This review assessed the benefits to the wider community of vaccinating healthy children against influenza. The authors concluded that vaccination appears to benefit both the children themselves and the wider community, both in terms of health benefits and cost-effectiveness. However, the poor quality of the evidence base indicates that further research is required before these findings can be confirmed.

To assess the effectiveness of vaccinating children against influenza in order to protect the wider community from influenza infection.

The Cochrane Controlled Trials Register, MEDLINE, EMBASE, CINAHL, NHS EED, HEED and DARE were searched up until January 2004; the search terms were reported. In addition, the authors searched specific internet sites, search engines, the reference lists of retrieved studies and registers of trials in order to locate further studies. Topic experts were contacted in an attempt to identify unpublished data. Although no language restrictions were applied, the authors were unable to translate six small studies which were only available in non-Roman script.

Study designs of evaluations included in the review

Comparative studies of any design were eligible for inclusion in the review.

Specific interventions included in the review

Studies that compared influenza vaccination with no influenza vaccination (placebo or another alternative vaccine) were eligible for inclusion. Seven of the included effectiveness studies compared influenza vaccination with placebo and three with no vaccination; one study compared influenza vaccination with hepatitis vaccination. The studies assessed both live and inactivated influenza vaccines which, in the majority of cases, were directed against both influenza A and B strains; the majority of the vaccines were trivalent. The vaccines were administered either intramuscularly or intranasally.

Participants included in the review

Studies of healthy children were eligible for inclusion in the review. Studies of children who were suffering from specific diseases such as asthma were excluded. The majority of the children assessed in the included studies were at school or daycare. The ages of the children varied, the youngest being 0.5 years and the oldest 19 years.

Outcomes assessed in the review

The studies must have assessed any measure of morbidity or mortality relating to influenza infection or pneumonia or any other respiratory infection in the contacts of the vaccinated children or the wider community. Measures of all-cause mortality and cost-effectiveness were also eligible for inclusion. The specific measures reported in the review included: influenza A/B infection in contacts confirmed by serology or culture; the incidence of respiratory tract infections (RTIs); days off school due to RTI; days off work due to RTI; physician visits; hospitalisations for RTI; prescriptions and over-the-counter medications for the treatment of RTI; all-cause and specific mortality; and the incidence of influenza-like illness.

How were decisions on the relevance of primary studies made?

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

The validity of the randomised controlled studies (RCTs) was assessed according to predefined criteria: random assignment; method of randomisation; confirmation of true randomisation; concealment of treatment allocation; blinding of contacts and outcome assessors; methods described for outcome assessment; adequacy of methods used to diagnose illness in contacts; baseline comparability of the treatment groups; equal treatment of the study groups; eligibility criteria reported; all appropriate data included in the analysis; loss to follow-up accounted for; power calculations performed; adequately powered outcomes for cluster of contacts; and clustering accounted for in analysis. The authors also commented on any other relevant methodological issues. Separate criteria were used to assess the validity of the cost-effectiveness studies (the details were reported).

Two authors independently assessed the validity of the studies.

Two authors independently extracted the data from the included studies. The study methods and designs used to assess outcomes for both the vaccinees and contacts, the influenza outbreak vaccine match and its adequacy, the follow-up period for contact outcomes, and the types of outcomes measured in both the vaccinees and contacts were described for effectiveness studies. Specific details pertinent to cost-effectiveness were recorded for the economic studies (the details were reported). For vaccine efficacy, the percentage efficacy with 95% confidence intervals (CIs) or p-values was reported where possible; other outcomes were reported as described by the original study authors.

How were the studies combined?

Given the heterogeneity, the studies were presented as a narrative synthesis with critical description and interpretation of the study data. The studies were grouped according to whether they addressed effectiveness or cost-effectiveness; effectiveness studies were discussed according to their effects on the incidence of influenza and influenza-like illness in household and school contacts, and their effects on the wider community.

How were differences between studies investigated?

Differences between the studies were evident from the data tables and were also discussed in the text.

Fourteen studies were included in the review: six RCTs (n=901 and 190 families), two observational subsets of RCTs (n=18,607), three non-randomised community studies (n over 3,159) and three economic studies (n not stated).

