Alison Bass, your Mayberg facts are wrong, wrong, wrong

Note: I attempted to post shorter versions of this at Alison Bass’s blog as a comment, but for some reason none of the comments would go through.

Alison, when I attempted to correct your errors before, I was less than brutally frank, was in fact rather generous, methinks, because I wanted to leave you a graceful way to admit error. But I see you have doubled down. I’d as soon leave this lie, as it were, but it’s important to correct the record, which you have filled with false and easily falsifiable information. 

So let me say more directly: In virtually everything you accuse Mayberg of in this and your previous post, you are wrong. You are wrong in your facts. You are wrong in your assessment of Helen Mayberg’s priorities,  motives, and character. And you are wrong, and seem to insist on continuing to be wrong, despite being already presented with evidence that would have allowed you — and in fact asked you — to correct your error. 

Alleged failure to disclose data:

Let me start with what you end with in your post above: Your repeated assertion, already corrected once in my comment on your previous post, that in her talk at the NASW, Mayberg did not report the data on the 20 patients in her Toronto study. 

Mayberg DID report the data on the Toronto study, as she reported it exactly as you cited it above. She did not give the data on a separate Atlanta study, because that Emory/Atlanta study is still unpublished. But she DID report the data on the Toronto patients, and I told you that before, but you keep repeating that she did not report it. I feel stupid repeating this, but saying it only once before apparently didn’t work. She showed a slide with the data, and she talked in specific about the outcomes as well. She noted improvement in 60% of the patients and full remission in 35%. 

Is that data encouraging or otherwise significant?

You think not, and in fact you imply that she’s making a big deal out of unimpressive data that is “not encouraging.” You’re not impressed with 35% full remission in patients who were almost inconceivably depressed. That’s your prerogative. But readers should understand what response and remission means in this group of 20 patients she treated in her pilot study. Mayberg related all of this at her talk.

As Mayberg noted, full remission, by the most common depression standard, means reducing symptoms enough that a patient won’t score as depressed in the Hamilton Depression Scale. The very top score possible in the Hamilton scale is 38. To be not depressed or in remission, you need to score under 8. Here’s the scale: 

0-7 = Normal 

8-13 = Mild Depression 

14-18 = Moderate Depression 

19-22 = Severe Depression 

≥ 23 = Very Severe Depression

Most of us have never seen people with scores over 20, for those people rarely get out and about. Their lives are deeply, intensely miserable. They were catastrophically, almost catatonically depressed. 

The average score of the patients in Mayberg’s Toronto trial was 25, and if I remember correctly. All scored over 20, and all had been severely depressed for years despite undergoing multiple drug treatments, CBT, and, for 17 of them, electroconvulsive therapy. None  — none — had ever responded significantly or lastingly to any other treatment, including (for 17 of the 20) CBT. This was by design; Mayberg didn’t want to be doing an experimental treatment on people who hadn’t exhausted every possible remedy. But this meant her patients for this pilot study were people for whom nothing worked, and who remained horrifically depressed, in an especially excruciating sort of hell. 

To be scored as cured — to reach remission — these deeply sick patients had to be brought to a Hamilton score below 8. Their depression scores had to be cut, in other words, by some 65%.  

To Mayberg’s experimental treatment, as you note, 60% had significant response, and 35% fully remitted. The average Hamilton score dropped from 25 to 15 — even though about a third didn’t quality as responders, because their scores didn’t drop more than 15%. (Mayberg is trying to figure out why those didn’t respond; she wants to either figure out how to make them well or identify why this procedure can’t make them well, if that’s the case, so she won’t needlessly use an intrusive treatment on people who won’t respond.) Meanwhile, 35% of these patients achieved full remissionn — they went from the most excruciating, debilitating depression to normal. Most of the other 25% who did respond went into the moderate or mild range — the low teens, and some the upper single digits, of the Hamilton scale. THey were depressed. But they again had lives. 

You assert that this is not impressive. The patients thought it was. It is also scientifically significant: Even as you throw stones at delusory targets, her work is helping to displace the neurochemical model of depression that has led to the egregious overuse of largely ineffective antidepressant drugs that you and I have both rightly lamented.

