Tamiflu for children questioned

It has been widely reported that Tamiflu may be of little benefit when given to children, with several news sources suggesting that the antiviral drugs Tamiflu and Relenza rarely prevent complications in children with seasonal flu, but still carry side effects. The news is based on a review of research on seasonal flu, which found that the antivirals could shorten the duration of seasonal flu in children by up to a day and a half, but had “little or no effect on asthma flare-ups, ear infections or the likelihood of children needing antibiotics”. The review did not look at swine flu.

The authors of the review are reported to have said that these drugs “are unlikely to help” children who catch swine flu. However, they offer a more cautious interpretation within the research paper, saying that the effects of antivirals on the “incidence of serious complications, and on the current A/H1N1 influenza strain remain to be determined”.

The Department of Health responded, saying that, "whilst there is doubt about how swine flu affects children, we believe a safety-first approach of offering antivirals to everyone remains a sensible and responsible way forward. However, we will keep this policy under review as we learn more about the virus and its effects”.

The review itself systematically identified and summarised the available research on the effects of the antiviral drugs oseltamivir (Tamiflu) and zanamivir (Relenza) for preventing and treating seasonal flu in children. The findings of this review will contribute to the decisions doctors make when considering the use of antivirals in children with seasonal flu. However, it is important to note that all of the studies in this review looked at seasonal flu, rather than the current pandemic flu strain. As the authors themselves note, the effects of these drugs on the current pandemic flu strain are not known, and studies are under way to address this question.

Where did the story come from?

Matthew Shun-Shin and colleagues from the John Radcliffe Hospital and the Oxford University Department of Primary Health Care carried out this research. The study did not receive any specific grant from any funding agency. It was published in the peer-reviewed British Medical Journal.

What kind of scientific study was this?

This was a systematic review and meta-analysis looking at the effects of the antiviral drugs oseltamivir (Tamiflu) and zanamivir (Relenza) in the treatment of seasonal influenza in children and prevention of transmission to children in households.

The researchers identified randomised controlled trials (RCTs) of oseltamivir and zanamivir by systematically searching online databases of medical and scientific literature and registers of clinical trials, including those by the manufacturers of oseltamivir and zanamivir (Roche Pharmaceuticals and GlaxoSmithKline, respectively). The manufacturers provided information about studies that had not been published.

The researchers identified further appropriate studies by looking at the reference lists of relevant papers, NICE guidelines and UK Health Technology Assessments. The researchers included only those RCTs looking at the use of oseltamivir and zanamivir (which are both neuraminidase inhibitors) to treat or prevent influenza in children under the age of 12.

Two researchers assessed the quality of the papers identified through this systematic search. They looked at how the participants were allocated into their groups, whether this could be influenced by the study researchers, whether the participants and researchers were blinded to which group the participants belonged to and how the study researchers dealt with any missing data.

Two researchers independently extracted data from each study, and any disagreements between their extractions were resolved by discussion with a third researcher. The researchers looked at data on children with laboratory-confirmed influenza and data on children who displayed symptoms of influenza but were not tested for the virus.

The main outcomes of interest to the researchers were how long the children took to recover from influenza in treatment studies and how many children caught influenza in the studies looking at preventing onward transmission. Other outcomes examined were the effects on children with asthma, side effects of the treatment, time to recover from individual symptoms of the flu and time to return to school, daycare or other normal activity.

The researchers pooled the data from the relevant studies where appropriate, and used standard statistical methods to look for differences between the groups and differences between the results of the individual studies.

What were the results of the study?

The researchers identified seven RCTs that met their inclusion criteria. Four of the studies were looking at the use of oseltamivir or zanamivir for the treatment of flu in children and three were looking at the use of the drugs for prophylaxis, specifically, preventing influenza infections in people who had been exposed to the virus from an infected individual in their household.

Most of the studies enrolled children who had influenza symptoms but had not been confirmed as having influenza through laboratory tests by the time they started the study. Some of the studies tested the participants for the influenza virus after the study started.

Treatment studies
Two of the treatment studies tested inhaled zanamivir and the other two assessed oral oseltamivir. These studies included a total of 1,766 children with flu symptoms, and 70% of these children had been confirmed as having the influenza virus (mostly influenza A). Only one of the studies was judged to be of a high quality. Children in these studies were mostly between five and 12, but one study included children as young as one year old. The children’s outcomes were mostly followed for a period of 28 days, with one study only following them up for five days.

These treatment studies found that oseltamivir or zanamivir reduced the time it took for children to recover from the symptoms of flu by an average (median) of between 0.5 and 1.5 days. In two of these four trials these reductions were reported to be statistically significant in children whose flu infection was confirmed in the laboratory. These two trials (one on oseltamivir and one on zanamivir) found that:

Zanamivir reduced the time to recovery by 1.25 days in those with laboratory-confirmed flu (from 5.25 to 4.0 days).
Zanamivir reduced the time to recovery by 0.5 days in those with flu symptoms (from 5.0 to 4.5 days).
Oseltamivir reduced the time to recovery by 1.5 days in those with laboratory-confirmed flu (from 4.2 days to 2.6 days).

