Retraction Watch

Tracking retractions as a window into the scientific process

Recent finding of misconduct by federal U.S. agency sparks debate

with 18 comments

Nasser Chegini

In 2011, the University of Florida assembled a misconduct report about one of its ob-gyn researchers, identifying falsified data in a 2010 paper. But when an investigator at the U.S. Office of Research Integrity reviewed the report, something didn’t feel right.

“I reviewed the data, and I thought [UF] didn’t do their due diligence,” said Kristen Grace, then an ORI investigator, now heading up the compliance department of the Office of Clinical Research at the University of Pennsylvania. “Because the extent of the the falsification was so great.”

So the ORI asked the UF to re-open its investigation, expanding it to include previous years of work by Nasser Chegini, now retired. The institution also hired a new director of research compliance, who oversaw the second investigation. That report, completed in October 2013, was significantly more extensive — it documented intentional falsifications or fabrications in nine papers published between 2003-2008. (Through a public records request, we obtained a copy of this second report, which you can read in full here.) But last month, the ORI issued a finding of misconduct against Chegini that focused on only one paper; the agency said it chose to take a “targeted approach,” since eight of the nine papers had already been retracted.  

The move has prompted a debate — while some argue it’s a pragmatic use of ORI’s limited resources, others (including Grace) are concerned:

I think there were many more findings that could be found by ORI, and I’m just disappointed that they decided to go with one finding.

John Dahlberg, former deputy director of the ORI, told us the narrowed finding sets a potentially dangerous precedent (and included this sentiment in a comment he posted to our story about the ORI’s finding). The federal agency depends on institutions to conduct thorough investigations, he said, which are not trivial endeavors.

…[At the ORI], we were appreciative of the fact that it can cost millions of dollars sometimes to pursue a case…Florida officials have to be angry that so much of their lengthy report has been ignored…If the evidence is strong enough, then there’s no justification for a weaker finding.

We asked Irene Cooke, who oversaw the second investigation at the University of Florida, what she thought about the government agency’s decision to limit itself to one finding. She told us:

UF conducted an internal investigation and shared its finding with ORI as is required under federal regulations. The university also sought to retract the affected publications, ultimately succeeding for eight of nine publications.

The university has long since moved on from this case, but UF shares ORI’s ongoing commitment to maintaining the integrity of scientific research and supports the office’s efforts in this area.

Grace notes that ORI’s mission is to ensure institutions are following regulations and conducting thorough investigations — and this narrowed finding may signal that it’s okay for institutions to limit their scope, as well:

Not following through with Chegini could be detrimental to the reputation of the ORI.

“Basically every paper…was falsified”

Indeed, there was much more that was problematic about Chegini’s work than just one paper, said Grace. So far, nine of his publications have been retracted. And when she examined the first misconduct report, she knew it had only scratched the surface of the problem. How?

Experience. Just looking at the data, and the papers. There was no way that type of falsification, I felt, was going to be limited to one paper. It was really just more of a gut instinct. But having worked in the field for a long time, you get a sense. There was a disregard for data.

Her suspicions were validated once she reviewed the second misconduct report issued by UF in 2013:

Basically every paper they looked at was falsified. And I knew it. I just knew it….A person doesn’t falsify that indiscriminately without having done it for a long time.

According to the report:

The overall conclusion of the Investigation Committee was that the Respondent had intentionally fabricated or falsified data in nine published manuscripts spanning the period of 2003‐2008 that were examined by the committee. …Together, these instances of fabrications or falsifications of data represent repeated and significant research misconduct. This research misconduct has had an impact on scientific research as other laboratories have been funded to repeat or conduct follow‐up research on the Respondent’s work resulting in a substantial waste of time, effort and valuable grant funding.

As part of the second investigation, the committee met formally 25 times, recording interviews with a number of sources. It determined Chegini had fabricated data points to produce the desired experimental results, and even added errors to give “the false representation of experimental error in each experiment.” He’d also falsified numbers — in some cases to suggest treatments had biological effects — and Western blot data, sometimes because the experiments hadn’t occurred.

Last month, the ORI sanctioned Chegini for tweaking the findings in seven figures of a 2007 paper, which hasn’t been retracted.

“A wise decision to maximize the outcome”

Earlier this month, the ORI told us the relatively narrow finding was part of a strategy:

ORI accepts the research misconduct findings of the institution that led to the eight prior retractions, compliments the institution on its handling of the matter, and makes new findings only on the ninth publication, so that the scientific record is corrected.  ORI’s targeted approach in this case enables ORI to conserve resources while timely [pursuing] targeted findings to support deterrence of research misconduct, protect Public Health Service funds, and correct the scientific record.

Not everyone believes the ORI’s decision sets a dangerous precedent. According to John Baumann, the associate vice president for research compliance at Indiana University:

Had ORI looked into every single allegation in which the University of Florida found research misconduct, and confirmed everyone of those research misconduct matters, the outcome would have been absolutely no different.

The remaining papers had already been retracted, Chegini appears to have retired from science completely, and the investigation had been going on for years, said Baumann — and the government agency has an “incredible backlog” of cases:

I know of many institutions that have cases that are awaiting ORI responses, for which the waiting has been more than a year.

He said he wouldn’t want this case to set a new precedent for the ORI’s approach to institutional investigations, but:

I think that in this case it was a wise decision to maximize the outcome, accept or acknowledge the determinations made by the University of Florida, impose sanctions that would have been the exact same sanctions had the ORI conducted the full investigation itself, then moved onto another case…If I were in Florida’s shoes, I would not be displeased with this outcome. Because it resolves the matter in a way consistent with [its] determinations.

Grace and Dahlberg conceded that in Chegini’s case, one could make the case for a smaller finding — many of his papers are already retracted, the investigation had been dragging for years, and he has retired from science. But, as Dahlberg said:

Retirement doesn’t necessarily mean the person wants to leave science.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

Written by Alison McCook

August 17th, 2017 at 11:13 am

Comments
  • Michael August 17, 2017 at 11:47 am

    I just do not get it. A previous RW report (Aug. 8, 2017) mentions he has agreed to be supervised for 5 years. Supervised for what? The individual has retired; like many other researchers and faculty found guilty of misconduct, he most likely went into happy retirement with full benefits.
    The US Penal Code needs to be fixed to accommodate researchers and professors found guilty of serial misconduct (they are nothing but professional con-artists with MD or PhD degrees) to ensure they get rewarded by a stay (however short or long) in the big-house. Plus, they (or the university) should pay back the federal funds that were abused. And, if patients were harmed in anyway (hint: bogus clinical trials based on such research), these Profs need to be honored with an extended stay in the said big house, commensurate with the number of patients harmed (or denied proper treatment).

  • Michael August 17, 2017 at 12:01 pm

    Correction…the said RW report appeared on Aug. 01, 2017, not Aug. 08, 2017. My mistake.

  • Ken Pimple August 17, 2017 at 12:56 pm

    When I read about an investigator, whether for ORI or the FBI, shows great confidence in her or his gut instinct, I get worried.

    John Baumann, whom I’ve know for several years, is right on target.

    • Kristen August 24, 2017 at 4:52 pm

      Agreed, perhaps “historical knowledge and experience” would have been better descriptors. Although, keep in mind, this was not referring to findings in-and-of-themselves, but a determination that a wider scope of investigation was necessary.

  • Tim August 17, 2017 at 1:19 pm

    Michael states an important point. Why are science cheaters not treated the same as anyone perpetrating other kinds of fraud? Maybe they should be treated in a more severe way than ordinary cheaters since scientific fraud has the propensity to be amplified in the scientific environment analogous to biological magnification of pollutants (unless caught early)!

    In terms of repayment, consider not only the direct costs (the grant funding), but other costs as well (for example the inflated salaries and bonuses for being SO productive; these lead to larger contributions to retirement plans, etc., etc.). Then, also, what is the logical argument against the fraudsters being held liable for legal costs incurred by the hosts (institutions) of these parasites… or the troubles and added responsibilities and attacks suffered by those assigned to investigate these dishonest people…or the collateral damage done to those who are innocent but associated with the cheater?

    This is serious business and should be taken as seriously as any other type of fraud. It’s NOT just “academic”; many people and institutions can be harmed in tangible and intangible ways by research fraud!

  • Bulova August 17, 2017 at 3:23 pm

    I agree with Michael.

    See this quote from the article: “and the government agency has an “incredible backlog” of cases”.

    Lots of academics will continue to cheat because there is effectively no punishment for academic fraud. It will continue until there is some.

  • rfg August 17, 2017 at 4:03 pm

    I agree with jail time for scientific fraud. Also, lifetime bans from ever getting paid in any way by a government grant.

    “Plus, they (or the university) should pay back the federal funds that were abused.”

    However, this would the more important deterent.

    I actually favor having the university, institute, company etc that got the grant be required to pay it back.

    Better still – include multipliers (2X, 3x etc.) for cover-ups and obfuscation.

    Too many cases of universities whitewashing for too-big-to-fail researchers, then in many cases going after the whistleblowers.*

    IF they know they’ll face even bigger financial loses, the whitewashing problem may go away.

    *OF COURSE, one should look out for exceptions, such as when the whistleblower is not acting in good faith.

    • fernandopessoa August 18, 2017 at 4:34 am

      “Better still – include multipliers (2X, 3x etc.) for cover-ups and obfuscation.”
      “Better still – include multipliers (2X, 3x etc.) for cover-ups and obfuscation.”

      Another cost is that the money did not go to people who did not cheat. The second and third highest “scorers” for the grant application likely did not cheat. The government (ORI and NIH are branches of the government) should mandate:-

      1. that the money be paid back to the government AND
      2. the money be paid by the cheater’s organisation to the rightful winner of the grant application.

      The confusion caused to other researchers is very hard to quantify, but perhaps another doubling of the money to be paid back to the government should be added on.

    • RetiredRIO August 18, 2017 at 12:03 pm

      “Better still – include multipliers (2X, 3x etc.) for cover-ups and obfuscation.”

      I don’t necessarily agree with the multiplier, but I do agree that universities and companies should be penalized for trying to cover up misconduct. In such cases, ORI has the power to do much more than fine the institution. ORI can take away a university’s or company’s assurance which would stop all NIH funding.

      • rfg August 19, 2017 at 7:49 am

        “ORI has the power to do much more than fine the institution. ORI can take away a university’s or company’s assurance which would stop all NIH funding.”

        Agreed. The risk of losing this assurance is incredibly important.

        This dramatic action should be brought into play more often.

        Unfortunately, I think many institutions that are “too-big-too-fail” view themselves as immune to this nuclear option.

        It also does not stop some institutions from behaving very badly when it comes to the scientific misconduct of a small minority of their faculty.

  • fgp August 18, 2017 at 7:36 am

    Fernando makes the salient point.
    In most contexts, funding for science is a zero-sum competition. For every winner there are losers, and for every big winner (as in multi-million $ NIH grants) there are many losers ranging from students to PI’s to their hosting institutions.
    All researchers are equipped to detect crap, so the fouling of the literature is not such a big issue. But the fouling of the funding landscape has devastating consequences.
    It is infuriating to hear cheat after cheat claim that their fiddlings had no effect on the “scientific validity of their findings.” Of course they may have been correct, but that is because they cheated to get into press first, in many cases with work based on ideas stolen from others owing to access to the review system that was gained by cheating their way in.
    Until their activities become genuine crimes, cheaters will indeed prosper in science.

    • Tim August 18, 2017 at 10:31 am

      I have to disagree with one issue raised here. There are many instances in which there is little indication that some particular data published or presented in some other way (poster, talk, grant proposal etc.) is “crap”. If that was the case, retractions wouldn’t be so problematic and misconduct investigations would be much simpler.

    • LX Zhang August 18, 2017 at 6:49 pm

      Claiming “scientific validity of their findings” does not make sense to me. Mathematicians distinguish “conjecture” and facts with valid proof. If one gives incorrect proof in math, people just will give any credit to that person. So, why biologists say this sentence why they are found to have image problems?

  • Paul Brookes August 18, 2017 at 8:53 am

    First, glad to see Kirsten Grace in a new position – I knew she’d left ORI, but she was always very civil and engaging in my dealings with the organization. It’s a pity John Dahlberg left too. Good luck to both of them.

    Second, the worrying trend identified by John Baumann is far worse than waiting one year. There are several cases for which ORI has not yet ruled at 5 years and counting. A key example is Bharat Aggarwal (now retired), but don’t forget about Dipak Das (now dead), Rakesh Kumar, Gizem Donmez or Falzul Sarkur. These are all high-profile cases for which ample evidence of problematic data is in the public arena (both here and on PubPeer), with academic institutional findings also published in some cases, but no word yet from ORI. One really has to ask what’s going on there? (especially with the departure of talented staff such as Dahlberg and Grace).

  • RetiredRIO August 18, 2017 at 11:55 am

    “Plus, they (or the university) should pay back the federal funds that were abused.”

    In order to make universities pay back the federal funds that were abused, you’d have to determine how much of the federal funds was used for the misconduct. If the funding resulted in 10 publications and only one image was falsified in one publication, should the university have to pay back the entire amount of the grant?

    What if the misconduct was done by a post-doc or grad student instead of the PI? You could make an argument that the PI is responsible for his staff, but what if the PI was the one that caught the misconduct and reported it properly to the institution? Should he lose his funding?

    There are several circumstances around misconduct cases so having a blanket financial punishment for the university isn’t as easy as it may sound.

    • fernandopessoa August 19, 2017 at 3:21 am

      I understand that it is important to determine the extent of misconduct and who committed the misconduct. Who said it was easy? It does need to be done.

  • rfg August 19, 2017 at 7:35 am

    “If the funding resulted in 10 publications and only one image was falsified in one publication, should the university have to pay back the entire amount of the grant?”

    Likely not.

    But, it also seems likely that plenty of robust examples exist on the other side of the spectrum: a single result obtained by falsification or fabrication was the key to winning a large grant or contract.

    Hypothetical examples: Compound R under condition S turns on profoundly important pathway T; drug V cured W percent of patients [or animals] with condition X; virus Y caused disease Z.

    I’m sure that one could construct or even point to examples in other fields

    Hypothetically speaking: identifying one clear-cut case where falsification or fabrication directly resulted in a university being required because of an ORI or other watchdog agency finding to pay back a large grant or contract plus a hefty fine would IMO have a major positive impact on how misconduct cases are handled by universities.

    “Who said it was easy? It does need to be done.”

    I could not agree more!

  • RetiredRIO August 19, 2017 at 7:39 am

    ORI’s new “targeted approach” is extremely disturbing from the universities’ perspective. Investigations can cost universities millions of dollars and resources. We don’t have the option of taking a targeted approach as we are required by ORI and regulation to conduct thorough investigations. Regardless of its current staffing issues, ORI should hold itself to the same standard.

    I’m curious to see ORI’s response when university investigations start making targeted findings on papers that haven’t been retracted.

  • Post a comment

    Threaded commenting powered by interconnect/it code.