2 February 2012
An updated Cochrane review has found that although oseltamivir effectively reduced the duration of seasonal influenza symptoms by about a day, it did not reduce the risk of hospitalisations to a statistically significant degree. This finding is controversial and relates to questions about which data are accessible and admissible for review.
Action
Health professionals considering the use of oseltamivir (or amantadine or zanamivir) in managing seasonal influenza should follow NICE guidance. This guidance does not cover the circumstances of a pandemic, impending pandemic, or a widespread epidemic of a new strain of influenza to which there is little or no community resistance. Users of the NPC patient decision aid relating to the use of oseltamivir for seasonal influenza in otherwise healthy adults should note that the evidence for the drug’s effectiveness in preventing lower respiratory tract complications requiring antibiotics is derived from the paper by Kaiser et al which is an important component of this controversy.
What is the background to this?
The controversy around the evidence for influenza management with these drugs first emerged in 2009, and was discussed in MeReC Rapid Review 912. The 2006 Cochrane review of neuraminidase inhibitors for preventing and treating influenza in healthy adults (now withdrawn from the Cochrane library) had concluded that oseltamivir had a beneficial effect on preventing complications of influenza in such people. However, the 2009 Cochrane review (also now withdrawn from the Cochrane library) concluded that no statistically significant benefit from oseltamivir on the risk of influenza related lower respiratory tract complications in otherwise healthy adults had been shown (risk ratio 0.55, 95% confidence interval [CI] 0.22 to 1.35). This difference arose because the authors decided not to include some of the data from a 2003 paper by Kaiser et al in its 2009 update, which had been included in 2006. The Kaiser et al paper was itself a summary of data from ten randomised controlled trials (RCTs) of which only two were fully published and available for review. The Cochrane group was not able to obtain sufficient data regarding the other eight studies in time to complete its review.
In both Cochrane reviews, the point estimates indicated a benefit on complications, but in 2009 (presumably because there were fewer people with fewer complications in fewer studies in the 2009 review) the 95%CI included 1. In other words, it could not exclude the possibility that there was no effect, or a harmful effect, beyond the level of ‘reasonable doubt’ conventionally used to define statistical significance. Roche, the manufacturers of oseltamivir, undertook to make unpublished data available regarding the effects on complications and the Cochrane group undertook to analyse it. Both Cochrane reviews concluded that oseltamivir is effective in alleviating symptoms in otherwise healthy adults with flu-like illness if taken promptly.
What are the most recent developments?
The story since 2009 has been summarised in two BMJ articles: a feature article by Cohen (a member of the BMJ staff) and an analysis article by Doshi, one of the Cochrane reviewers. Within a few weeks of issuing its undertaking to do so, Roche provided clinical study report data to the Cochrane reviewers. However, it provided only part of the clinical study reports – the summary of the study methods and the results. The company says this is enough for the Cochrane group to conduct their review, but the Cochrane reviewers disagree. The Cochrane team then obtained data from US, European and Japanese regulators. These included published data and unpublished data obtained under freedom of information arrangements. The authors explain how they used these data in their full report.
The authors included and analysed data from 25 studies (15 oseltamivir and 10 zanamivir studies). They rejected a further 42 studies due to insufficient information or unresolved discrepancies in their data. The included trials were predominantly conducted in adults during influenza seasons in both hemispheres. A small number of studies were conducted in older people residing in care homes and in people with underlying respiratory diseases. They judged that the studies had adequate randomisation and blinding procedures, but that imbalances in the analysis populations available (intention to treat [ITT] influenza-infected) left many of the studies at risk of attrition bias. All the studies were sponsored by manufacturers of the drugs. One important limitation in the data is the lack of standard definitions of complications among the clinical trials. The reporting of cases of ‘otitis media’, ‘pneumonia’ or ‘bronchitis’ was based on local centre definitions.
What does the most recent Cochrane review conclude?
