Background: Recent measles outbreaks in highly vaccinated populations have highlighted the role of vaccine failure as a barrier to the elimination of measles. We sought to estimate the rate of secondary failure (clinical measles after vaccine-induced seroconversion) of measles vaccines using metaanalysis.
Methods: We identified 1411 studies of which 125 were relevant. From these we found 10 original studies of healthy subjects with sufficient details to calculate a pooled secondary failure rate. We performed a test for homogeneity before any pooling.
Results: Although significant heterogeneity prevented their pooling as a single group, the studies fell into three homogeneous groups suitable for pooling. Group A studies used killed vaccine whereas the other two groups (Groups B and C) of studies used live vaccine. These latter groups differ in that the studies in Group B share higher failure rates and are difficult to interpret with respect to the lack of verification of vaccination, immunization before 12 months of age and a non-North American study site and vaccine manufacturer. Those studies in Group C, in which US subjects were older than 12 months at vaccination and received a live US-manufactured vaccine that was documented in a medical record, had a failure rate of 0 of 2031 with a 95% confidence interval of 0.0 to 0.147%.
Conclusions: Although reports of measles related to secondary failure exist, studies that permit the calculation of the rate of secondary failure demonstrate that the rate appears to be < 0.2%.