Closing schools has proved to be one of the most effective ways of curbing the spread of the H1N1 influenza virus - also known as swine flu - in the US. But a study has found that shutting schools for less than two weeks during an epidemic does not reduce infection rates. Instead, researchers suggest they should be closed for at least eight weeks.

Last year, the federal government urged schools to close only as a last resort, but around 1,000 still shut during the spring and autumn outbreaks.

The new study was based on a computer simulation model of Allegheny County, Pennsylvania, built from data on the county’s population, school systems, hospitals, workplaces, households and communities. Differing scenarios were entered into the programme, then the simulation modelled the movements of residents between their homes, schools and workplaces, and how the influenza would spread.

According to the researchers, the study showed that closing schools for short periods of time might increase infection rates by returning susceptible pupils in the middle of an epidemic when they were most vulnerable.

The report, published this month in the Journal of Public Health Management and Practice, details work done by the University of Pittsburgh, the Allegheny County Health Department, and RTI International, a non-profit research organisation based in North Carolina.

“Since children are more susceptible to most influenza strains than adults, closing schools seems an obvious strategy to slow the spread of flu,” said the study co-author Philip C Cooley, assistant director of bioinformatics and RTI fellow.

“Computer simulations indicate that such closures are ineffective unless sustained for at least eight weeks after implementation.”

Cooley added that closing schools quickly at the start of an outbreak was less effective than keeping them closed continually throughout the epidemic.

The model was developed to compare three school-focused strategies for dealing with H1N1 outbreaks: isolating sick children at home but leaving the school open; closing the entire school; closing individual schools.

The study found that identifying sick children and keeping them at home had minimal impact. Additionally, there were no significant differences between individual closures and system-wide closures in mitigating an epidemic.

RTI’s Geospatial Spatial and Technology Group developed complex synthesised populations of 1.2 million people, including 200,000 school-aged children, 500,000 households and nearly 300 schools. The study is part of the Models of Infectious Disease Agent Study (MIDAS) funded by the National Institutes of Health to simulate the spread of infectious diseases and advise health officials.

Its findings follow UK research, published last year in the Lancet Infectious Diseases journal, which found that pre-emptive school closures could help reduce infections at the peak of an outbreak, but only if children were kept apart outside school.