Following up a case of scientific misconduct (UPDATED)

May 11, 2010

I recently noted the case of scientific conduct of one Scott J. Brodie, DVM, Ph.D. (ORI Notice). The incomparable writedit has a few thoughts on the matter as well. I, of course, originally just got smart about the mention of PowerPoint presentations.
With the extensive list of NIH grants and papers that were involved in the Finding of Scientific Misconduct, however, I got to thinking. And pulling on the threads a little bit.
First a PubMed search for Brodie, S.J. identifies 48 publications. The ten earliest stretch from 1958 to 1966 then there is a 20 year gap, so I’m going to assume the most recent 38 are from the subject of this misconduct case.
Only one of them (Brodie, Journal of Leukocyte Biology 68:351-359, 2000) has a retraction link on the PubMed listing. And that link points to the wrong retraction, this is the right one. This is the fourth paper listed in the ORI Notice as containing falsified figures/data.


The first paper listed in the ORI Notice is Brodie et al, American Journal of Pathology 54:1453-1464, 1999. Checking PubMed, PubMed Central and the Journal website I find no mention of a problem with the paper.
The second paper listed in the ORI Notice is Brodie et al, Journal of Clinical Investigations 105:1407, 2000. No link on the author search but if you click on the abstract or look at the PubMed Central version you will find a link to a 2008 Expression of Concern:

According to the report issued by John T. Slattery, Vice Dean of Research and Graduate Education at the University of Washington in Seattle, Washington, “Dr. Brodie was found to have falsified images that appeared as Figure 5A in the publication. These images respectively appeared in JCI as representing neomycin gene-marked CD8+ cells before patient infusions, with neo-positive cells showing yellow-red fluorescence and neo-negative cells being purple-blue; in one unfunded NIH grant application labeled as cells harboring HIV DNA (PCR in situ hybridization for gag DNA); and in a second unfunded NIH grant application as depicting alveolar macrophages from HIV+ persons treated with LPS, tuberculin or HIV tat protein-stain for viral RNA. The University’s investigative committee concluded that two or more of these images was falsified.”
There is an ongoing investigation into potential scientific misconduct in the performance of this study, reportedly by the Office of Research Integrity. We will inform our readers of the outcome of this investigation when it is complete.

The third paper listed as containing false data in the ORI Notice is Berrey et al, Journal of Infectious Diseases 83:1466, 2001. In this case there is a note of an erratum that turns up in the summarized author search but there is no direct link there or on the abstract page. The abstract page at the journal lists over two dozen citations of the article but no mention of the erratum. No link on the full text page either. Doing the hard work from the unlinked cite listed on PubMed, it is possible to find the erratum.

A recent investigation of scientific misconduct involving experiments performed by Dr. Scott Brodie suggests that panel D of figure 2, which was provided by Dr. Brodie, is not, as he had claimed, a picture of in situ polymerase chain reaction (PCR) performed on HIV‐1 DNA in lymphoid tissue but, rather, is a picture of in situ hybridization to neomycin of adoptively transferred neomycin‐marked CD8+ T cells. Therefore, the data on page 1470, second column, lines 4-9, suggesting that treatment with zidovudine, lamivudine, and indinavir does not lower HIV‐1 DNA in lymphoid mononuclear cells at the 1‐year time period, is not correct. Subsequent evaluation of HIV‐1 DNA levels in mononuclear cells purified from lymph nodes by use of Taqman PCR indicates that treatment with zidovudine, lamivudine, and indinavir does reduce HIV‐1 DNA levels, from a median of 815 copies/μg of DNA, in untreated individuals, to 9.7 copies of HIV‐1 DNA/μg of DNA, in treated patients, a finding that is similar to what has been reported by Wong et al. [1] and Ngo‐Giang‐Huong N et al. [2]. All other figures, tables, and statements in the article are correct.

As I mentioned above, the fourth article listed in the ORI Notice links a straightforward retraction.
So. About 6-7 years after publication of the faked data, the first indications of a concern were published. Some two years prior to the issuance of the ORI Notice in three of four cases, hopefully the fourth journal was just waiting for the ORI to conclude its process and will rapidly issue a retraction notice.

Now….how about those grants?

