Tamiflu: Know the Research Before You Prescribe

No Comments

Some people I know are very diplomatic and couch their opinions with soft words like “may,” “perhaps” or “could” in order to allow some wiggle room. Not David Newman. His recent column in the Huffington Post pulled no punches: “The News About Tamiflu: It Doesn’t Work.1”

If flu-like symptoms have you reaching reflexively for oseltamivir, consider three things: the cost, its very limited efficacy and the problem of creating resistant strains of flu. 


Some people I know are very diplomatic and couch their opinions with soft words like “may,” “perhaps” or “could” in order to allow some wiggle room. Not David Newman. His recent column in the Huffington Post pulled no punches: “The News About Tamiflu: It Doesn’t Work.1”

He referenced many studies, including one by a very smart friend, Mark Ebell. Mark founded a literature review service that focused on high-quality clinical studies called Essential Evidence Plus2 that was ultimately sold to Wiley. He has served as the Editor for the Journal of Family Practice and an Editor for the American Family Physician (the official publication of the American Academy of Family Physicians), and is a prolific author of studies that challenge day-to-day clinical practice. Suffice it to say, he’s a smart guy and has been a major contributor in the evidence-based medicine movement in primary care.

In any case, back to the question, does oseltamivir work? Given that many were suspicious whether oseltamivir really produced benefit for flu patients, there was a move to get Roche to release their unpublished studies and raw data on Tamiflu (why would there be unpublished studies if they all showed that oseltamivir worked? Selective publication?). Many organizations petitioned for the data for years, but, according to Newman, it was finally released to my friend, Mark Ebell.



Payne, D., Br Med J 345:e7303, October 29, 2012

The World Health Organization and the CDC-P have claimed that early treatment with oseltamivir (Tamiflu, Roche) can limit the spread of influenza and reduce the risk of lower respiratory complications due to the flu. Both organizations have recommended its use, which prompted massive stockpiling of this drug by governments worldwide. This author, an editor of the British Medical Journal, comments on the lack of evidence to support these claims and the inability of BMJ to persuade Roche to release unpublished data on oseltamivir trials. Only two of ten randomized controlled trials of oseltamivir (both funded by Roche and authored by Roche employees) have been published, and a Cochrane review concluded that “paucity of good data has undermined previous findings for oseltamivir’s prevention of complications from influenza.” The author chronicles BMJ’s unsuccessful efforts to persuade Roche to release all of their unpublished data, in addition to stonewalling by the WHO and the CDC-P when asked to elucidate the basis for the recommendations for the use ofoseltamivir. The author calls for organized efforts to ensure accountability on the part of the pharmaceutical industry. BMJ has developed a website for the purpose of persuading Roche to provide access to complete data on oseltamivir. The site (www.bmj.com/tamiflu) displays correspondence with all involved parties relating to this issue (Roche, the WHO, the CDC-P, the European Medicines Agency, the European Ombudsman and NICE). 10 references ([email protected] – no reprints)

Copyright 2013 by Emergency Medical Abstracts – All Rights Reserved 6/13 – #25


Ebell and his colleagues e-published in Family Practice in September of 2012 an analysis of the eight unpublished trials and the three published ones. The in-print version came out in April of 2013. The major results of the analysis were:

  1. There was a reduction of symptoms of about 24 hours in patients confirmed to have the flu. This being the case, a cost-benefit analysis needs to be done to determine if spending $120 of somebody’s money (patient or insurance company) to buy one less day of illness is worth it. And that assumes that the conclusion of the authors is correct. One study showed that symptom reduction was 29 hours when started within 24 hours of onset, but only 15 hours if started between 24-36 hours of onset.
  2. Oseltamivir did not decrease hospitalizations, but pneumonia may have been a tad less common with treatment
  3. Regarding side effects, nausea and vomiting have been reported with oseltamivir
  4. Not noted by the Ebell review are neurologic and psychiatric effects noted primarily in Japanese children 16 years of age or younger such as delirium, hallucinations, confusion, abnormal behavior, convulsions and encephalitis. The Japanese protocol for the use of oseltamavir in children is similar to the U.S. protocol. Theories as to why these effects have been essentially limited to Japanese children have been set forward but have not been substantiated.


Ebell, M.H., et al, Fam Pract 30(2):125, April 2013

METHODS: The authors, from the University of Georgia, performed a systematic review of eleven RCTs (4,769 patients over age 12), including eight unpublished studies (largely industry sponsored), comparing oseltamivir and placebo in patients with confirmed or suspected influenza.