The other two trials also found reductions in time to recovery, but in one trial the reduction was not statistically significant, and in the other the statistical significance was not reported. The researchers did not pool these results because of inadequate reporting of trial data and differences between the studies.

Other findings were:

There was no difference in the proportion of children experiencing asthma exacerbations with neuraminidase treatment and control.
There was a small (6%) reduction in antibiotic use with neuraminidase treatment in children with laboratory confirmed flu when study results were pooled, but this reduction was not statistically significant.
Three trials found no difference in the proportion of children aged between five and 12 experiencing inflammation of the middle ear (otitis media). However, one of these trials found a 16% reduction in the proportion of children aged one to five experiencing otitis media (reduced from 31% to 15%).

Prevention studies
Two of the prevention studies tested inhaled zanamivir, and one assessed oral oseltamivir. These followed households that contained one person with flu, looking at how antivirals affected infection rates in other householders. These households were randomly assigned to receive either the active drug or control conditions (a placebo or no prophylaxis). These households included 863 children.

These three prevention studies treated the person who brought flu into the household in different ways: one study assigned them the same treatment method as the other people in their household (either zanamivir or placebo), another study gave them all oseltamivir and the third gave them no antiviral treatment. The three trials were each judged to be of moderate quality.

The results of the prevention studies were pooled and found that, overall, zanamivir and oseltamivir reduced the risk of developing confirmed symptomatic flu by 8% in households exposed to the flu. This reduction was statistically significant. This meant that 13 people would need to be treated to prevent one additional case of flu (with a confidence interval range of nine to 20 people).

Side effects
The treatment trials reported that 97% of children took eight or more of their 10 doses of zanamivir, and 90% took all 10 doses of oseltamivir. There was no significant difference in the proportion of children withdrawing from the drug and placebo group because of side effects.

Other findings were that:

Zanamivir did not increase risk of vomiting compared to control subjects in treatment studies.
Oseltamivir increased the proportion of children experiencing vomiting by 5%, which represents an extra one in twenty children experiencing vomiting. Among the untreated children with flu symptoms, 6.7% experienced vomiting.
The overall risk of nausea and diarrhoea was low (3.4% and 6.6%, respectively), and was not increased by neuraminidase inhibitor treatment.
No deaths were reported.

What interpretations did the researchers draw from these results?

The researchers conclude that oseltamivir and zanamivir “provide a small benefit by shortening the duration of illness in children with seasonal influenza and reducing household transmission”. They also say that, “their effects on the incidence of serious complications, and on the current A/H1N1 influenza strain remain to be determined”.

What does the NHS Knowledge Service make of this study?

This study has systematically identified and summarised the available research on the effects of the neuraminidase inhibitors oseltamivir and zanamivir for preventing and treating seasonal flu in children. The authors note that their study has some limitations, which include:

They considered most of the studies included to be only of moderate quality, which could affect the reliability of their findings.
Studies varied in the outcomes they assessed and their reporting of these, and this limited the authors’ ability to pool study results.
The studies included few children who had other underlying health conditions. Therefore, the effects in children with such conditions (for example, those with chronic heart conditions or who are immunosuppressed) are not clear.
The studies differed in the number of children that had been vaccinated against flu, and the outcomes of influenza could be less severe in those who were immunised.
The studies were not large enough to detect differences between groups in terms of the serious complications that can occur with flu, such as pneumonia.
The studies did not look at neuraminidase treatment in children under one. Therefore, the effects in this group are unclear. The authors note that oseltamivir is not licensed in the UK or US for use in children of this age.
Their search may have missed some relevant trials. They note that they did obtain details of one unpublished trial from the drug manufacturers, and that seven RCTs of neuraminidase inhibitors in children are currently under way. They also say that there are some studies looking specifically at the current pandemic strain of flu, and their outcomes could affect their review’s findings.

The findings of this review will contribute to the decisions doctors make when considering the balance of benefits and risks of neuraminidase inhibitor treatment in children with seasonal flu. However, it is important to note that none of the studies in this review looked at the current pandemic flu strain, rather they looked at seasonal flu.

As the authors themselves note, the effects of these drugs on the current pandemic flu strain are not known, and they report that studies are under way to address this question.

One of the reasons for using these drugs is that they may reduce the risk of serious complications from influenza. If the complication rate and severity of type A/H1N1 influenza differ from those of seasonal flu then it may not be possible to work out the balance of risks and benefits for swine flu based solely on studies done in patients with seasonal flu.