Influenza cases show up several weeks earlier in preschoolers than in adults.

Current immunization policies recommend universal flu vaccination for children aged 6-23 months, but shots are advised for older children only if they have high-risk medical conditions. New data compiled by researchers at Children's Hospital Boston and Harvard Medical School, reported in October 1 issue of the American Journal of Epidemiology, suggest that otherwise healthy 3- and 4-year-olds drive flu epidemics, a pattern that may warrant consideration when formulating immunization policy.

The researchers leveraged a real-time computerized biosurveillance system linking five diverse health-care settings in Greater Boston, and examined medical visits from 2000 to 2004. Children aged 3 to 4 clearly led influenza epidemics, presenting with flu-like respiratory illness as early as late September. Children aged 0-2 began arriving a week or two later, while older children first arrived in October and adults began arriving only in November.

Moreover, flu-like illness in children under age 5, compared with all other age groups, was the most predictive of pneumonia and influenza deaths in the general population as determined from a Centers for Disease Control and Prevention database. Visits by children aged 0-2 provided the best prediction of mortality, but those of 3- and 4-year-olds followed close behind, suggesting that preschoolers, not just infants and toddlers, are important spreaders of flu to vulnerable groups.

"The data make sense because preschools and daycares, with their close quarters, are hotbeds of infection," says Dr. John Brownstein, the paper's lead author and a faculty member of the Children's Hospital Informatics Program at the Harvard-MIT Health Sciences and Technology program. "The data suggest that when kids are sneezing, the elderly begin to die. Three- and 4-year-olds are sentinels that allow us to focus our surveillance systems."

Influenza kills tens of thousands of Americans each year. Previous studies have shown decreases in household flu transmission and in adult flu mortality when children are immunized. Additional studies have also suggested that preschoolers drive flu epidemics, but they are based on simulations.

"Our study was not a simulation," says senior investigator Dr. Kenneth Mandl, an attending physician in Children's Department of Emergency Medicine and an informatics program faculty member. "This was real life."

With that previous report in mind consider this next report. H5N1 bird flu is currently very lethal to children.

The disease has been particularly deadly for children. In Thailand, 89% of patients under the age of 15 years died an average of nine or 10 days after illness onset.

Protection of children is the biggest problem in a pandemic. During a pandemic of high lethality schools will be closed. But lots of kids are in day care and kids come into contact with each other while playing. Also, with so many working mothers kids end up under the care of others during the day. This sets up transmission belts of influenza from home to children to day care environments to other kids and then to other homes.

Some ideas to protect children from a deadly flu pandemic:

• Keep all boarding schools open but quarantined. Staff and teachers should live on the premises and agree that if they leave they can't come back. Only occasional delivery trucks should enter the grounds to drop off large crates of supplies which then get collected by staff after the delivery trucks leave.
• Some summer camps for children should get converted into year-round boarding schools with the same quarantine rules as those put in place for boarding schools.
• Child labor laws should get temporarily lifted so that adolescents can join the workforce to work and live at quarantined workplaces (which I call "workplace cocoons").
• Farm houses out in plains states should take on children as boarders to live in isolation. The farmers would have to agree to avoid all contact with outsiders except for rare carefully made large supply deliveries.
• Where economically feasible mothers could stay home and home school their children. If they have spouses the spouses could stay somewhere else in order to avoid bringing infection into the home.
• Turn abandoned towns (such places exist in the US plains states) into quarantined camps for children.
• Take hotels and turn them into quarantined boarding schools.
• Poorly behaving children should be put in separate facilities to reduce quarantine rule breaking and skin contact when kids beat on other kids.
• Identify isolated islands (e.g. Catalina) and valleys where entry and exit could be controlled and where children could move to stay in homes and in camps set up over larger areas. This would allow much larger areas in which kids could roam, go to school, and function in more normal fashion.

To move children or adults into longer term quarantine facilities requires intermediate quarantine facilities where people stay even more isolated for a week or two to establish that they do not already have a flu infection. So one problem is how to process large numbers of children through about 10 day cycles to separate out the infected from the infection-free. Then the infection-free could be transported under quarantine to quarantined longer term living facilities of the sorts listed above.

Children do not work in the economy. Therefore their isolation from society should be easier to manage than the isolation of working adults. At the same time, children are the biggest transmission belts of influenza. So their isolation would do more to reduce the spread of flu virus than the isolation of most other members of society. Plus, they would be at great risk of dying from an H5N1 pandemic.