RESULTS: The mean reduction in the duration of symptoms with active treatment was 20.7 hours in an intention-to-treat (ITT) population with clinically suspected influenza and 25.4 hours in an ITT population with laboratory-confirmed influenza (ITTI). Symptom reduction was less profound in adults older than 65. In one relevant study, the mean reduction in symptom duration was 28.8 hours when oseltamivir was started within 24 hours of onset, but 14.8 hours when started at 24-36 hours. Overall rates of hospital admission (reported in the ITT population) were comparable in the two groups (1.3% vs. 1.2%), as was admission for respiratory complications, sepsis or dehydration. Pneumonia was less common with active treatment in the ITTI population (0.5% vs. 1.6%, NNT 111), but not in the general ITT population (0.4% vs. 0.9%). The composite of otitis media, sinusitis, pneumonia and acute bronchitis (all originally considered indications for antibiotics) was less frequent with active treatment in the ITTI population (risk difference -2.8%), but not after exclusion of acute bronchitis (not typically considered an indication for antibiotics) (risk difference -0.1%). The only reported death involved a control patient in a study of patients with chronic cardiopulmonary disease.

CONCLUSIONS: The retail cost of a Tamiflu dose pack (ten 75mg capsules) is about $120. A generic formulation is not available in the USA. This review suggests that the benefit of treatment is limited. 23 references ([email protected] – no reprints)

Copyright 2013 by Emergency Medical Abstracts – All Rights Reserved 11/13 – #18

In the meantime Roche has seen a windfall in the stockpiling of its drug throughout the world. Mike Leavitt, Secretary of Health and Human Services under the second Bush administration, had wanted enough treatment courses to cover a quarter of the U.S. population – about 75 million courses. In a communication from Alain Li Wan Po and colleagues from the National Genetics Education and Development Center in Birmingham, UK to the CDC3 it is observed that more than 70 countries are stockpiling oseltamivir and in 2003 at least 220 million treatment courses had been purchased worldwide at a cost of $6.9 billion (comes out to be about $31 per treatment course). With an estimated world population of 6.8 billion, the authors observe that on a global basis stockpiles are very inadequate.

Furthermore, observers note that “replace by” dates are approaching on some of the world’s stockpile and cash-strapped countries may have a challenge keeping up with replacement costs.

Given that current literature is fairly unenthusiastic about oseltamivir, one could challenge the expense associated with the stockpiling of oseltamivir, however, in the event of a pandemic, depending on the virus’s susceptibility, it is possible that the drug may be clinically very useful. And oseltamivir has been shown to be effective prophylaxis for influenza; not as effective as the typical annual vaccine – but then again, vaccines to new strains take months to develop and predicting next year’s predominant flu strains isn’t always perfect. Given the lack of other options, the stockpiling is defended. To help defray costs, the authors outline methods to safely increase shelf life. Given that the 1918 influenza pandemic killed 50 million people worldwide (many in developing countries) the authors stress the importance of developing effective shelf-life extension methods.

But what about the routine patient who presents to the ED with flu-like symptoms? There are a number of considerations: Many will not have the flu – particularly at edges of the annual flu season. It is unclear if oseltamivir will help these folks at all. That brings up another question – what about the flu tests? This is complicated. If knowing that a patient has the flu prevents a bunch of unnecessary blood tests, x-rays and antibiotics, they may be very defendable. If the providers were not going to do the tests and give the antibiotics in the first place then flu tests seem to be a waste of time and money (their sensitivity is generally low – about 70% – but specificity is high). But if, at most, you are going to buy one day less illness with a course of $120 Tamiflu, it seems the case for the use of flu tests is pretty bleak.

Here are two papers, one going back to 2006, discouraging routine use of oseltamavir (and this is before the new data revealing oseltamivir’s uselessness at preventing hospitalizations):



Burch, J., et al, Lancet Infect Dis 9:537, September 2009

BACKGROUND: These British authors point out that, in publicly funded healthcare systems, choices must be made concerning which interventions can or cannot be supported.

METHODS: The authors performed a systematic review and meta-analysis of 26 randomized controlled trials of zanamivir (Relenza) (13 trials) or oseltamivir(Tamiflu) (13 trials), commissioned by the British National Institute for Health and Clinical Excellence (NICE), to evaluate the effectiveness and cost-effectiveness of these agents as treatment for seasonal influenza in adults.

RESULTS: Among individuals with typical signs and symptoms of influenza, pooled data from placebo- controlled studies in otherwise healthy adults are consistent with a median reduction in the interval to alleviation of symptoms of 0.57 days with zanamivir (six trials) and 0.55 days with oseltamivir (four trials). In at-risk populations (e.g., children, the elderly, and patients with COPD or asthma), there was a 0.98-day decrease in the median time to alleviation of symptoms with zanamivir (seven trials); although trends favored oseltamivir in this population (six trials), there was no clear evidence of a decrease in time to resolution of symptoms with this agent. The studies provided insufficient evidence concerning the effects of active treatment on the development of influenza complications in healthy adults or in at-risk populations.

