Historically, measles incidence has shown clear seasonal patterns driven by the school calendar, but since the start of mass vaccination in developed countries there are only occasional outbreaks, which may have changed the effect of school vacations on transmission. In 2013–2014, a large measles epidemic took place in a low vaccination coverage area in The Netherlands, allowing us to quantify current-day measles transmission and the effect of school vacations.
We fitted a dynamic transmission model to notification and hospitalization time series data of the Dutch 2013–2014 measles epidemic. Our primary aim was to estimate the reduction in contact rate during school vacations and the number of cases averted due to the vacation. In addition, because the summer vacations were time-staggered in three regions, we could distinguish within-region from across-region effects of school vacations.
We estimated a 53% (95% credible interval: 45%, 60%) reduction in contact rate during school vacations, resulting in 4900 (3400–7100) averted cases (estimated outbreak size: 16,600 [12,600–23,200]). There was a shift from mainly local transmission during school term to mainly cross-regional transmission during vacations. With seroprevalence data, we derived a current-day estimate of 15 to 27 for R 0 (number of secondary cases per primary case in a susceptible population).
School vacations are associated with greatly reduced overall measles transmission. However, transmission is not eliminated, and increased long-distance travel may even promote spread to other areas. Therefore, we estimate that school closure is unlikely to prevent measles epidemics unless there are still few cases and the community is well vaccinated.
From the aCentre for Infectious Disease Control, Netherlands Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands
bDepartment of Medical Statistics and Bio-Informatics, Leiden University Medical Center, Leiden, The Netherlands.
Submitted April 11, 2017; accepted April 3, 2018.
The final datasets used for model fitting (after preprocessing) and the computing code (R script and JAGS script) are available from the authors upon request. We thank staff at Municipal Health Services, laboratories, and clinicians for reporting and investigating cases. We thank Dr Liesbeth Mollema for help with the seroprevalence data and Dr Helma Ruijs for helping us understand the Orthodox Protestant Individual (OPI) community.
The authors report no conflicts of interest.
Supplemental digital content is available through direct URL citations in the HTML and PDF versions of this article (www.epidem.com).
Correspondence: Don Klinkenberg, Centre for Infectious Disease Control, Netherlands Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, The Netherlands. E-mail: email@example.com