A model based on how flu is passed around in and out of school holidays has provided a disappointing prediction of how effective closing schools would be at curbing a potential pandemic. But public health experts emphasise that such closures would be implemented as part of a package of measures against the disease.

“Our work should temper the expectations of the scale of the reduction in overall illness and mortality achievable through this measure alone,” authors Simon Cauchemez and colleagues conclude in Nature this week.

In the absence of a vaccine against pandemic flu and a limited stock of antiviral medicines, health planners have proposed a range of ‘non-pharmaceutical interventions’ to try to limit the spread of a potential pandemic. One of the interventions most often suggested is the closure of schools, but this remains controversial, as it would probably take a heavy financial and social toll if it went on for long, and there is not much evidence to show that it would work.

“The current absence of quantitative data on the role of schools during influenza epidemics means there is little consensus on the probable effectiveness of school closure in reducing the impact of pandemics,” say Cauchemez, of the MRC Centre for Outbreak Analysis and Modelling at Imperial College London, and colleagues.

By looking at how flu transmission changed as kids changed their social mixing patterns during school holidays, the team hoped to be able to put some numbers onto how school closures affect  seasonal influenza epidemics and use them to model what might happen during a potential pandemic.

“In France, holidays are staggered across three geographic zones… and the timing varies from region to region and from year to year. This provides conditions resembling those of a natural experiment,” explain Cauchemez and colleagues.

The team used data collected routinely since 1984 by a network of GPs around the country (The Sentinel network). They used the data to build and test a model of how seasonal flu transmitted among the children and their household members. Then they expanded this to predict what might happen if schools were closed in a pandemic situation.

“Holidays prevent 16–18% of seasonal influenza cases (18–21% in children),” they report. “By extrapolation, we find that prolonged school closure might reduce the cumulative number of cases by 13–17% (18–23% in children) and peak attack rates by up to 39–45% (47–52% in children).”

"It is an excellent study and a great example of how modern epidemiology can inform public health planning,” comments Steven Riley, from the Department of Community Medicine, The University of Hong Kong. “However, this is a largely negative result — school closures will be less effective than has been assumed previously. This study does not support the use of school closures as a primary intervention during an influenza pandemic."

The conclusions do suggest that school closures might have less of an impact than other studies, says Eric Toner, Senior Associate with the Center for Biosecurity of the University of Pittsburgh Medical Center. But the impact is not insignificant, if closures were implemented as modelled (which would be problematic).

“The [model] assumptions seem reasonable, but it is important to understand that they are guesses. Extrapolating this work to a pandemic scenario requires numerous assumptions for which there is little evidence.”

Whether schools should be closed for long periods is a decision that requires a deep understanding of the individual community involved, Toner suggests. “Based on this work and others, school closures may be worth considering in some circumstances — such as a small rural community in which the kids can be cared for at home when they are out of school,” he says. “But I have serious reservations about the use of early and prolonged school closures in a large urban setting.”

Previous studies have shown that more needs to be done to slow a pandemic than simply closing schools, emphasises Robert Glass of the National Infrastructure Simulation and Analysis Center, Sandia National Laboratories, Albuquerque, New Mexico. Other ‘social-distancing measures’ will be required, he says.

These will probably depend on the transmission dynamics of the particular flu strain, as well as the communities involved. Analyses of the course of previous pandemics show that for a 1957-like pandemic closing schools and making kids and teenagers stay at home did a good job. But for a 1918-like pandemic, the social networks of adults would also have to be changed by extra social-distancing measures, changing working practices, for instance.

“I do not believe that anyone in the pandemic policy arena is advocating only closing schools in the face of a severe (high death rate) pandemic,” Glass says. “[The work] supports current pandemic policy. School closure must be combined with other interventions to form a robust mitigation strategy for community containment.”