What a horrible idea

This idea is so bad that I might even agree with a Scientologist about it (OK, not really). A company I will not name or link to has developed a home genetic test for bipolar disorder. What could be so horrible about making it easier for people to diagnose diseases?

Well, first there is a problem of "begging the question": does the test do what it is purported to do?

(Test X)⢠- tests for two mutations in the GRK3 gene that are associated with bipolar disorder. Patients who have either of these two mutations, are Caucasian, of Northern European ancestry and have a family history of bipolar disorder, are three times more likely to have bipolar disorder.

Three times more likely than whom?

And what does "3x more likely" mean? That is a statement of risk, but without knowing the baseline risk, it's hard to know what to think. Using prevalence data from the NMH , there are about 5 million people in the U.S. with bipolar disorder. That's about 1 in 60 people. "Three times more likely" might mean your chances go up to 3 in 60 (from 1.7% to 5%, correct ethnicity only, please). But it probably doesn't mean that at all. The only way to really know what it means is to get a hold of the test characteristics, which I can't seem to find. Accuracy of diagnostic tests depends on many different factors including how common the disease is in a population. It is unlikely that patients and doctors will know how to interpret a positive or negative result---I know I won't be able to.

Then there is the problem with defining "bipolar disorder". Bipolar Spectrum? Bipolar I? Bipolar II? What about the patient's experience? Most mental illnesses are defined by patient experience since the underlying etiologies are not known. Neither are mental illnesses homogeneous disorders. Each diagnosis might actually encompass many different "diseases".

Does this test add anything to our current diagnostic paradigm? I can't see how. Once again, for an illness of unknown etiology, based on description of behavior and thought, a genetic test is not useful.

And I won't even touch the ethical issues of medical genomics.

Empowering patients, whatever that means, is a good thing. But giving them "noise" in the guise of information isn't helpful. This is not much different than the CT scanners that were doing huge business a few years ago, promising a great look at your insides. It provided that, but without any context for interpretation. For a new diagnostic test to "go live", it must have data to back up it's utility. It's not enough that it accurately measure a particular parameter---that parameter must have meaning.

Genetic testing will probably gain in importance in all fields of medicine, including psychiatry. But selling unproven tests is bad medicine.

Tags

More like this

GrrlScientist wrote href="http://scienceblogs.com/grrlscientist/2008/02/diagnosing_bipolar_disorder_wi.php">a post last month about a potential genetic test for bipolar disorder.  Read that first to get some background. Now, it turns out that a company is selling a testing kit that you can…
tags: researchblogging.org, bipolar disorder, clinical depression, mental illness, mood disorder, functional genomics, blood test, biomarkers Image: Florida Department of Law Enforcement. According to the United States Department of Health and Human Services, serious mental illnesses affect…
The Archives of General Psychiatry has an open-access article about bipolar disorder in childhood ( href="http://archpsyc.ama-assn.org/cgi/content/full/65/10/1125">Child Bipolar I Disorder).  I started to write about that.  But then, as often happens, I stumbled upon something else. The LA…
Mark Henderson has a great piece in the Times exploring the impact of personal genomics on the practice of medicine. The basic theme should be familiar to anyone who has been following the emergence of the personal genomics industry: doctors are currently almost completely unprepared for the…

Couldn't agree more. To add my own slightly broader condemnation is the whole genre (for lack of a better word at the moment) of patient-directed testing. The immediate question I think is raised by PalMD is "what does one do with the results of the home bipolar test?" Patients can/do request CTs, various blood tests (home or other), hair sample analyses (my personal favorite)yet have no idea how to interpret them. It seems to me that we're evolving a marketing scheme praying on our insecurities and irrational fear of death&disease, rather than trying to improve the health of our population. I'm sure I'm not the only clinician who has been brought results of a test I would never order (because I can't interpret it or wouldn't know what action to take based on the results), which has been done elsewhere, and who has to deal with the aftermath of a panicked poorly-informed patient whom you can't possibly adequately educate/console/pacify/treat in the 15 minutes alloted.

