This post is meant to be an update to yesterday’s post on the CDC recommendations regarding Tamiflu.  Most of this information will be based off of a 2014 Cochrane Review. Toward the end of this post I will delve into the controversial story of how Roche turned Tamiflu into a household name, and netted the company billions of dollars despite dubious efficacy.


Cochrane: Neuraminidase inhibitors for preventing and treating influenza in adults and children

-The objective of the Cochrane review was to describe the potential benefits and harms of NI’s for influenza in all age groups

-Reviewed all clinical study reports of published and unpublished randomized, placebo-controlled trials and regulatory comments

-Main Results

  1. Time to first symptom alleviation:
    1. Adults: Reduced time to first alleviation of symptoms by 16.8 hrs (7 days to 6.3 days).
    2. Children: No effect in asthmatic children. In healthy children there was a mean difference of 29 hours
  2. Hospitalizations:
    1. Adults: No significant effect on hospitalizations.
    2. Children: No significant effect in children or in prophylaxis
  3. Serious influenza complications or those leading to study withdrawal: In both adults and children there was no reduction in serious influenza complications
  4. Pneumonia:
    1. Adults: Significantly reduced SELF REPORTED, INVESTIGATOR MEDIATED UNVERIFIED PNEUMONIA. No significant effect in trials that used more detail diagnostic form for pneumonia (no studies reported effects on radiographically confirmed pneumonia).
    2. Children: No effect on unverified pneumonia in children
  5. Bronchitis, sinusitis, otitis medial: No effect in children or adults
  6. Harms/benefits of treatment:
    1. Adults: Increased risk of nausea (NNTH = 28), vomiting (NNTH = 22). Decreased risk of diarrhea (NNTB = 43), cardiac events (NNTB = 148). Dose-response effect on psychiatric events at 150mg and 300mg
    2. Children: Increased risk of vomiting (NNTH = 19)


The Tamiflu Story

Tamiflu was a nice little earner. It reflected opportunistic action by a multinational corporation, which was able to be a little bit sharper in its marketing practices than you could now, given the debates over the disclosure of clinical data and how effective the drug was.” – A City financial analyst

-In 2009 Tamiflu, owned by a Swiss pharmaceutical company Roche, was approved by the FDA for treatment of influenza in adults. The details of this process are important. The trial reports supplied to the FDA are called “clinical study reports” which span thousands of pages and provide every detail of the trials. The approval process, and most of the voluminous data are kept secret to protect commercial interests. (4)

-Specifically, the FDA approved Tamiflu for shortening illness duration, but not for preventing complications (4)

-Some of the trials performed by Roche were subsequently published in peer reviewed journals (JAMA, Lancet). However the majority of the data were kept secret.

-In 2006 Cochrane published a review of neuraminidase inhibitors, concluding that Tamiflu was effective at shortening influenza symptoms and reducing complications. In particular, this review supported Tamiflu’s use as part of government preparation for an influenza pandemic (6)

-FDA approval, the Cochrane review, and studies published in major academic journals seemingly provided legitimacy for Tamiflu. This allowed Roche to bolster sales, in particular to government agencies in preparation for an influenza pandemic

-Since Roche launched Tamiflu in 1999 it has generated sales in excess of $18bn. Approximately half of this due to governments and companies stockpiling the drug for pandemic preparations. The US alone spent more than $1.3bn (1, 2)

-Tamiflu is relatively costly and complex to produce, beginning with extracting raw material from the Star anise plant (pictured above) (1)

-Roche’s margins on Tamiflu were impressive by any standards. This was due to the fact that relatively few clinical trials were performed and marketing expenses were limited as many contracts for the drug were negotiated with a handful of decision makers in governments (1)

-Initial sales were driven early in the 2000s by concerns for emerging infectious diseases and by outbreaks of SARS and H5N1 “bird flu.” Emergency preparedness was also top of mind, in the wake of Hurricane Katrina, a heatwave in France, and hand foot and mouth disease in the UK (1)

-Throughout the 2000s questions were raised regarding the safety and efficacy of Tamiflu. Many side effects not initially disclosed by Roche began accumulating in public databases

-Fast forward to 2009. Enter Dr. Hayashi, a pediatrician in Japan. He questioned the conclusion of the Cochrane review team, noting that the conclusion of the review was not based on their own data analysis. Instead Cochrane had derived their report from a meta-analysis by Kaiser et al. The Kaiser study used data from 10 RCTs. Of these only 2 were published in peer reviewed journals, the rest were presented as proceedings of the congress, as abstracts in meetings or conferences, or were available as data from the sponsor (5)

-This prompted Cochrane to attempt to perform a complete analysis of the full clinical trial dataset. Roche balked at this request resulting in a campaign by BMJ and Cochrane to persuade Roche to release the data (6)

-In 2013 Roche released the data much of which was analyzed by Cochrane. The results prompted Cochrane to revise their recommendation, questioning the usefulness of Tamiflu in pandemics and urged governments to review their use of the drug. (6)


Some Cool Links

Art inspired by the Tamiflu saga



(1) BMJ 2014;348:g2524

(2) US Government Accountability Office. Influenza pandemic: lessons from the H1N1 pandemic should be incorporated into future planning. 2011. 320181.html

(3) Treanor JJ, Hayden FG, Vrooman PS, Barbarash R, Bettis R, Riff D, Singh S, Kinnersley N, Ward P, Mills RG JAMA. 2000 Feb 23; 283(8):1016-24.


(5) Gupta YK, Meenu M, Mohan P. The Tamiflu fiasco and lessons learnt. Indian Journal of Pharmacology. 2015;47(1):11-16. doi:10.4103/0253-7613.150308.