I found two grants for which Brodie was the PI.
(NIH RePORTER has apparently been fixed to allow direct linkage of search results! Huzzah!)
The second (and final listed) year for 5R01DE014149-02 indicates it was originally a four year grant with a start in 2001.
The only listed year for 1R01DE014827-01 indicates it was likewise intended as a four year award.
No award $$ data on RePORTER but Research Crossroads claims $547,200 total costs for the two year award and $373,444 for the one year award. Lotta money but, phew, at least the NIH only wasted two years and one year, respectively, on these faked-data applications, right? Nope. Click on the history tab for DE014827 or DE014149 and it looks like other investigators were substituted as the PI.
The projects each ran for the originally awarded four years of funding.
Of course this was probably in the early stages of the investigation, since the first expressions of concern and reference to the University findings of misconduct were published in 2008. And awards are to the institution and not the PI after all.
But it is still…annoying.
__
UPDATE 05/12/10: Following a link posted at writedit’s blog, we find that the investigation into Brodie was concluded in 2003 and only reached the light of day in 2007 due to a lawsuit involving a newspaper.

The exhaustive 16-month long investigation of Scott J. Brodie, accused of manipulating computer images of cells to resemble two distinct images, was concluded in December 2003, but the report was not made public until legal proceedings between The Seattle Times, the UW and Brodie ended last week, in favor of The Seattle Times, which fought to release the investigative documents.

Also of interest is that ORI knew about the case right from the start.

The UW was notified in August 2002 by the U.S. Department of Health and Human Services’ Office of Research Integrity that an individual had noticed some anomalies in Brodie’s work. Investigations began shortly thereafter.

13 Responses to “Following up a case of scientific misconduct (UPDATED)”

  1. Anonymous Says:

    “But it still…annoying”
    Annoying? Wait until you read about the slap on the wrist that Brodie received.

    Like

  2. becca Says:

    I, for one, am outraged- OUTRAGED- that hundreds of thousands of my taxdollar-funded NIH dollars are going to…test early antiretrovial therapy in HIV patients, resulting in “immune preservation…decreased opportunistic infections, and decreased frequency of respiratory and mucocutaneous conditions”. Oh wait. No I’m not. That’s a *good* thing to know. Mind, I’m still outraged the money went to this guy, when it sounds like there are plenty of people who would have done the study without any falsification issues to cast a doubt on the sound aspects of the science. But, to be very honest, these studies sound *important*. Even without the thought of potential ignorant trainees in his labs, or innocent collaborators (who may be able to continue good work because U Wash could reassign the grants), I’m not sure halting the studies would be in the best interests of the public. And while I understand you find it “annoying” that NIH continued to fund the studies… would it really have been cheaper and better for public health to do halt them? I’m not sure.
    Additionally, I am right and fully perplexed by why somebody would say that their therapy did *not* reduce HIV-1 DNA when it *did*. Seems to me he lied to make his data seem less impressive than it is. Unless there’s an ax to grind about the possibility for resistance to develop (more likely in the presence of continued DNA production???) or something like that, I’m just baffled. Anyone want to enlighten a poor ignorant grad student here?

    Like

  3. jojo Says:

    What the hell does a “retraction” even mean if the paper isn’t ever retracted?
    This seems to happen a lot. Going to original links of many retracted papers, the paper itself is totally intact, and if anything there is only a reference in the side bar to a “retraction statement” from the author or editor about how the paper is all lies and shouldn’t be read or cited.
    Why do we keep the text of retracted papers publicly available on journal websites? They should be retracted, for real, because they are misleading to people trying to learn the field! At the very least, the first page you see when you click the link to the paper should be the retraction statement, and then at the bottom you can click “yes, I still want to read this retracted, falsified piece of crap” in order to read the full text of the paper.

    Like

  4. DSKS Says:

    “I’m not sure halting the studies would be in the best interests of the public.”
    When a serious level of fraud is implicated, of course it’s in the best interests of the public to do just this until a thorough review has established that the original grant was not awarded on the basis of fraudulent preliminary data (which might throw the entire project’s feasibility into question). As it is, the good science is having a run for its money trying to combat stuff like HIV, so I see no reason to think that bringing the hammer down on crap science is going to be all that costly relatively speaking.

    Like

  5. becca Says:

    DSKS- would you say that if your child/parent/loved one were the one being treated in the clinic? If I were the patient, or family of a patient involved in this protocol, I wouldn’t want them to stop the early-treatment protocol just because it was found the HIV DNA actually *did* go away. The devil is in the details, and they are really strange in this case.
    No one is suggesting Dr. Fraud be allowed to continue. What DM finds annoying is that U Wash’s funds weren’t taken away. Which strikes me as possibly hard on grad students/techs/postdocs/other minions who were *not* involved in the shady data, and rather shockingly hard on the patients involved in this still obviously somewhat scientifically sound course of treatment. Unless I’m missing some data that suggests this was heinously unsound medical practice…?

