The Tamiflu Debacle

Background:  Oseltamivir (Tamiflu), a neuraminidase inhibitor, was approved by the FDA in 1999.  The majority of the evidence supporting the use of the medication came from trials funded by Roche, the maker of the drug. Safety issues with the drug began sprouting up in 2009, due to case reports in Japan of neuropsychiatric events and these events eventually led to a label warning. The Cochrane collaboration published analyses of the available data in 1999, 2003, and 2006, supporting the use of the drug. However, in 2009, the  Cochrane collaboration began to question Roche about the completeness of the data they were using, which was data from another meta-analysis with 10 RCTs.  Only 2 of those RCTs (Nicholson 2000 and Treanor 2000) were published in peer-reviewed journals.  The other 8 RCTs were presented as proceedings of congress or abstracts in meetings.  Cochrane decided to undertake a complete analysis of full clinical trial data, but had difficulties accessing the data until 2013.  This post will serve as a review of the evidence for and against the use of Oseltamivir (Tamiflu) after the full clinical trial data was finally released.

Jefferson T et al. BMJ 2014 Cochrane Review [1]

What They Did: Systematic review of randomized placebo controlled trials on adults and children with confirmed or suspected exposure to influenza

Outcomes (Divided into Treatment and Prophylaxis Outcomes):

  • Primary Outcomes in Treatment Studies:
    • Time to first alleviation of symptoms
    • Influenza outcomes
    • Complications
    • Admissions to hospital
    • Adverse events
  • Secondary Outcomes in Treatment Studies:
    • Symptom relapse after finishing treatment
    • Drug resistance
    • Viral excretion
    • Mortality
  • Primary Outcomes in Prophylaxis Studies:
    • Influenza (Symptomatic and asymptomatic, always with laboratory confirmation)
    • Influenza-like illness
    • Admission to hospital
    • Complications
    • Interruption of transmission (i.e Viral spread from index cases and prevention of onset of influenza in contacts)
    • Harms
  • Secondary Outcomes for Prophylaxis Studies:
    • Drug resistance
    • Viral excretion
    • Mortality


  • Randomized controlled trials evaluating the effects of oseltamivir for treatment, prophylaxis, and post-exposure prophylaxis of influenza


  • Open label studies and those not involving naturally occurring influenza
  • Also excluded patients who were immunocompromised (i.e. HIV, chemotherapy)


  • 83 trialsobtained for review
    • 23 trials (9623 participants) used for reliability and completeness screening
      • Risk of Bias: 48% (11/23) of studies adequately reported random sequence generation, 65% (15/23) showed adequate allocation concealment, 48% (11/23) showed adequate blinding of participants and staff, and 83% (19/23) showed adequate blinding of outcome assessors
      • High risk of bias for included outcomes as a result of missing data
      • All but 3 trials were under-recruited
    • 20 trials used for formal analysis
      • 11 trials on treatment in adults and 4 trials in children
      • 5 trials on prophylaxis on adults (2 in adults, 2 in elderly, and one in households)
    • In Adults:
      • Oseltamivir reduced time to first alleviation of symptoms by 16.7 hours (95% CI 8.4 – 25.1hrs; p<0.001), and reduced unverified PNA (NNT = 100)*
        • *In trials that used more detailed verification and diagnostic imaging (verified PNA) for PNA, there was no statistical significance
      • Oseltamivir increased risk of nausea (NNH = 28), and vomiting (NNH = 22), neuropsychiatric events (NNH = 94) and headaches (NNH = 32)
      • No difference in admission to hospital, sinusitis, otitis media, or complications classified as serious that led to study withdrawal
      • In prophylaxis trials, oseltamivir reduced symptomatic influenza participants (patients with symptoms of flu) by 55% (NNT = 33) and households (NNT = 8)
    • In Children:
      • Oseltamivir did not reduce time to first alleviation of symptoms in children with asthma but did reduce time to first alleviation of symptoms in otherwise healthy children by 29 hours (95% CI 12 – 47hr; p = 0.001)
      • No difference in unverified PNA, unverified bronchitis, otitis media, sinusitis, or any complication classified as serious that led to study withdrawal
      • Oseltamivir increased risk of vomiting (NNH = 19)


  • Identified and retrieved manufacturer funded and non-manufacturer funded clinical trials and their clinical study reports for evaluation
  • Included only randomized controlled trials testing the effects of oseltamivir for treatment, prophylaxis, and post-exposure prophylaxis of influenza
  • Analyzed intention-to-treat and safety
  • Because of the novelty and size of clinical study reports, the authors subdivided the extraction, appraisal, and analysis of the data into a two stage exercise (i.e. reliability and completeness screening and formal analysis)
  • Studies were only included for analysis if they were complete according to the CONSORT statement, and had internal/external consistency of data
  • Extracted data on quality of the studies included as corroborated by a face to face meeting with all authors
  • Independently cross checked the data on outcomes for statistical analysis to ensure the numbers presented in forest plots matched the actual data from clinical study reports
  • Used the Cochrane risk of bias tool to appraise clinical study reports