CONCLUSIONS: The authors suggest that, although “it is reasonable to recommend precautionary treatment to people who are at an increased risk of suffering influenza-related complications,” the use of antiviral drugs for patients with influenza is “unlikely to be the most appropriate course of action” given the “debatable clinical importance of their effect on symptom duration.” 62 references ([email protected] – no reprints)

Copyright 2010 by Emergency Medical Abstracts – All Rights Reserved 2/10 – #21


Jefferson, T., et al, Lancet 367:303, January 28, 2006

METHODS: The authors, from the Italian Cochrane Vaccines Field and the University of Queensland in Australia, performed a systematic review of 52 randomized, controlled trials of antivirals for the prophylaxis or treatment of influenza.

RESULTS: In prophylaxis trials, amantadine prevented 61% of cases of influenza A and 25% of cases of influenza-like illness, but was associated with a significant increase in withdrawals due to adverse effects, when compared with placebo. In treatment trials, amantadine shortened the duration of symptoms (fever) by about one day but did not influence viral shedding. Rimantadine appears to have similar effects. In prophylaxis trials, the neuraminidase inhibitorsoseltamivir (Tamiflu) and zanamivir (Relenza) had no effect on influenza-like illness or asymptomatic influenza but did appear to be effective against symptomatic influenza. For post-exposure prophylaxis, the reported efficacy of oseltamivir was 58% for household exposure and 68-89% among contacts of index cases. In treatment trials, when the medication was initiated within 48 hours of symptom onset both agents shortened the duration of symptoms, but by an unknown amount (evaluable only as hazard ratios due to “cryptic reporting”). Active treatment also significantly decreased nasal viral titers, and oseltamivirdecreased the rate of lower respiratory complications. The limited available information on the effects of oseltamivir for avian influenza does not permit the formulation of firm conclusions.

CONCLUSIONS: The authors discourage the use of any of these agents for seasonal influenza control, but suggest there might be a role for the neuraminidase inhibitors during serious epidemics or pandemics. 71 references ([email protected])

Copyright 2006 by Emergency Medical Abstracts – All Rights Reserved 5/06 – #24

And here’s another reason not to use oseltamivir – it doesn’t make sense from a microbiological point of view to give the same drug over and over to attack the same bug. Why? Because bugs mutate under the stress of being exposed to the same drug all the time. You end up with resistance. Maybe it would be one thing if we had a bunch of alternatives ready to take oseltamavir’s place when resistance develops – but we don’t:


Poland, G.A., et al, Clin Infect Dis 48:1254, May 1, 2009

The authors, from the Mayo Clinic in Rochester, MN, comment on the problem of increasing resistance to antiviral agents used for the treatment of influenza. Both amantadine and rimantadine are ineffective against influenza B, and their clinical utility is currently limited due to high levels of resistance of influenza A to these agents. Use of the neuraminidase inhibitor, oseltamivir (Tamiflu) appears to be preferred to use of zanamivir (Relenza), likely due to ease of use, but in December 2008 the CDC-P reported a high level of resistance of influenza A/H1N1 viruses to oseltamivir. Although oseltamivir-resistant viruses are usually susceptible to zanamivir, a switch to increasing use of zanamivir monotherapy might be associated with an increase in zanamivir resistance. The authors suggest that preferential use of a single drug for the treatment of viral illnesses and/or sequential use of antiviral drugs (starting one drug after development of resistance to another) are potentially hazardous therapeutic strategies that increase the likelihood of resistance. They believe that it is time to consider a more rational approach that would include increased immunization against influenza, point-of-care testing and development of algorithms that would target for treatment those patients at the greatest risk for life- threatening complications. They note that the CDC-P’s Advisory Committee on Immunization Practices appears to be poised to recommend annual influenza immunization for all persons aged six months or older. Such a policy would reserve antiviral treatment for those patients who should not be vaccinated or who develop life-threatening influenza illness despite vaccination. They further suggest that consideration be given to multidrug treatment regimens for influenza to replace the current monotherapeutic approach and reduce the risk of development of resistance. 27 references ([email protected] for reprints)

Copyright 2009 by Emergency Medical
Abstracts – All Rights Reserved 10/09 – #21

So it seems there are a bunch of reasons not to use oseltamivir in an ED patient with influenza-like illness: It costs $120 to maybe buy one day’s sooner relief (if you really have the flu and you see your clinician within one or two days of symptom onset). It’s unlikely to prevent serious complications from the flu. It is precipitating the development of resistant strains of flu. And it may be causing some funky neurologic effects in kids (at least in Japan).

With flu season upon us there will be lots of pressure to dole out lots of oseltamivir prescriptions (especially if patient expectations are created through lots of direct-to-consumer TV ads). Doctor, aren’t you going to prescribe the flu pills? So it won’t be easy, but it is worth a try. We are in the supposed age of informed decision making. Just tell the patient the contents of the prior paragraph and let them help decide. It is the fair, reasonable and ethical thing to do.

Leave A Reply