We can argue, of course, over whether 60% response and 35% remission in the 20 Emory patients is “encouraging.” But we should not be arguing over whether Mayberg disclosed that information at the talk, for she reported it quite explicitly, and it’s in the published literature, as I told you before in comments attempting to correct your errors at the previous post, even providing the  links (paper; pdf).  She published it and she talked explicitly of it, and showed a slide of it, at the talk. Yet you keep saying she’s hiding it and ask why she didn’t tell us about it at the talk. 

Did she disclose her industry ties?

You say she didn’t. She did. She noted she received equipment donations and that she had consulted for ANS/St. Jude Medical. She showed this information in a slide during the talk, and at my request today sent the slide to me today. Here is the slide:

This was the second slide of the talk (after the title slide), and she drew attention to it by saying she wanted to make some disclosures. This information was the only content of the slide. The letters were cleanly laid out and plenty big enough to read, and she left it up there long enough for anyone interested — as you claim to have been — to read it. It’s not terribly ambiguous. Many of these relevant interests were also listed later on slides that described the results of individual studies. I myself do not recall exactly what she said out loud about any consultancy with ANS, and there’s no video posted, so I can’t check. But I do remember her drawing attention to her relationships, and, more to the point, the consultancy was plainly and explicitly listed on the slide. Given your intense interest in her disclosures, I don’t understand how you missed it. But you did. Your repeated accusation that she failed to disclose her consultancy is just plain wrong

You complain that you thought St. Jude Medical was a hospital, and you imply that your confusion arose from some intention of Mayberg’s to deceive. Yet Mayberg  disclosed both ANS and St Jude Medical as partners in the research — see above — and she identified ANS as a device maker and a consultancy client. St Jude Medical is the company that bought ANS. So in separately identifying ANS and St Jude Medical, Mayberg effectively twice named the entity in question, which is, and has been since 2005, ANS/St. Jude Medical, the instrument maker.  That you mistook one of those names as that of a hospital named St. Jude Medical Center— a hospital that apparently has a neuromodulation division, at least in your read of the slide— does not mean Mayberg was trying to mislead you. It means you made a mistake. If Mayberg had been wanting to mislead you, she not have listed ANS, and in fact would not have opened her talk with a slide about disclosures, and would not have listed those relationships on it

You’re making a tangle of a simple truth here. Mayberg made no attempt to hide these relationships; she in fact offered an explicit right-up-front disclosure of them in her second slide; and she simply did not make the failures to disclose you accuse her of. Your memory and impression directly contradict the reality that was before you that day.

“History of failure to disclose…”  

Here you are almost half-right but actually, alas, again completely wrong. 

When the Nemeroff et al VNS paper you refer to above was published in 2006, the paper did indeed fail to disclose conflicts of interest. However, your accusation that Mayberg failed to disclose is wrong. She did disclose. She and the other authors all disclosed their interests (they were paid for their time in a working group that did the review that was published), as was requested by a journal editor. But for some reason — and whatever it was, it was not under Mayberg’s control — the journal did not publish that disclosure paragraph in the print issue or the original web version. Later the journal published a corrected version with proper disclosures online.  

This is obviously not ideal. But again, the failure was the journal’s, not Mayberg’s. She disclosed; the journal didn’t. The journal’s failure became a big deal because the lead author, Howard Charles Nemeroff, who was head of Emory psychiatry at the time (and thus Mayberg’s new boss; she had recently moved to Emory from Toronto), was the journal’s top editor and had been involved in other failures to disclose. Nemeroff indeed did have a history of failure to disclose. And the failure on this paper was not, as your post above seems to suggest, the sole reason he lost the chairmanship. It was just the final straw

As for Mayberg, however, she disclosed her interest but the journal didn’t print them. You err in saying she failed to disclose. I can understand you concluding that she didn’t disclose, since no disclosure was originally printed. But the fact remains that you are wrong in saying she did not. This is not a trivial distinction, since you use this one failure to disclose to claim she had “a history of failure to disclose.”

Altogether, of course, the VNS paper was problematic, both because of the failure to disclose, because the company’s input created conflicts, and because the first draft was reportedly written, either largely or in part, by a ghostwriter, who was thanked in the acknowledgments but not listed as an author. (A Cyberonics scientist also took part but was listed as an author.) Such ghostwriting is a practice that many, including me, find questionable. 

