A vaccine is assessed by the reduction of the risk of disease that is produced by vaccination, the vaccine's efficacy. In contrast, in the field, the effectiveness of a vaccine is the practical reduction in risk for an individual when they are vaccinated under real-world conditions. Measuring efficacy of influenza vaccines is relatively simple, as the immune response produced by the vaccine can be assessed in animal models, or the amount of antibody produced in vaccinated people can be measured, or most rigorously, by immunising adult volunteers and then challenging with virulent influenza virus. In studies such as these, influenza vaccines showed high efficacy and produced a protective immune response. For ethical reasons, such challenge studies cannot be performed in the population most at risk from influenza – the elderly and young children. However, studies on the effectiveness of flu vaccines in the real world are uniquely difficult. The vaccine may not be matched to the viruses in circulation that year; virus prevalence varies widely between years, and influenza is often confused with other influenza-like illnesses.
Nevertheless, multiple clinical trials of both live and inactivated influenza vaccines against seasonal influenza have been performed and their results pooled and analyzed in several recent meta-analyses. Studies on live vaccines have very limited data, but these preparations may be more effective than inactivated vaccines. The meta-analyses examined the efficacy and effectiveness of inactivated vaccines against seasonal influenza in adults, children, and the elderly. In adults, vaccines show high efficacy against the targeted strains, but low effectiveness overall, so the benefits of vaccination are small, with a one-quarter reduction in risk of contracting influenza but no significant effect on the rate of hospitalization. However, the risk of serious complications from influenza is small in adults, so unless the effect from vaccination is large it might not have been detected. In children, vaccines again showed high efficacy, but low effectiveness in preventing "flu-like illness". In children under two the data are extremely limited, but vaccination appeared to confer no measurable benefit. In the elderly, vaccination does not reduce the frequency of influenza, but seems to reduce pneumonia, hospital admission and deaths from influenza or pneumonia.
Overall, the benefit of influenza vaccination is clear in the elderly and vaccination of children may be beneficial. Routine vaccination of adults is not predicted to produce significant improvements in public health. The apparent contradiction between vaccines with high efficacy, but low effectiveness, may reflect the difficulty in diagnosing influenza under clinical conditions and the large number of strains circulating in the population. In contrast, during an influenza pandemic, where a single strain of virus is responsible for illnesses, an effective vaccine could produce a large decrease in the number of cases and be highly effective in controlling an epidemic. However, such a vaccine would have to be produced and distributed rapidly to have maximum effect.