The quality of the RCTs included in the review was variable. Only one study described the method of randomisation used. Similarly, only one study reported an appropriate method to conceal treatment allocation. Other problems with the studies included a lack of information on blinding, the absence of a power calculation for cluster outcomes, and the lack of a statistical analysis that allowed for the effects of clustering.

Influenza and influenza-like illness in household or school contacts (6 RCTs).

Only two studies were primarily designed to assess the effects of vaccines in contacts. Two RCTs (n=190) showed no significant differences in influenza infection (confirmed by serology or culture) in family contacts of any age. Three RCTs (n=557) showed a statistically significant difference that suggested protection against respiratory infection in the family contacts of vaccines; efficacy was variously defined and ranged between 16 to 30% for RTS and 42% (p=0.04) for respiratory infections with fever. The highest quality study comparing intranasal inactivated virus with placebo in 127 children with recurrent RTI showed a significant protective effect in family contacts (infection rate 24%, p<0.0001) with fewer parental workdays lost (84%, p<0.0001) and fewer physician visits for respiratory illness (39%, p<0.0001).

Effects on the wider community (3 non-randomised studies).

One study suggested a protective effect of vaccination but had numerous methodological flaws. A similarly weak observational study suggested a fall in deaths from influenza and pneumonia, which corresponded to period of high uptake of influenza vaccination. The third non-randomised study reported a small effect in illness rate between intervention and control communities, which translated into a substantial absolute number of infections, but this appears to have had little effect on herd immunity.

The authors assessed the findings of three economic studies, including two cost studies and one cost-effectiveness study. All of the studies were carried out in the USA; two of the studies were carried out in pre-school children. All three studies concluded that the universal vaccination of children was cost-saving in settings where children could be vaccinated without the need for parents to miss work. Individual savings varied between $1.20 to $21.28 and $3.99 to $34.79 per vaccine for the cost studies, and was $29.67 per influenza-like illness for the cost-effectiveness study. These savings were driven largely by the reduction in work days missed, although vaccination was marginally cost-saving in one study if indirect costs to contacts were not included and it was still highly cost-effective in another study when productivity losses were not included.

The evidence presented suggests that the universal vaccination of children against influenza is not only of benefit to the children themselves but could also produce significant benefits to the community as a whole, both in terms of health benefits and cost-effectiveness. However, these findings remain inconclusive given the limitations in the methodology of the identified studies. Further research is required.

This review was based on a clearly defined research question with clear inclusion criteria. The authors searched a number of literature sources for both published and unpublished data, and no language restrictions were applied. Multiple reviewers were involved in both assessing the validity of the primary studies and extracting the data; this helps reduce bias and errors. However, the authors did not report how many reviewers were involved in the study selection process, so it was unclear whether similar steps were taken to avoid selection bias.

The authors' decision to adopt a narrative approach to synthesis was appropriate given the variations in study design, intervention, population and outcomes. Their synthesis considered the review findings in light of both study quality and heterogeneity, and was accompanied by extensive data tables which helped summarise the data; however, only two studies specifically addressed the review question. Given the data presented, the review findings appear valid and the recommendation for further research seems appropriate.

The authors stated that neither of the funding bodies influenced the methods of analysis used in the study.

Practice: The authors did not state any implications for practice.

Research: The authors stated that due to the paucity of good-quality data, further research is required to confirm their findings. Ideally, a large cluster randomised trial that includes adequate numbers of communities in the treatment and control arms and ensures high vaccine coverage is required. However, given that such a study would most likely need to be conducted over more and possibly even several influenza seasons, it may not be practically or financially viable. An alternative option would be to create a model with a European focus that looks at the transmission characteristics of the influenza virus using a range of vaccination scenarios. These should incorporate estimates of effectiveness, cost-effectiveness and adverse effects, whilst also considering issues regarding vaccine uptake and acceptability. The authors also stated that policy-makers require a clear evaluation of the ethical considerations and acceptability to parents of annual vaccinations for children.

European Scientific Working Group on Influenza; Health Protection Agency.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.