Editorial Commentary: Influenza Vaccine: Glass Half Full or Half Empty?

Abstract

In this issue of Clinical Infectious Diseases, Ohmit and colleagues report the results of an assessment of influenza vaccine effectiveness during the 2010–2011 influenza season, using a nonrandomized, prospectively followed cohort [1]. This intriguing study had 2 surprising findings. Remarkably, the rate of polymerase chain reaction–documented influenza illness was not substantially different between those subjects whoreceivedinfluenzavaccine and those who did not, and the study was unable to demonstrate statistically significant vaccine effectiveness even after adjustment for age and the presence of high-risk conditions. The second surprising finding is that receipt of influenza vaccination in a previous season may have impacted the effectiveness of the vaccine in the current season. The population under study was predominantly young healthy adults and children with a low frequency of chronic health conditions, the types of subjects that typically mount the best immune responses to vaccine. In addition, there was a close antigenic match between the vaccine and circulating viruses. As we are currently struggling through one of the most vigorous influenza seasons in recent memory, the apparent failure of influenza vaccine under optimal conditions seen in this study is indeed troubling. Influenza vaccines are unique among currently licensed vaccines in that the continuous antigenic evolution of the virus requires frequent updating, reformulation, and annual administration of the vaccine. Thus, each year’ si nfluenza season essentially represents a test of a new vaccine against a new virus. Partly for this reason, there has been substantial interest for developing systems to monitor the performance of influenza vaccines on an annual basis. Prospective, randomized controlled trials are the method that is least susceptible to bias as the assignment of vaccinated or unvaccinated groups is made randomly. Organizing such trials on an annual basis would be quite challenging, however. Alternatively, several large networks have been developed that monitor vaccine effectiveness using a test-negative casecontrol design. In these studies, selected persons seeking care for acute respiratory illness are tested using a sensitive and specific test for influenza, and the vaccination histories of those testing positive are compared to those testing negative. The selection of cases based on a prespecified case definition and diagnostic testing without knowledge of vaccine history may eliminate some of the biases inherentinotherobservationalstudies[2], but the assessment is typically limited to medically attended outcomes. Because the majority of influenza infections do not result in a medical visit, this provides a somewhat incomplete assessment of the impact of vaccination. The innovative prospective surveil