    Like

  6. antipodean Says:

    Becca
    How does a patient give informed consent to be in that trial when the rationale is now known to be fraudulent?
    This isn’t rat science afterall

    Like

  7. neurolover Says:

    “Why do we keep the text of retracted papers publicly available on journal websites? ”
    Because retraction doesn’t actually mean that we get to rewrite history, and disappear something that existed. The appropriate scheme should have the retraction appear prominently in the paper, and on downloadable PDF’s of it (that new PDF could state that it was altered on date xxxx in order to reflect the retraction).
    This case is super icky.

    Like

  8. ginger Says:

    Becca – maybe Brodie had a pet theory about HIV latency and treatment cessation? The paper’s clinical and in-vivo RNA data all supported the idea that earlier HAART treatment improved outcomes. They also found that there was no difference in drug resistance between immediate and delayed treatment. Brodie’s fraud was supplying a piece of evidence that purportedly showed no difference between delayed and immediate treatment in viral DNA present in lymph node cells.
    If viral DNA was lower in immediately than delayed treatment people, then you’d have a more potent argument to continue treatment in people who weren’t experiencing severe symptoms (as is the case in our reality-based universe.) In Brodie’s world, you might not be able to justify delaying treatment, but you might be able to safely stop treatment after a year for a treatment vacation if the patient’s not feeling sick and he’s suffering a lot from medication effects.
    I’m not saying this is right – just that it’s one possible explanation.
    Also, jojo, we don’t make retracted papers vanish with a poof because 1. sometimes, as in this case, they still have valuable information from non-fraudy authors and 2. in order to provide a context for the fraud and figure out how much the field has been affected, people need to be able to publish the fraudulent or incorrect paper. What I don’t understand is why there’s not a watermark that says “SEE ERRATA” on the versions provided as pdfs. Maybe that’s a throwback to the fact you can’t do that with hard copies already in libraries.

    Like

  9. DSKS Says:

    “DSKS- would you say that if your child/parent/loved one were the one being treated in the clinic?”
    Whether I would or wouldn’t has no bearing on what constitutes appropriate ethics on this subject. It’s a little like defending torture with the old, “But what if the subject was hiding your daughter in a meat locker?”
    Well, obviously I’d take the pliers to him… and then I’d rightly go to jail.

    Like

  10. Kevin Says:

    Why isn’t the article taken down, corrected, and put back up? That seems like the best thing for those who want to read the paper. I suppose it’s a lot of work to re-write the whole thing to change, as in this case, just a few sections of it but it would definitely avoid people missing the erratum message.

    Like

  11. Drazen Says:

    I agree with Kevin. Those kind of mistakes cannot be allowed. But hey, as long as we learn from our mistakes it should be fine.

    Like

  12. formerscientist Says:

    Unfortunately this kind of thing is way more common than ORI finds. I think most grad students and postdocs know of or are guilty themselves of “Photoshopping” and many of us have seen other investigators reuse data with different captions and Methods. Raising these issues when you witness them are usually the end of a science career