Limitations (This meta-analysis/systematic review is very sound in its methodology and the limitations related to it are due to the limitations of the included studies):

  • Incomplete reporting of viral resistance and viral nasal voidance meant that those outcomes could not be analyzed in this review
  • Duration of effect on symptom relief is open to question due to the fact that data on relapse after 5 day treatment period were not reported in included trials
  • Oseltamivir trials did not detect any influenza related mortality events and therefore no conclusions can be drawn for this endpoint
  • Incomplete reporting of influenza-like illness does not allow a proper determination of the effect of oseltamivir on prevention/transmission
  • Treatment trials were mostly under-recruited (i.e. did not achieve sample sizes planned)


  • This is the first review that includes all of the collected data as opposed to the data selectively released by Roche initially. Prior studies only had a proportion of the studies due to unpublished or selectively unpublished data
  • Previous findings that symptoms can be reduced by about half a day remain, but with all the evidence available, it has become clear that reductions in risk of hospitalization or complications is lacking. Prior recommendations stated “For people with a high risk medical condition, treatment with an antiviral drug can mean the differences between having a milder illness instead of a very serious illness that could result in hospital stay.”
  • Other issues with prior publications:
    • No dates for unblinding of the trial database were reported
    • Authorship and accountability for writing of clinical study reports were unclear
    • Study protocols or their amendments were postdated after participant enrollment and some protocols were missing altogether
  • This Cochrane analysis was published after Roche finally revealed all the trial data. This systematic review and meta-analysis found that oseltamivir had a modest effect on time to first alleviation of symptoms; 16.8 hours in adults and 29 hours in children.  Roche has benefited by >$18 billion since its launch of Oseltamivir in 1999.

Author Conclusion:“In prophylactic studies oseltamivir reduces the proportion of symptomatic influenza.  In treatment studies it also modestly reduces the time to first alleviation of symptoms, but it causes nausea and vomiting and increases the risk of headaches and renal and psychiatric syndromes.  The evidence of clinically significant effects on complications and viral transmission is limited because of rarity of such events and problems with study design.  The trade-off between benefits and harms should be borne in mind when making decisions to use oseltamivir for treatment, prophylaxis, or stockpiling.”

Clinical Take Home Point:

In both treatment and prophylactic studies, oseltamivir moderately reduced time to first alleviation of symptoms and the proportion of symptomatic influenza patients, however this came at the expense of increased nausea/vomiting, headaches and neuropsychiatric illness.  Furthermore, most if not all the studies evaluated Tamiflu vs placebo (not standard supportive care – APAP, NSAID, etc…) and additionally there was no benefit for complications, hospitalization, or transmission of influenza.

Okoli GN et al. PLOS One Systematic Review 2014 [3]

What They Did: Systematic review and meta-analysis to assess the effectiveness of neuraminidase inhibitors for use in rapid containment of influenza

Outcomes:Community transmission (Definition = epidemiologically linked cases in settings other than hospitals, care homes, nursing homes, boarding schools, and places of detention)


  • 17 eligible studies
    • 9 randomized controlled trials (10,532 pts) and 8 observational studies (8,740 pts) met inclusion criteria (19,272 pts total)
    • 12 (71%) of studies evaluated transmission in households or discrete household-type settings
    • 5 (29%) of studies evaluated individual transmission
    • 0 (0%) of studies were available for prevention of population-wide community transmission of influenza
  • Neuraminidase inhibitors provided 67 – 89% protection for individuals following prophylaxis
  • Neuraminidase inhibitors provide protection for individuals (OR 0.11; 95% CI 0.06 – 0.20; p<0.001) and households (OR 0.23; 95% CI 0.09 – 0.59; p<0.002) regardless of pre or post exposure in laboratory confirmed seasonal influenza and influenza H1N1 infection


  • Risk of bias in individual studies was assessed at both the study and outcome level in compliance with the PRISMA statement
  • Most RCTs were at low risk of bias for blinding of participants and personnel, incomplete data and selective outcome reporting


  • No studies on avian influenza that met study eligibility criteria
  • Most of the included RCTs were judged to be at high or unclear risk of bias for sequence generation, allocation concealment, blinding of outcome assessors and other sources of bias
  • Nearly all RCTs and observational studies had patients that were both vaccinated and unvaccinated for influenza which causes confounding of the modification of effect for neuraminidase inhibitors
  • Although there was no evidence of publication bias in any of the meta-analyses carried out, it cannot be fully excluded due to the few studies available
  • Variation in treatment compliance between studies is a potential limitation as this information was not provided in many studies
  • There is no safety data reported in this systematic review and meta-analysis


  • The authors of this publication included observational data, including data generated during the 2009 – 2010 pandemic period to potentially inform clinical and public health practice.

