ArticlesEfficacy and effectiveness of influenza vaccines: a systematic review and meta-analysis
Introduction
The main strategy for prevention and control of seasonal and pandemic influenza for the past 60 years has been vaccination.1, 2 The first population-scale use of an inactivated influenza vaccine was in US military personnel in 1945.3 In 1960, the US Surgeon General, in response to substantial morbidity and mortality during the 1957–58 pandemic, recommended annual influenza vaccination for individuals with chronic debilitating disease, people aged 65 years or older, and pregnant women.4 This recommendation was made without data for vaccine efficacy or effectiveness for these high-risk populations. Instead, it was made on the basis of studies showing efficacy in young, healthy military recruits with clinical illness or seroconversion as primary measures of infection. In 1964, the Advisory Committee on Immunization Practices (ACIP) reaffirmed this recommendation but noted the absence of efficacy data.5 Because of the longstanding public health recommendation of annual vaccination in the elderly and other high-risk groups, such patients have been excluded from placebo-controlled randomised clinical trials in the USA for the past 50 years. The ACIP supports the widely held view that inclusion of individuals at high-risk of influenza in placebo-controlled trials would be unethical.2
In 2010, the ACIP established the first recommendation of national universal seasonal influenza vaccination.2 Vaccination every year is now recommended with trivalent inactivated vaccine (TIV) for all individuals aged 6 months or older, or live attenuated influenza vaccine (LAIV) for healthy non-pregnant people aged 2–49 years.2 In the USA, TIV has been used since 1978 and accounts for approximately 90% of influenza vaccine given at present.6 The LAIV was first approved for use in the USA in 2003 and accounts for approximately 9% of the vaccine given.7, 8 The universal influenza vaccination recommendation came after a decade of incremental changes during which the ACIP expanded recommendations to include an ever-increasing proportion of the US population.
Previous meta-analyses of TIV or LAIV efficacy and effectiveness have included studies that used diagnostic endpoints with little sensitivity or specificity to confirm influenza infection in recipients of vaccine and placebo.9, 10, 11, 12 For example, the use of serology to confirm influenza infection in participants who were vaccinated with an inactivated vaccine had been recognised as problematic since the 1940s and 1950s.13, 14, 15, 16 Investigators noted that the increased antibody titres after vaccination in individuals given an inactivated vaccine made it difficult to document a four-fold rise in hemagglutinin antibodies necessary to confirm an influenza infection. Studies into the efficacy and effectiveness of TIV continue to use serology as a primary endpoint for confirmation of influenza infection in study participants, without addressing concerns raised by the studies done in the 1940s and 1950s. Petrie and colleagues17 showed that, in participants who had received TIV, only 23% who had RT-PCR-confirmed H3N2 influenza had serological evidence of infection. By contrast, 90% of cases confirmed by RT-PCR in the placebo group had serologically confirmed infection. This biased case detection contributes to overestimation of the effect of vaccines in studies of TIV that rely on serological confirmation of influenza infection.
To assess the highest quality evidence about the efficacy and effectiveness of licensed influenza vaccines in the USA, we did a meta-analysis of randomised controlled trials and observational studies that used RT-PCR or viral culture to confirm influenza infections.
Section snippets
Definitions and outcomes
We defined influenza vaccine efficacy as the relative reduction in influenza risk after vaccination as established by a randomised placebo-controlled clinical trial. We defined influenza vaccine effectiveness as relative reduction in influenza risk in vaccinated individuals in observational studies that used medically attended, laboratory-confirmed influenza as the primary outcome of interest.18 Observational study designs included case-control (with test-negative controls), case-cohort, and
Results
We identified 5707 studies on influenza vaccines in human beings with our PubMed search (figure 1). Of these, 992 were identified as cohort studies, case-control studies, clinical trials, randomised controlled trials, or did not have MeSH terms. A review of the abstracts of these studies suggested 176 (18%) potentially eligible studies; 73 (41%) were randomised controlled trials estimating vaccine efficacy and 103 (59%) were observational studies estimating vaccine effectiveness. 31 of these
Discussion
Our analysis differs from previous reviews of influenza vaccine efficacy and effectiveness because of our use of restrictive study inclusion criteria to minimise bias and confounding. Our approach uses only very specific outcome endpoint data for virologically confirmed influenza. When these more stringent criteria were applied, we noted substantial gaps in the evidence base for some age groups with regard to efficacy data for TIV and LAIV.
There are no randomised controlled trials showing
References (71)
- et al.
A meta-analysis of effectiveness of influenza vaccine in persons aged 65 years and over living in the community
Vaccine
(2002) - et al.
Efficacy of vaccination with live attenuated, cold-adapted, trivalent, intranasal influenza virus vaccine against a variant (A/Sydney) not contained in the vaccine
J Pediatr
(2000) - et al.
Influenza vaccine concurrently administered with a combination measles, mumps, and rubella vaccine to young children
Vaccine
(2010) - et al.
Efficacy and safety of a live attenuated influenza vaccine in adults 60 years of age and older
Vaccine
(2009) - et al.
Effectiveness of inactivated influenza vaccine in children aged 9 months to 3 years: an observational cohort study
Lancet Infect Dis
(2011) - et al.
Estimating the influenza vaccine effectiveness in elderly on a yearly basis using the Spanish influenza surveillance network—pilot case-control studies using different control groups, 2008–2009 season, Spain
Vaccine
(2010) - et al.
Effectiveness of seasonal 2008–2009, 2009–2010 and pandemic vaccines, to prevent influenza hospitalizations during the autumn 2009 influenza pandemic wave in Castellón, Spain. A test-negative, hospital-based, case-control study
Vaccine
(2010) - et al.
Influenza vaccination and all-cause mortality in community-dwelling elderly in Ontario, Canada, a cohort study
Vaccine
(2010) - et al.
Mortality benefits of influenza vaccination in elderly people: an ongoing controversy
Lancet Infect Dis
(2007) - et al.
A nested case-control study of influenza vaccination was a cost-effective alternative to a full cohort analysis
J Clin Epidemiol
(2004)