M. Eichner, H.H. Diebner, C. Schubert, H.W. Kreth and K. Dietz:
Estimation of the time-dependent vaccine efficacy from a measles epidemic.
Statistics in Medicine 21, 2355-2368 (2002).
Abstract:
During the last years there has been a resurgence of measles in
industrialised nations. Vaccine efficacy and duration of immunity have,
therefore, been questioned. We studied a measles epidemic that occured
in the municipality of Ansbach, Germany, from June 1992 to June 1993.
All general practitioneers and pediatricians of this region were asked
to fill out forms concerning measles cases they had seen. According to
this survey, 530 children had developed measles (89 were vaccinated and
441 were unvaccinated). The data reveiled a large fraction of
vaccinated children among the cases, particularly among the older ones.
The data were complemented by results of school surveys which gave us
some idea of the general vaccination coverage and allowed us to
estimate the age-dependend vaccination rate. Using a maximum likelihood
approach, we estimated the time-dependend efficacy of the measles
vaccine used in the Ansbach region. According to our results, the
measles vaccine hardly provided any protection for those who were
vaccinated before 1978 (estimate: 0%; 95% confidence interval: 0-47%).
In the following years, the vaccine efficacy increased to 80 % and
reached 96 to 97% (95% CI: 93-99%). The estimate for the time after
7/1990 yielded a lower value of only 89% which was based on few data
and, therefore, had a wide confidence interval (95% CI: 74-97%). An
extended model which allowed the vaccine protection to wane over time,
yielded an estimate of the loss rate of 0% per year (95% CI: 0-0,3 %
per year).
For the full paper see:
ExploSys - Research