not necessarily connote a strong or even sound argument. Clarke47 points to this trap in noting that it is insufficient to, ‘look at the length of the two, rather arbitrary, lists of arguments and then come to the unwarranted conclusion that the two opposed sets are of equal force’. An argument is not intrinsically significant simply because it has appeared in the literature. It must be tested against opposing positions. That said, there is some value in a tally, to understand what authors to date have considered important.

In the last 5 years all but one (1A.6) argument has continued to appear, highlighting that the debate is ongoing, with little resolution. The debate, such as it is, has adopted a to-and-fro pattern defined by parallel lines of argument, rarely intersecting and relying heavily on fundamentally opposed positions on core principles of bioethics with no reconciliation in sight. That no arguments have ‘dropped off’ the list testifies to a prevailing inertia. No argument has gained supremacy over others. The same core arguments are cited repeatedly, with only subtle variations, suggesting that a different approach may now be required. Clinical practice has forged on ahead of the discourse and predictive genetic testing in minors has become a reality.57 Formal guidelines have evolved to admit some flexibility into considering requests for predictive testing in minors, although they do not provide a framework or criteria to guide this process.

In reviewing the arguments listed in this review, two groups of claims can be identified. A minority are value claims supported by a preordained broader ethical stance, where there is no possibility of empirical investigation. Three (of 33) arguments clearly fit into this category (‘testing minors fails to respect their future autonomy', ‘testing minors breaches their confidentiality when their results are disclosed to their parents' and ‘it is important to avoid professional paternalism’). Of greater interest is the second group of arguments, which are all testable empirical claims that have not yet been tested. Examples of such testable, but so far untested, claims include: ‘A positive test result may lead to depression’ (1A.5); ‘Testing can have a negative effect on family relationships’ (1B.1); ‘Testing can decrease uncertainty’ (2A.1) and ‘Harm could result from not testing’ (2D.1). This group of untested claims forms the ethical subtext of the debate over predictive genetic testing in minors. This subtext concerns the ethics of collecting the evidence to settle the question. If most arguments are empirical claims, eminently testable, the matter at hand is whether to test these claims or if the potential outcomes of this research are too harmful to even attempt. Some argue that the harms and benefits of testing should be assessed in empirical research,5 while others argue that the potential for harm is too great to ethically conduct this research.46

Many of the arguments for and against testing are mirror images of one another. Some authors highlight particular outcomes as net negatives, whereas others look to the benefits associated with these same outcomes. For example, the new knowledge represented by a test result is viewed in negative terms as burdensome or else, more positively, as empowering knowledge enabling future decision-making, thereby promoting autonomy.20 There are underlying, mostly unacknowledged, values working to direct the commentators to one or the other interpretation. Working from the same set of ethical principles and concerns (eg, harm, benefit and autonomy), the commentators on either side of the debate have reached opposing conclusions and this sticking point is the position we remain in.

Whereas arguments in favour of predictive genetic testing of minors highlight consequences of receiving both gene-positive and gene-negative results, the arguments against predictive testing have focused heavily on the scenario in which a minor receives a gene-positive test result, largely excluding any serious analysis of potential outcomes. Of the arguments listed above, 11 of 16 assume an unfavourable test result (1A.1, 1A.3–1A.6, 1B.2 and 1C.1–1C.5). Authors appear to believe that the negative ramifications associated with this 50% probability (a gene-positive test result) are so profound that they outweigh any potential positive impact derived from a gene-negative result. Completely missed from this binary analysis is the possibility that there may be positive ramifications from a gene-positive result, or negative impacts from a gene-negative result and, importantly, the potential adverse effects of not testing.

In keeping with the overall lack of systematic work in the area, commentators have rarely discussed which of their arguments specifically apply to infants and younger children, older children or adolescents. A developmental perspective is indispensible, considering each argument as it applies along the trajectory of child and adolescent development. For this review, the authors attempted to categorise each argument by the age group most affected, specifically ‘older’ minors, ‘younger’ minors and arguments applicable to all minors. Here, ‘younger’ minors are ‘young people who do not possess the cognitive capacities that allow them to appreciate the implications of predictive testing’.5 ‘Older’ minors are considered to have developed the capacity to discuss and comprehend, to some meaningful extent, the testing process and potential long-term repercussions associated with testing. Some arguments clearly applied to one developmental stage; however, in most cases, making this distinction was challenging and made it more complicated when the original work of other authors were referred. For example, we felt that argument 1D.2 (‘testing minors breaches their confidentiality when their results are disclosed to their parents’) related more to younger minors where the request for a test comes from a third party (typically parents) but also acknowledge that some commentators argue that this applies equally for older minors. Ultimately, a developmental stratification was omitted from this review, because there was insufficient raw data, in that the literature to date has not adopted a developmental perspective, generalising, for the most part, across the entire child and adolescent developmental trajectory. Ethical issues vary significantly along the developmental pathway and it is recommended that future work make specific reference to the developmental context for their argument(s) in favour or opposed to predictive testing in minors.