Time to first alleviation of symptoms in people with influenza-like illness symptoms (i.e. ITT population) was a median of 160 hours (range 125 to 192 hours) in the placebo groups and oseltamivir shortened this by around 21 hours (95%CI −29.5 to −12.9 hours, p < 0.001; five studies). There was no evidence of effect on hospitalisations based on seven studies with a median placebo group event rate of 0.84% (range 0% to 11%): (odds ratio [OR] 0.95, 95% CI 0.57 to 1.61, p = 0.86).
How does this relate to the rest of the evidence?
A Health Technology Assessment (HTA) of antiviral drugs in influenza was published in December 2009. Like the 2009 Cochrane review, it did not use the Kaiser et al paper per se, but did include some data from RCTs of oseltamivir not published in full. Like the 2009 Cochrane review the HTA also found that on data judged appropriate for inclusion in the analysis a statistically significant reduction in complications in otherwise healthy adults was absent. However, oseltamivir did reduce the use of antibiotics both in otherwise healthy adults with flu-like illness (OR 0.37, 95%CI 0.29 to 0.48, p < 0.00001) and otherwise healthy adults with confirmed flu infection (OR 0.52, 95%CI 0.27 to 1.00, p = 0.05).
In a review (published in June 2011) of a systematic review of published and unpublished data, the European Centre for Disease Prevention and Control commented ‘The important result here is the confirmation of the original Kaiser meta-analysis result namely that early oseltamivir use was associated with reductions in the risk of some complications of influenza in previously well adults and adolescents in the period 1997-2001.’ The Kaiser et al study was also quoted in support of a beneficial effect of oseltamivir on complications in a review in January 2011 by the US Centers for Disease Control and Prevention (CDC).
The UK summary of product characteristics for oseltamivir states that in a pooled analysis of 2413 influenza-positive adults and adolescents 13 years of age and older the proportion of subjects who developed specified lower respiratory tract complications (mainly bronchitis) treated with antibiotics was reduced from 12.7 % (135/1063) in the placebo group to 8.6 % (116/1350) in the oseltamivir treated population (p = 0.0012). These exact denominators appear in the Kaiser 2003 meta-analysis. However, the US-approved prescribing information for oseltamivir states that oseltamivir has not been shown to prevent serious bacterial infection complications of influenza.
So what?
Health professionals considering the use of oseltamivir (or amantadine or zanamivir) in managing seasonal influenza should follow NICE guidance. This recommends oseltamivir or zanamivir for the treatment of influenza in ‘at risk’ adults and children (as defined in the guidance) if national surveillance schemes indicate that influenza virus A or B is circulating and the person can start treatment promptly after the onset of symptoms (as in the drugs’ licensed indications). This guidance does not cover the circumstances of a pandemic, impending pandemic, or a widespread epidemic of a new strain of influenza to which there is little or no community resistance
There seems to be no question over whether or not oseltamivir shortens the duration of seasonal influenza symptoms by about a day on average. However, the issue over whether it has an effect on complications is unresolved. This is an ongoing controversy and as and when new important information emerges this MeReC Rapid Review will be updated.
It is important to note that ‘no statistically significant difference’ is not necessarily the same as ‘no difference’. Strictly, a finding that the result is not statistically significant means that insufficient data have been provided to reject the null hypothesis (that there is no difference between the two treatments). From the data discussed by the Cochrane reviewers, it might be that there truly is no difference between oseltamivir and placebo in preventing complications (and observed differences are due only to chance), or it may be that oseltamivir truly has an effect but the studies considered lacked sufficient statistical power to detect this difference ‘beyond reasonable doubt’. The virulence of the seasonal influenza viruses circulating at the time oseltamivir was being examined in clinical trials may have a bearing on this: if this was relatively mild, and the risk of complications consequently low, it would have been harder to detect a benefit from oseltamivir, if one truly existed, to a statistically significant extent. During the 2009 swine flu pandemic, the Chief Medical Officer noted that international observational evidence supported early antiviral treatment in swine flu. This evidence was summarised in an editorial from the US CDC, published in December 2009.
Further information on influenza and its treatment can be found on NHS Evidence
Please comment on this rapid review using our feedback form.
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you.