As more and more of these home-tests come on-line, who is responsible for interpreting and acting on the results? If I order a test in the clinic and get an abnormal result, I can be found liable if I don't take appropriate action (miss a malignancy report on a colon-polyp path report and the patient dies of metastatic CA...) Does the company marketing the test or the store selling the test bear any responsiblity (legally probably not, although I'd argue they should.)

Now wait... If you had bipolar disorder, wouldn't you have a pretty good idea that there was something wrong without a genetic test? Even if it is accurate, what good does it do? If you suspect that you have bipolar disorder, you go to a doctor. Who needs a DNA test when you just knocked out 3 walls for the kitchen remodeling project of the century last week, just to end up in bed, too depressed to get up this week.

If you suspect that you have bipolar disorder, you go to a doctor. But if, instead of merely suspecting it, you actually have bipolar disorder, voluntarily seeking the advice of a doctor is the last thing you do. And if your best efforts to avoid such a diagnosis happen to fail, you come up with five hundred reasons why the diagnosing doctor is a quack. Let it be anything but that. In one respect, a home genetic test for bipolar disorder might acutally have some value as a diagnostic tool: if the subject agrees to the test, it's probably not bipolar disorder.

Do they tell us the specificity of the test? IOW how likely is it to come up positive if you're not bipolar?

There are some classic statistical problems lurking in here - it's possible that more people could be told they're susceptible and never become bipolar than who are correctly predicted to become bipolar.

If you had bipolar disorder, wouldn't you have a pretty good idea that there was something wrong without a genetic test?

I take your point to be that a much more sensible way to diagnose would be to simply look directly for the usual diagnostic criteria and that it seems unlikely that a genetic test for predisposing genes would be more informative. If so, I agree entirely.

I understand -- or more accurately, I read somewhere, once from a source which may not have been completely full of crap but probably was not necessarily infallible -- that in the case of bipolar disorder, the manic phase feels so good and people in the phase are so genuinely productive and creative, that is not unusual to be difficult to get them to see that there is in fact a problem. (Because, if I understand correctly, until the depressive phase hits, it's not really so much a problem. If anything it sounds like kind of a bonus.)

My pedantic point being, in response to the query, "Wouldn't someone with bipolar disorder know there's something wrong?", it is possible that they might not always know there's something wrong.

However, I suspect that the people who are unaware that anything is wrong are also not the ones using home genetic tests, so this would not in any way contradict what I took to be your general point, that if one is wondering if they have bipolar disorder, the genetic test is going to be nearly useless and a much better guide would be, well, things like "do you experience sudden changes between manic phases and depressed phases?" and so on. It seems to me unlikely that a person with bipolar disorder who believes they're fine, would take the test, see the positive result (i.e. that they possess the genetic risk factors and meet the other criteria) and have a moment of realization that they might be bipolar. It seems to me likely that they would think "I have the risk factors but I must be one of the people who have the risk factors but doesn't have it, because clearly I'm fine."

In summary, yeah, it sounds like a useless test in the first place, and also in danger of being misinterpreted and misunderstood and just making things worse.

By El Christador (not verified) on 25 Mar 2008 #permalink

This is not a horrible idea. What's wrong with knowing what's
going on with my body? Why should I not learn whether I have
0, or 1, or 2, of the genetic defects shown by the UCSD group
to be associated with the incidence of bipolar illness? My
psychiatrist is great in her use of counseling and the mimimum
effective drug regimen to keep my condition under control - but
why does she not care about biological characterization of her
patient population - i.e., correlating which of her patients
have one or both of these genetic defects with how they're doing
on their respective drug regimens. Wouldn't it help her to
understand better the biological basis of the illness and what
medications are more likely to be effective?

