In TA 168 (Feb 2009), NICE recommends the use of oseltamivir and zanamivir for the treatment of influenza in adults and children if all the following circumstances apply:
• national surveillance schemes indicate that influenza virus A or B is circulating• the person is in an 'at-risk' group (as defined) • the person presents with an influenza-like illness and can start treatment within 48 hours (or within 36 hours for zanamivir treatment in children) of the onset of symptoms as per licensed indications.
When surveillance data indicate that there is a substantial likelihood that people presenting with an influenza-like illness are infected with an influenza virus, GPs are able to prescribe antivirals according to this guidance (for treatment or prophylaxis), at NHS expense.
The effectiveness of oseltamivir has however been the subject of much debate. The Cochrane group published a meta-analysis on this topic in 2014 – this concluded that oseltamivir had modest benefit but documented problems obtaining original data from Roche. The current meta-analysis included more complete data (all available data from nine published and unpublished placebo-controlled trials RCTs; n=4328) and was based on individual patient data. Although the analysis was funded by an unrestricted grant from Roche, it was conducted by an independent research group.
In the intention-to-treat (ITT) patients who were found to be infected with the influenza virus (ITT infected; 66% of the oseltamivir group and 68% of the placebo group), oseltamivir was associated with a median reduction of 25.2 hours in the time to alleviation of all symptoms. In the entire ITT group (including those in whom influenza was not confirmed), the benefit was smaller but remained statistically significant (difference −17.8 hours).
Two indicators of complications were also assessed. The first was lower respiratory tract complication >48 h after randomisation requiring antibiotics – this was reduced with oseltamivir; with an absolute risk difference of −3.8% for the ITT-infected patients and −3.0% for the ITT population. This was not a pre-defined primary outcome for this study however and was non-specific, as there were no trial criteria for the prescription of antibiotics and no specific diagnostic tests were required to confirm a bacterial infection. The second was admittance to hospital for any cause – this again was lower with oseltamivir, with absolute risk difference of −1.1% in the ITT-infected group; the difference in the ITT population was not statistically significant. This outcome was also non-specific, as there were few admissions to hospital (nine in the oseltamivir group and 22 in the placebo group) and the causes were varied and not specified.
The risk of nausea (difference of 3.7%) and vomiting (4.7%) was increased with oseltamivir and the researchers say that careful consideration should be given as to whether the magnitude of the observed benefits outweigh these harms. As there was no benefit observed for oseltamivir in patients who were not infected with influenza, the balance of benefits and harms becomes less favourable if more non-infected participants are treated with oseltamivir. Treatment strategies need to avoid the treatment of patients who are not infected to avoid these harms. The authors of a related Comment article note that a high proportion of the study population had diagnosed influenza infection, and that this is unlikely to be seen outside the trial environment.
The researchers note that a further individual patient data meta-analysis of three Roche-sponsored randomised trials plus two other randomised trials in children is underway and will be published separately.