Author Conclusion: “Oseltamivir and zanamivir are effective for prophylaxis of individuals and households irrespective of treatment of the index case.  There are no data which directly support an effect on wider community transmission.”

Clinical Take Home Point: Neuraminidase inhibitors may provide some effectiveness to direct transmission of influenza at the individual and household level, but there is no evidence that confirms or refutes the impact at a community transmission level.

Qui S et al. Infectious Diseases Updated Systematic Review 2015 [4]:

What They Did: Updated systematic review and meta-analysis of RCTs evaluating oseltamivir, including studies in Chinese, to evaluate its effectiveness and safety

Outcomes (Unclear, which outcome is the primary outcome):

  • Duration of influenza
  • Prevention of serious complications
  • Hospitalization


  • Only randomized clinical trials
  • Trials including patients diagnosed as having influenza with epidemiological exposure history
  • Full-text articles or essential data could be obtained


  • Trials that were self-controlled
  • Duplicate articles
  • Animal studies or cell-line studies
  • Trials that were assessed as having a lower quality


  • 12 studies with 107,712 patients eligible for analysis (6 in English and 6 in Chinese)
    • Due to heterogeneity of included studies a random effect model was used
    • The weighted mean difference (WMD) = -19.39; 95% CI -32.94 – -5.84; p = 0.005
    • This indicates a positive effect of oseltamivir in shortening duration of symptoms
  • Oseltamivir significantly reduced the duration of fever and influenza-like symptoms
    • Due to heterogeneity of included studies a random effect model was used
    • The WMD = -20.48; 95% CI -28.43 – -12.53; p<0.001
    • This indicates a positive effect of oseltamivir in shortening duration of fever
  • Oseltamivir also reduced the rates of hospitalization (RR 0.79; 95% CI 0.68 – 0.90; p = 0.001), antibiotic usage (RR0.56; 95% CI 0.42 – 0.74; p<0.001), and otitis media (RR 0.78; 95% CI 0.65 – 0.93; p = 0.006)
  • No significant difference was observed with respect to the risk of adverse reactions (RR 1.01; 95% CI 0.95 – 1.09; p = 0.710)


  • Included studies were evaluated the revised Jadad scale for quality assessment
  • Included RCTs in Chinese increasing the size and power of results


  • Due to large variability between studies, data was analyzed using a random effects model
  • Studies in Chinese had lower quality compared to studies published in English
  • In the Chinese studies patients in the control group usually received routine symptomatic treatment or traditional Chinese herbs rather than placebo, which could cause a biased estimation of both benefits and harms of oseltamivir


  • The authors of this paper reperformed a systematic review and meta-analysis due to the exclusion of Chinese trials written in Chinese.

Author Conclusion:“Oseltamivir can effectively alleviate the symptoms of influenza and reduce hospitalization, antibiotic usage, and the risk of otitis media without significantly increasing the rate of adverse drug reactions.” 

Clinical Take Home Point: Due to the large heterogeneity between studies, the use of Chinese herbs in the control group, and the lower quality of studies in Chinese it is difficult to draw strong conclusions based on the studies included, and the authors conclusions should be taken with caution. 

Clinical Bottom Line: 

Despite the recommendations from the WHO, advocating for the use of Oseltamivir, health care providers should discuss the available evidence of benefit, harms, and costs of this medication. Until new evidence demonstrates conclusive proof of benefit over harm, oseltamivir should not be routinely recommended as treatment or prophylaxis in adults or children (Excluding immunocompromised patients).


  1. Jefferson T et al. Oseltamivir for Influenza in Adults and Children: Systematic Review of Clinical Study Reports and Summary of Regulatory Comments. BMJ 2014. PMID: 24811411
  2. Krumholz HM et al. Neuraminidase Inhibitors for Influenza: The Whole Truth and Nothing but the Truth. BMJ 2014. PMID: 24811413
  3. Okoli GN et al. Use of Neuraminidase Inhibitors for Rapid Containment of Influenza: A Systematic Review and Meta-Analysis of Individual and Household Transmission Studies. PLOS One 2014.PMID: 25490762
  4. Qiu S et al. Effectiveness and Safety of Oseltamivir for Treating Influenza: An Updated Meta-Analysis of Clinical Trials. Infectious Diseases 2015. PMID: 26173991
  5. Heneghan CJ et al. Neuraminidase Inhibitors for Influenza: A Systematic Review and Meta-Analysis of Regulatory and Mortality Data. Health Technology Assessment 2016. PMID: 27246259

Post Peer Reviewed By: Anand Swaminathan, MD (Twitter: @EMSwami)

Cite this article as: Salim Rezaie, "The Tamiflu Debacle", REBEL EM blog, October 24, 2018. Available at:

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share via Email

Want to support rebelem?