So what are we to make of Mayberg’s participation in this review and her co-authorship of this paper? It’s obviously not her finest moment. But to know what to make of it, we must ask whether that episode appears to be part of a pattern of behavior and an acceptance of serious conflicts of interest and failures to disclose, or whether it seems to be an aberration that she learned from. I would argue the latter. Mayberg’s conduct before and since then suggests that was the case. She does not court or accept big lucrative relationships or sponsorships with device makers — even though some of those makers would very, very much like to have such relationships with her. She assiduously avoids the sort of hype and commercialization that is far too common in the medical device industry. She takes pains to describe her procedure as experimental, a pilot, and in need of double-blinded trials. She always notes that it’s appropriate for only a tiny portion of the depressed population. And I don’t believe you’ll find her name on any more ghostwritten papers. She saw the ugly, and she recoiled. 

Aside from that stands another, more proximate problem in your post above: You haves used that one failure to disclose, which never actually occurred, along with  others you falsely her of at the NASW talk, to assert in your very headline that Mayberg has a “history of failures to disclose conflicts of interest.” As I just explained, your one example is flawed, false, and erroneous — she didn’t fail, the journal did. Your other example is her alleged failure to disclose her consultancy in her talk at the NASW conference, and there too you are demonstrably, completely wrong. So where is the rest of this history? 

Getting it wrong

Getting the facts wrong is not good, but, well, it happens. The bigger problem in all this is that you have so eagerly harnessed these errors — and so far refuse to acknowledge or correct them — in an effort to paint Mayberg as someone she is not:  as someone who oversells results, hides data, and exploits her research, and by implication her patients, for financial gain. There are people out there like that, and I fully support and indeed take part in criticizing them. They are the bane of medicine. But here you have mistaken your mark. Helen Mayberg is not the shill you suggest. She is, rather, and has always been, a) deeply concerned with figuring out a model of depression to replace the flawed neurochemical model that has led to so much overmedication, and b) intensely, deeply, movingly concerned with the welfare of her patients and c) consistently vocal in her concern that this treatment not be overused or oversold. One of her patients wrote you to tell you this, and of Mayberg’s caution about expectations in generating expectations about the treatment. The patient describes Mayberg’s caution as almost excessive. I saw that myself. When I did the reporting for the story I wrote about her work in the Times in 2006, Mayberg took pains to convey — and to urge me to include in the story — her concern that the treatment might become too trendy and get used too widely and by people who didn’t use it well and/or on patients who were not as depressed and without other options as the patients she works on. 

So to sum — and if I repeat myself, it’s because saying all this once didn’t seem to take effect — your posts assert two main things: that Mayberg failed to disclose relationships and data in the talk at NASW, which she in fact did plainly disclose and discuss; and that she has a history of failure to disclose, which failure does not exist.

You also assert, by implication and in your direct use of accusations based on facts that don’t exist, a cynical and erroneous assessment of her motives, principles, and character. Quite frankly, you seem to carry a bias so dense that facts can’t penetrate it. You fail to see and hear things plainly displayed and said before you. You fail to acknowledge corrections when they are offered in good faith.  It’s as if you went to her talk having perceived some cues — device maker; depression treatment; past association with Nemeroff — that provoked a conditioned response — Attack — and a confirmation bias that nothing can penetrate. Your target image went faulty. Mayberg simply isn’t what you think she is, and she has not done the things of which you accuse her. 

I’m at the neuroscience meeting, and a couple of days ago I was talking, not for attribution, to a medical ethicist: someone keenly aware of the dangers of conflicts of interest. This person knows Mayberg not by chance, but because Mayberg too is keenly concerned with conflicts of interest, and because she has talked with this person both in private conversations and in public and semi-public conversations at conferences about the use of medical devices, about conflicts of interest and the eagerness of some in the medical device community, both doctors and companies, to oversell the effectiveness of devices so as to make money. 

I told this person about Mayberg’s talk and your accusations. This person laughed and said, “She’s whacking Mayberg? Picked the wrong example. It’s hard to think of someone who’s more careful about these things.”

If we journalists are to have credibility when we call out conflicts of interest, we need to check our facts, and if we get them wrong, we need to acknowledge that when corrected. I’d be pleased to see you do that in this case.  


MAY 19, 2011:  Fixed a couple typos, added some underlines and bold.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s