We need to determine who benefits most from flu treatments

BMJ 2003; 326 doi: (Published 05 June 2003) Cite this as: BMJ 2003;326:1239
  1. Lucy Hansen (s0198608{at}, consultant physician1
  1. 1 University of Edinburgh, Edinburgh EH9 1QH

    Influenza accounts for about 20 000 deaths and 110 000 hospital admissions each year in the United States alone.1 In their meta-analysis of the neuraminidase inhibitors zanamivir and oseltamivir Cooper and colleagues have included the small amount of information available that allows separation of subjects into healthy adults and high risk individuals.

    Both zanamivir and oseltamivir reduce the median time to resolution of symptoms by up to one day based on intention to treat, with similar results on confirmation of flu positivity. No clear difference between the healthy and high risk groups is apparent. The prophylactic use of each drug resulted in a more impressive 70-90% risk reduction in both post-exposure prophylaxis and prophylactic treatment during the time of year when flu is most common (seasonal prophylaxis).

    Subjects were monitored for only three to four weeks, however, and a large minority remained symptomatic at the end of this time. This is consistent with my observations that many patients admitted to hospital with complications following a bout of flu have a four to eight week history of symptoms. No studies have compared the response to treatment in vaccinated versus non-vaccinated subjects, but one study of vaccinated, elderly residential patients treated with seasonal prophylaxis reported a 92% relative reduction.2 This emphasises the often forgotten fact that the vaccine is only 70% effective and has only short term benefits. Rather than neuraminidase inhibitors being an alternative to vaccination, they might be an additional treatment in high risk groups, particularly during epidemics or local outbreaks.

    It is difficult to see what important new information about the treatment of flu this meta-analysis offers. As is often the situation with new drugs, information from new studies is essential before neuraminidase inhibitors will become widely used: characterisation of the type and severity of symptoms and end points such as “return to normal activities” should be automatically included; trials should continue for longer; and data collection should provide more details of the type and severity of complications and admissions to hospital. Studies concentrating on the different high risk groups may define those who will gain most benefit from treatment and should incorporate information on vaccination status. In addition to comparative studies of the two neuraminidase inhibitors, combination therapy (vaccination, the M2 inhibitors (amantadine and rimantadine), and neuraminidase inhibitors) may prove an effective means of reducing morbidity and mortality in both treatment and prevention of flu. Results of studies to date do not provide adequate evidence of a cost effective treatment for flu,3 but new, more clearly directed research will hopefully clarify which groups will benefit from treatment with neuraminidase inhibitors, alone or in combination with other established treatments.


    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    View Abstract