A lot has been written about the risks & mostly theoretical & associated with genetic testing. Genetic information is highly personal, often difficult to understand, subject to considerable interpretation, and of widely variable predictive power. Two main kinds of worries arise. One pertains to how the individual will react to genetic information about him- or herself. Another pertains to how others & including employers, insurers, etc. & will react to it.
But there’s an interesting twist, here. Part of the risk of genetic data lies in how that data is interpreted, and how it’s interpreted depends in part upon what kinds of studies have been done on the relevant bit of genetics. The fact that you happen to have a certain string of DNA is meaningless until interpreted in light of relevant scientific studies. But here’s the problem. If you get tested today, and get your test results next week, new information released next month may alter the interpretation of those results. That may change the risk you face.
The result is that at least some of the risks posed by a genetic test are entirely outside of the control of the company that sells that test. That’s pretty odd, as product safety goes. Who determines just how safe a chainsaw is? The maker of that chainsaw. Who determines how many safety features are built into a new car? The maker of that car. Who determines the risk of a genetic test? That’s much harder to answer.
Biotech Ethics 
Your own examples suggest some answers. Once upon a time, how safe a chainsaw ought to be was left up to the chainsaw manufacturer (vetted, of course, by the judgment of the chainsaw buyer). Similarly, once upon a time the safety of an automobile design was more or less a matter for purely internal risk/benefit analysis by auto companies (although of course, mindful of civil liability, their analyses typically would take into account at least some of the probable social cost of their decisions).
But when it comes to chainsaws (at least in some jurisdictions) and cars (in many jurisdictions) governments have come to the conclusion that regulators can provide a more accurate accounting of probable social costs than will manufacturers left to their own devices. Previously uncontrollable (or, in any case, uncontrolled) risks can be made controllable through regulation.
So, in the case of automobiles, e.g., government regulations may require that manufacturers make technical service bulletins publicly available or initiate a non-voluntary recall (although it only rarely comes to that) or even override existing intellectual property rights in order to make proprietary information available to aftermarket manufacturers (including so-called “right to repair” laws like California’s).
The automotive model, it seems to me, could quite easily (and probably with good moral justification) be carried over into genetic testing. It does not seem unreasonable, for instance, that companies which sell DTC genetic tests could be required to maintain customer records for substantial period of time (say, 2o years) and to inform customers in case the scientific information necessary to interpret test results changes after initial test results are conveyed to the consumer (analogous to a service bulletin or, if mandated, a recall). Of course, ab initio, this might leave the judgment about what constitutes a sufficiently large change in information to warrant contacting customers up to testing companies. But having a regulatory framework in place would allow, over time, for better justified and better implemented regulatory requirements. Just like what’s already the case for chainsaws and cars.
Will:
That all makes sense. Except that the risks I’m trying to point to are ones against which the testing company is entirely impotent, as far as I can see. The company can (and in some cases perhaps ought) react to new information. But they cannot *prevent* that information from coming to light. And in some cases, new information coming to light will itself pose new risks. (Hypothetical example: that mutation you have that’s related to the size of your ears? We just found out that it’s also causally related to Huntington’s disease, which changes how your family and your employer relate to you. And there’s nothing the company that ordered the test can do to eliminate that risk.)
Chris.
True enough. But the company can follow (or could be compelled to follow) the instructions of their customers if new information about an allele becomes available. An analogy might be the linkage analysis tests that were used to assess for Huntington’s disease in the days before the mHtt mutation had been identified: Some family members would want to know their individual results, others would not.
I do now get your point, though, that there is a narrower problem here that really is a problem: It’s difficult to imagine how a patient/customer could retroactively revoke informed consent to the disclosure of genetic testing results (or, in the patient/customer’s own case, ‘unlearn’ the results of tests that have already taken place). Assessing in advance the likelihood that they might want to is indeed a case of ‘unknown unknowns’.
Viewed that way, however, perhaps this just provides one more consideration in support of tight regulation of the genetic testing market and tight control over the dissemination of genetic testing information.
Will:
I doubt the risk here would outweigh the free speech argument. Anyway, it should certainly give companies pause.
But I think it’s interesting that — from a company’s point of view — certain risks are just simply beyond their control. I’m sure that’s not unique. In fact, here’s another example: the risks of driving your subcompact car change when some OTHER company decides to start making Hummers!
Chris.