Read Original Story

Deep dive into the ethics and evidence underlying genetic testing in depression treatment


5 Star

Genetic tests for psychiatric drugs spur hope, doubts

Our Review Summary

pillsThis story, originally reported by the New England Center for Investigative Reporting and published in partnership with the Boston Globe, describes the history and ethical implications of a loophole allowing genetic tests to bypass FDA approval before doctors prescribe them. Specifically, it outlines concerns regarding the use of genetic testing to help psychiatrists identify which drugs may work best for a patient with mental illness. The story evaluates the research that genetic testing companies use to demonstrate the effectiveness of their tests — and reports that most of these studies are funded or conducted by the companies themselves. The story makes clear that using these genetic tests to guide decisions on prescribing drugs poses risks for patients. The article overall is very well-researched and covers an emerging and important topic.


Why This Matters

According to the CDC, approximately 25 percent of U.S. adults have some form of mental illness. Finding the right drug and dose for treating a mental illness can be time-consuming — and mentally, physically and emotionally exhausting for both patients and their loved ones. As a result, the idea of using a genetic test to identify drug treatments that are most likely to be effective is incredibly appealing for all concerned. However, it’s important to recognize the limitations of the current research supporting such testing. The tests are costly and may not offer any benefit. The tests may slightly help doctors pick the right drug more quickly, or they may not, but the drugs that are picked after the genetic tests are done are subject to the same scrutiny as always.


Does the story adequately discuss the costs of the intervention?


While the story looks at genetic tests for prescribing psychiatric drugs broadly, it also uses one test in particular as a focal point. The story notes that this test, called GeneSight, cost one patient $3,800. The story also notes that GeneSight is covered by Medicare. Those are great points to make in a story like this one. However, since the story also notes that there are a lot of other, similar tests on the market, it would have been good to tell readers a little bit more about cost. Is $3,800 about average for testing costs? Is there a range that these tests fall into? Do other insurers pay for GeneSight? The story would have been stronger if it had touched on these issues.

Does the story adequately quantify the benefits of the treatment/test/product/procedure?


We would have liked to see more detail about the company-funded studies mentioned, especially sample size and the odds ratios or degree of benefit. It is easy and tempting to discredit studies because they are company-funded, but this does not necessarily mean they are wrong — simply that we should proceed with caution in interpreting them. And it’s true that in many cases, the company is the only group interested in investing the time/money to do the necessary research.

With that being said, the story does note that “Finding a test to more precisely target treatment could help troubled people find help more quickly, cut down on side effects, and save money.” And, when discussing GeneSight in particular, the story points to “company-funded studies showing that when [patients with depression] took the medications recommended by the test, they had fewer symptoms after eight weeks compared with patients who did not use the test.” We’ll award a Satisfactory and leave the suggestions for next time.

Does the story adequately explain/quantify the harms of the intervention?

Not Satisfactory

This area could have been stronger. The story begins with an anecdote about a patient who was prescribed a drug based in part on the results of a genetic test. That patient became suicidal within three weeks and checked himself into a mental hospital. That sort of story tells readers very little about risks associated with these genetic tests, precisely because it is anecdotal — and there was very little else in the story about potential harms. For example, the story quotes one researcher as saying that claims about such genetic tests are “not harmless and may be quite dangerous.” But the story doesn’t tell us what that means. Why aren’t these claims harmless? Why may they be dangerous? We suspect that the biggest harm here is in the high cost of the test. But what if the test is wrong?  It might also cause delays in finding the “right” drug for patients — especially if the genetic test takes time to run and if doctors weigh the test results over clinical judgment or other factors, leading to poorer decisions.

Does the story seem to grasp the quality of the evidence?


The focus of the story is on quality of evidence. It emphasizes the lack of independently-funded studies that have evaluated the efficacy of these genetic tests. Based on a review by the New England Center for Investigative Reporting, “virtually all the evidence that these psychiatric tests work is based on limited studies funded by the companies themselves or researchers they fund, including all five studies used to promote GeneSight on the company’s website.” It would have been good to put some numbers there — “virtually all the evidence” is a somewhat vague description — but we think it does address the Quality of Evidence question in a meaningful way. Worth noting: an annotated version of this story that appeared on the New England Center for Investigative Reporting’s site includes links to supporting information that is of interest. It would have been good if the version of the story that appeared in the Globe had let readers know that the annotated version (and supporting documents) were available.

Does the story commit disease-mongering?


No disease mongering here.