    Like

  13. Lisa Glazatcheff Says:

    Deputy Assistant Secretary
    Office of Grants and Acquisition Policy and Accountability
    U.S. Department of Health and Human Services
    Hubert H. Humphrey Building, Room 336-E
    200 Independence Avenue, S.W.
    Washington, D.C. 20201
    April 3, 2014
    Dear Ms. Gunderson:
    Documents recently obtained through the University of Washington Office of Public Records show that the spoliation and withholding of evidence occurred during the investigation of Dr. Scott J. Brodie for alleged scientific misconduct. These recently produced documents show that data was deleted during
    the investigation from one of the key computers used by Dr. Brodie (hereinafter “SB Residence”). The data deletion occurred during the University of Washington’s investigation, and thus, predated the administrative law judge’s decisions, the decision to debar Dr. Brodie, and those of the federal courts in this matter. The spoliation of Dr. Brodie’s research records denied ORl, and Dr. Brodie key evidence.
    “Spoliation” refers to the destruction or material alteration of evidence in pending or reasonably foreseeable litigation which denies respondent access to evidence from which it could develop its defenses adequately” Henderson v. Tyrrell, SO Wash. App. 592, 604-07, 910 P.2d 522, 531-32 (1996); See also, Silvestri v. Gen. Motors Corp., 271 F.3d 583, 593-94 (4th Cir. 2001).
    DISCUSSION.
    In ORI’s investigation of misconduct the “[a]uthentication of a scientific image requires access to the original data.” ORI’s instructions to investigators emphasize that electronic media, such as the images from computer hard drives need to be “reflective of the actual results”, and “shown to have
    originated with a particular individual.” These instructions are part of the institution’s obligation, and that of ORI, to preserve all research records.
    See Office of Research Integrity, ORI “Forensic Images Samples”,
    4th bullet point, for the quick examination of scientific images, available
    at http://ori.hhs.gov/samples (last visited Jan. 29, 2013). http :/ /www.ohsu.edu/xd/research/uploa… (see Slide #3), http://www.upenn.edu/researchs…, (see Slide #52), http:/ /healthj oumalism.org/resources-tips-de…,
    http://ori.hhs.gov/images/ddbl… vol17 no4.pdf, p.3, ORI and OHRP Compliance Oversight: Recent Cases and Initiatives, Health Care Compliance Association’s “RESEARCH Compliance Conference (Apr. 22,2010, audio mp3)
    Recent documents produced by UWOPR show that documents were apparently destroyed. Mr. Diem admits that a computer was purchased for Dr. Brodie; that the computer was not set up or used in the lab; and, that it was sent directly to Dr. Brodie’s home. Exhibit l, p.4. Mr. Diem also admits that the computer apparently was not returned to the lab “until after the investigation began.” id. Mr. Diem was able to clearly identify the computer. Mr. Stensland, who worked with Mr. Diem, admits that the computer “did mysteriously appear in the lab after the investigation started.” Exhibit 1, p.
    In the December 3, 2003, email, Diem wrote:
    To all concerned:
    The computer with the Lab Med tags 1193702 and 30021731 was
    ordered specifically for use by Scott Brodie at home. When it arrived in the
    T-293X lab it was transported to Scott’s house without being set up or used in
    that lab. It did not return to that lab or to the Rosen lab until after the
    investigation began. We have subsequently deleted most of Scott’s files and
    have using it for general Office software purposes. Significantly, Mr. Diem
    admits that: “We have subsequently deleted most of Scott’s files …. “Exhibit 1,
    p.4.
    The acts of Mr. Diem and others caused Dr. Brodie actual prejudice in the defense of his case. The deletion of data from Dr. Brodie’s computer prevented him from having the opportunity to access data that he had prepared, and unlike the SB Home computer, had been in his exclusive control.
    The deletions also made it impossible to refute claims by Dr. Mullins that many
    of the questioned images found in Dr. Mullins grants came from Dr. Brodie.
    A review of the inventories by Dr. Brodie’s wife, who has been assisting her husband in this matter, discloses that there is no reference to the SB Residence computer identified by Mr. Diem. The UW 2003 report, and its 2004 errata, moreover, do not appear to reveal that the data on Dr. Brodie’s computer was deleted after the investigation began. See Declaration of Elizabeth H. Glazatcheff, attached.
    Neither you as the debarring official nor the ALJ was aware that
    the spoliation of evidence had occurred during UW’s investigation or that data from Brodie’s home computer, “SB Residence”, was deleted during the administrative proceedings. The evidence relied upon by the ALJ came from a computer identified as “SB Home”, which is not to be confused with the computer from which Dr. Brodie’s data was deleted.
    It is now clear from the UW public records responses in 2013 that Dr. Brodie’s requests for his data could not be fulfilled because the records were deleted by UW employees.

    CONCLUSION.
    Evidence critical to Dr. Brodie’s defense was deleted during the course of the investigation. The fact of the spoliation was not known to Dr. Brodie, ORI or the ALJ.
    Dr. Brodie, accordingly, was never given his entire computer generated data or research data for his defense. In turn, the ALJ was misled into believing that Dr. Brodie was provided all his information necessary to defend against the claims of misconduct.
    Dr. Brodie, requests that you reopen this matter and rescind the debarment of Dr. Brodie.

    Signed by Jack Young
    Counsel for Dr. Scott J. Brodie

    Like


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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

%d bloggers like this: