Skip to content

You won’t believe this story: Tamiflu conflict of interest

Paul Alper writes:

Maybe it is time to return to really important things such as medical swindles in particular, Tamiflu. Consider Tamifu and its financially-influenced and influential supportors as seen from the fabulous Susan Perry of Minnpost:

The group of researchers who conducted the Lancet study [supporting Tamiflu]was described in a commentary that accompanied their study as also being “independent” But is that true? No. As [Jeanne] Lenzer points out (in her BMJ article and in a later follow-up piece), all four coauthors of the Lancet meta-analysis have received speaker’s or consultancy fees, grants or contracts from either Roche, the company that makes and sells Tamiflu, or Gilead, the company that holds the patent to the drug.

And the conflicts of interest don’t end there. The listed funding source for the Lancet study — the Multiparty Group for Advice on Science (MUGAS) — received an unrestricted grant from Roche specifically to pay for the study.

It gets even murkier. MUGAS was co-founded by the managing director of a Belgium-based public relations firm, Semiotics, which receives 20 percent of its funding from Roche, according to Lenzer.
For the record: The authors of the Lancet study deny that Roche had any influence on their study’s design or its results.

And from Lenzer:

The CDC Foundation, created by Congress to “connect CDC to the private sector to advance pubic health,” provides funding to the CDC. The foundation confirmed to The BMJ that the CDC received a directed donation from Roche via the foundation for the [“Take 3”] campaign, stating, “Roche provided a grant of $198,000 to CDC Foundation [which] has an administrative fee of 13.5%, so $174,800 was provided to [the CDC to] support qualitative research into influenza prevention and treatment messaging.”

This is not the only money the CDC takes from the industry. Unbeknownst to many, the CDC receives substantial industry funding through the CDC Foundation. A spokesperson said that over the past three years the foundation has received an average of about $6.3 [million] from the industry a year, 21% of the foundation’s overall funding. Since 1995 the foundation has received funding from more than 150 corporate “partners,” including Gilead, which holds the patent on oseltamivir [Tamiflu], as well as Genentech and Roche, the drug’s manufacturers.

Neither the CDC nor the foundation provided data on how much funding Roche, Gilead, and Genentech have donated to the CDC in addition to that for the Take 3 campaign.

I can relate to this, as I’m working right now with Novartis on a couple of projects which help support the Stan project. It all seems ok to me but I can see the general potential for problems.

P.S. Alper points to this news article by Ben Goldacre, who wrote on 10 Apr 2014:

Today we found out that Tamiflu doesn’t work so well after all. Roche, the drug company behind it, withheld vital information on its clinical trials for half a decade, but the Cochrane Collaboration, a global not-for-profit organisation of 14,000 academics, finally obtained all the information. Putting the evidence together, it has found that Tamiflu has little or no impact on complications of flu infection, such as pneumonia.

That is a scandal because the UK government spent £0.5bn stockpiling this drug in the hope that it would help prevent serious side-effects from flu infection. But the bigger scandal is that Roche broke no law by withholding vital information on how well its drug works. . . .

Wow. Evilicious indeed.


  1. Dan Riley says:

    Not commenting on Tamiflu in particular (which does appear to have very limited effectiveness), but the general framing: Congress created a CDC Foundation to “connect CDC to the private sector”, who did they expect to contribute to the foundation? Of course drug companies, medical device manufacturers, etc. I’d have to guess that Congress created the CDC foundation specifically so that corporations could fund their pet projects. I wouldn’t have been surprised to find that 100% of the foundations funding came from industry, not just 21%, but for perspective it should have been noted that foundation funding is a very small slice of CDC’s $7billion budget.

  2. BenK says:

    Not to be too glib, but I perceive some more general distinctions across the professional communities regarding the tamiflu and other antivirals… as for who has clout, when and where, this becomes important when considering blind spots and conflicts, which are not necessarily one and the same.

  3. Sam Watson says:

    The trouble with Tamiflu, and other similarly politicised public health issues, is that both sides of a debate might skew evidence to support their narrative. The Cochrane review examined seasonal influenza (as opposed to pandemic influenza for which it was stockpiled) and it was significantly underpowered to examine the serious clinical outcomes such as mortality. However, if we think of say Bradford Hill’s criteria, they are generally fulfilled in the case of Tamiflu. Indeed the Cochrane review did find evidence of a reduction in other less serious or proxy outcomes. The point being that there is a lot of uncertainty that remains on the topic, yet many claims seen sure that it’s an industry led scandal.

    • Keith O'Rourke says:

      Agree that this is clear as mud.

      But the scandal regarding lack of access to data should be clear.

      Cochrane did an initial meta-analysis without much of the data, then got much of the data and redid (a not very thoughtful but _standard Cochrane group think_ analysis) was horrified at how much the answer _changed_ and then (as indicated below) considered closing up shop or at least providing read at own risk labeling of their work.

      A more thoughtful analysis (recalling Stephen Senn’s comments somewhere) does seem to suggest some benefit of the product.

      Tom Jefferson, et al (of The Cochrane Collaboration). Risk of bias in industry-funded oseltamivir trials: comparison of core reports versus full clinical study reports

      The conclusion:
      “This approach is not possible when assessing trials reported in journal publications, in which articles necessarily reflect post hoc reporting with a far more sparse level of detail. We suggest that when bias is so limiting as to make meta-analysis results unreliable, either it should not be carried out or a prominent explanation of its clear limitations should be included alongside the meta-analysis.”

  4. Phil says:

    I have no knowledge of, nor comment on, Tamiflu etc., nor the ethics of the big drug questions. But this does raise a question. Suppose a big drug company thinks it has an effective drug. If they do their own study, people will of course be skeptical. But they can’t get the government to fund the study. They say “OK, we will pay the CDC to fund the study and let the chips fall where they may.” Is this a bad thing?

    Until recently I worked at Lawrence Berkeley National Laboratory. A few years ago a company wanted to quantify the effectiveness of their proprietary data analysis approach (related to commercial building energy consumption…not a life and death issue like drugs). We had no funding for such a thing, but they offered to cover the costs. We (I and a colleague) said OK, as long as we can design the study and define the metrics of effectiveness. We provided the company with a portion of a dataset to which they didn’t have access, and asked them to predict the rest. They gave us the answers and we compared them to reality. We did the same with some other statistical models. We found that the company’s approach had about the same accuracy as the best open-source model.

    It didn’t seem to me at the time that we had done anything wrong..and it still doesn’t. But this seems pretty much identical to one aspect of the drug testing issue that Alper implies is unethical.

    I’m not saying it isn’t unethical and I’m certainly not saying the drug companies aren’t doing something evil, I’m just asking whether there’s something wrong with a government agency taking money from a company to do an “independent” study of the company’s product. I put scare quotes around “independent” but…well, what can I tell you, I really do think the study _I_ did was independent.

    • Were the results of your test used as the primary reason for the US Govt to pay the company a billion dollars to run their proprietary technique, conveniently leaving out the fact that free software that already exists could do the same job at no cost? (Obviously, probably not)

      • Rahul says:

        So had the US Govt. been a potential buyer of the billion dollar technique, what mode of vetting do you suggest?

        Finally, someone has to pay someone else to test. We are only arguing about who’s allowed to pay who.

        • Well, whatever the mode of testing. I’d say that all the data about the effectiveness of the method vs the effectiveness of other methods should be made publicly available, and the Govt should base its purchasing decision on seeing the full data.

          To me, it sounds like what Phil did was probably pretty good, and not in any way unethical by itself. But, it’s perfectly plausible that the company might USE the output of that analysis in an unethical way. For example, if it showed they were practically speaking the same as the best alternative with a difference in the 3rd decimal place… and that both of them were very poor, then they marketed their technique to the government claiming “US Government testing shows that our technique outperforms every other technique tested” and then didn’t link/footnote to the study.

          1) Taken a negligible difference and used it to rank themselves first.
          2) Pretended that because they were the best that it implies that they were good, which in my imaginary scenario wasn’t true, all the techniques performed very poorly.
          3) Failed to refer anyone to the actual results so that evaluators could look at the details.

          I think if that were the case, it would be an unethical use of such a study, even though the person carrying out the study (Say Phil in this example) did a good job, and performed the study ethically.

          It’s my impression that a lot of this kind of shenanigans has gone on in the past in the medical industry… defunding studies that early on look to give bad results, leaving out preliminary results and relying only on final testing results that happen to show p < 0.05, recruiting from one group for testing while pretending that the drug is effective for other groups as well… whatever.

    • James C. Whanger says:

      A related question is whether the funding provided was adequate to fully test the hypotheses necessary to actually determine effectiveness. One can imagine a company strategy that would limit the funding to allow for a minimally adequate sample size such that a negative result could be explained away as a sampling issue. Though sometimes this is done unknowingly rather than willfully.

    • Keith O'Rourke says:


      I agree, you just need to be upfront and ideally open to possible outside audits.

      In fact, here we argued that it would be bad for patients to overly restrict such activities and went with full disclosure instead –

  5. Rahul says:

    Naive question: How was the grant to MUGS from Roche “unrestricted” if it was meant “specifically to pay for the study”?

    Does “unrestricted grant” have a specific meaning that I’m not getting?

  6. Simon Gates says:

    As a footnote, the EU has funded a trial of oseltamivir in primary care across Europe- see (not terribly informative):
    It’s using Bayesian methods and an adaptive design so an all round good thing. Starting soon.

  7. Fernando says:

    Regarding conflict of interests I’d like to see a more balanced approach.

    All things considered I’d wager industry has less conflicts of interests day-to-day than the academic researches.

    As I commented here some time ago, a full academic disclosure might read something like:

    My salary, grants, promotions, professional standing, and career all depend on publishing significant findings frequently. I declare these incentives have in no way influenced the integrity of the present research.

    I have never seen that written down in any journal despire considerable chest beating about ethics…

Leave a Reply to Phil