Maybe Kelsoe is just discouraged that the therapeutic psychiatric
community hasn't shown much interest in his group's study.
Why couldn't one of the three hospitals I've been in have
suggested participation in Kelsoe's study? Do they only care
about bringing their patients quickly under control, using some
mind-numbing drug such as Depakote? (N.B. Actually, the I above
is somebody else I care about.)

Also, do you read only the press releases of the pharmaceutical
cos.? PR folks never get the details correct - you know that.
The UCSD web site gives a thorough characterization of their study population.

Did you bother reading that the test results for the bipolar test
are sent to the individual's doctor - not directly to the subject.

Surely, a pre-knowledge of the biological basis of a condition
must somehow be related to the appropriate choice of drug therapy -
and practicing physicians and therapists ought to order
any available genetic tests to help see what works for whom.

By Fred Schwab (not verified) on 25 Mar 2008 #permalink

Fred Schwab wrote: "This is not a horrible idea. What's wrong with knowing what's
going on with my body? Why should I not learn whether I have
0, or 1, or 2, of the genetic defects shown by the UCSD group
to be associated with the incidence of bipolar illness? My
psychiatrist is great in her use of counseling and the mimimum
effective drug regimen to keep my condition under control - but
why does she not care about biological characterization of her
patient population - i.e., correlating which of her patients
have one or both of these genetic defects with how they're doing
on their respective drug regimens. Wouldn't it help her to
understand better the biological basis of the illness and what
medications are more likely to be effective?"

For me the question becomes "what are you going to do with the information?" Please help me understand; with today's body of knowledge, are treatment regimens for bipolar based on specific genetic variants? (I'm not aware that psychiatric treatment was that scientifically based yet, given that we're not even sure how our drug regiments ultimately cause the behavioral changes they do.) I'm not arguing that genetic testing doesn't have its place, but unless the results help you you make a treatment recommendation to this patient, right now, I would argue that you shouldn't order the test (outside of a research protocol.)

"Did you bother reading that the test results for the bipolar test
are sent to the individual's doctor - not directly to the subject."

No, I didn't know this, but it really doesn't matter to me (see my original post, the first response to PalMD).

"Surely, a pre-knowledge of the biological basis of a condition
must somehow be related to the appropriate choice of drug therapy -
and practicing physicians and therapists ought to order
any available genetic tests to help see what works for whom."

You assume that our current state of knowledge of the disease and of treatment is refined to the point that this information can impact decisions *today.* Unless someone has done the study that genetic variant X is best treated by drug Y, whereas variant A is best treated by drug B, you're just doing N-of-1 studies on your patients if you're making treatment decisions based on this test.

This test doesn't sound great but a test for bi-polar would be awesome.

It's been said that depressives view the world as it is. This can be in their psychiatrists office knowing that the man has no definitive way of knowing if they really are manic depressive. There is currently no physical test for bi-polar disorder, there is a lot of guess work a culture of perhaps over diagnosing the problem and also culture of demonizing the problem.

While some people might view you as more creative because of this diagnosis, law enforcement and a few others can see you as much more dangerous because of the same diagnosis. Your insurance might start seeing you as much less desirable and any employers can start seeing you as 1000% less reliable.

I'm soo manic depressive. Yes, I know there is a problem. I just want to be sure the problem is nailed down right. I have the wisdom to not trust my own judgment as well as the audacity to not trust the judgment of the diagnostician. I would love a clear incontrovertible test. To prove to some people that I'm not faking it. To prove to myself sometimes that I do need to stay on the meds. If there was a simple test to tell you what phase you were in and what meds to take that would be even better. Some meds treat the depression but are the last thing you want to give someone who is manic.

The manic part is often great. You can get enormous things done. You can inspire people around you and truly warp their reality and impression of you. When you crash you realize how fallible your reality was and how distorted the shared reality was, this knowledge informs informs your distrust of the diagnostician.

My big problem with the manic phase is that I spend spend spend money and make investments (perhaps unwisely). I also start lots of things and don't get to finish them all. In the down time I hate myself for not completing any project no matter how overly ambitious.