Does the story use independent sources and identify conflicts of interest?


A very strong point for the story. It quotes multiple industry and expert sources, with industry sources clearly identified.

Does the story compare the new approach with existing alternatives?


The story looks at the field of genetic testing to identify psychiatric drugs most likely to be effective to treat specific mental conditions in a given patient. The alternative to using such testing is for psychiatrists to prescribe drugs based on their diagnosis, which is often determined by symptoms. The story doesn’t state this explicitly, and that would have been useful. But the story does note that psychiatry does have “a lack of definitive tests for conditions and treatments alike,” and that doctors often take a “trial-and-error approach to medications.” We think the story did enough here to earn a Satisfactory rating.

Does the story establish the availability of the treatment/test/product/procedure?


The story makes clear that many of these tests are currently available.

Does the story establish the true novelty of the approach?


The story notes that “Genetic tests to identify the most effective psychiatric drugs are the hot new technology in the race to create personalized treatments based on people’s DNA” and that “[m]ore than 600,000 of these tests have likely been administered in the last three years.” While it may have been nice to mention that genetic testing is done in some other fields (eg. to determine breast cancer risk) to provide context, the story does enough to earn a Satisfactory rating.

Does the story appear to rely solely or largely on a news release?


This work doesn’t appear to be based on a news release at all — and even if it were, it clearly goes far beyond anything a news release might include.

Total Score: 9 of 10 Satisfactory

Comments (3)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Beth Daley

October 7, 2015 at 10:58 am

Thanks so much for the robust, smart analysis.

I thought I would respond to a few points to give more context and clarification. I think most of the reviewers’ points are spot on. Unfortunately, the story was cut for space reasons by over 800 words and that forced some important information out of the story that, in hindsight, I should have figured out how to leave in. Yes, test cost ranges from a few hundred dollars to over $4,000 and other insurers do cover Genesight, although it appears some of them have taken their lead from Medicare as is often the case.
Good point on sample sizes (almost all small). I had originally included detailed analysis by two academics that one of GeneSight’s double blind, randomized trials were neither, as well as other criticisms, but it was cut. Who does the research work is a challenging question because it is true that without the company, the research wouldn’t be done, but that still mean the public should trust it. This is not to say FDA regulation is the right answer, but some sort of independent body appears needed to review science, claims, marketing etc. of the now more than 15,000 genetic tests on the market.
To push back on the anecdote rating, I see the reviewers point, but I thought the example sufficed for showing harm. The cost factor is a huge one, certainly, including to taxpayers when a federal insurer is paying, but my goal was to show that its not a way to a sure cure (and also costs a lot of money). The problem is with anecdotes, of course, is that its easy to sway – but the fact is because of the lack of independent validation it is exceedingly difficult to distinguish a good test from a bad one.
Once again, I did have other types of genetic tests to speak to the broad nature of genetic tests, but that dropped to the cutting room floor.
Many thanks again,

best, Beth Daley


    Kevin Lomangino

    October 8, 2015 at 7:47 am

    Thanks so much Beth for taking the time to reply. As journalists, we are all too familiar with the feeling of having important details and context axed for space or readability. We never run out of suggestions for things that could have improved a story, but we realize full well that this is an ideal that is often not attainable. And I see your point about the anecdote showing potential for harm. It does — though perhaps not as explicitly as the reviewers would have liked. Certainly a close call and perhaps our comments should have reflected that more clearly.

    Best regards,

    Kevin Lomangino


Johanna Ryan

October 7, 2015 at 7:55 pm

So glad to have this article! As a “patient activist” I’ve been leery of this trend for awhile but lacking in facts. One of the dangers, I think, is that this is pushed as a substitute for (horrors) LISTENING to patients and EDUCATING us as well–so we know that if we start to have strange upsetting experiences it is probably the drug not us, and we should report it. This would save lives. There is also the opportunity cost. Why spend $4000 trying to figure out which third rate antidepressant will screw up my unique biology the least? How much exercise, good food, counseling, social support could be had for that money? Last but not least I was disgusted to see them pitching these tests as providing patients with actual insights into the “genetic or biological basis” of their depression (and supposedly reducing stigma by doing so). From what I understand most of the genes these tests identify are for much more prosaic things, like the enzymes in your liver that determine how fast your body breaks down the drug. They are not tests for the Genes that Cause Depression — correct me if I’m wrong but we don’t know what those are yet, or even whether that’s what we should be trying to find, do we?