The death of an apparently healthy six-year-old child has helped to put swine flu back on the front pages. The Health Protection Agency says that 5 to 14-year-olds remain the group predominately affected by the illness — 1 in 600 of them — 1,500 a day — went to their GPs with symptoms last week. The death of a child naturally provokes our shock and sympathy, and such events are now so rare that they are unfamiliar to most people. Statistically, 7 is the safest age to be in England and Wales — there are 650,000 seven-year-olds and about 60 die each year. That’s 1 in 10,000.

So what do young people die of, and what might swine flu do to those risks? The Office for National Statistics reports that of 6.3 million children aged between 5 and 14 in England and Wales, 721 died in 2007. The statistics rather coldly amalgamate 721 individual stories, each of which will be gone over endlessly by parents and others who were touched by their short lives.

We tend to hear about the 135 accidental deaths — including 34 pedestrian fatalities, 18 killed on bicycles, 5 on motorbikes, 22 in cars, 2 in trains, 1 who fell from a tree, 8 drownings, 1 electrocution, 3 deaths in fires and 6 accidental poisonings. These events are so rare that it is unsurprising when they make local or even national news. We also hear about the 24 homicides in this age group — about half of them committed by the child’s parents — though we hear less of the 16 suicides.

But most of the deaths are from natural causes, and although we might see photos of plucky children fighting disease, when they actually die families are usually left to grieve in private. Many of these deaths are from genetic or pre-existing conditions: 170 were classified as cancer, 15 as asthma and 15 as epilepsy. Deaths from simple infections are very uncommon: these included 2 from chicken pox, and the deaths of only 2 children aged between 5 and 14 were primarily classified as “influenza” in 2007, and they may have had other health problems.

A brief look in any churchyard shows that a child’s death was not always so rare. When Dickens published Oliver Twist in 1838, those lucky enough to reach 5 years old (a quarter of those born didn’t) had a 1 in 10 chance of dying before they were 15. That’s about 100 times the risk that children face today.

So will swine flu put us back to Dickensian levels of risk? It is difficult to estimate how many children might die if no vaccine were available: there have been about 500 deaths out of 100,000 confirmed swine flu cases worldwide, a death rate of 1 in 200 death rate.

But the main UK research team recently warned that the only thing we can be confident about this figure is that it is wrong. The true denominator will be many more than 100,000, as only the most severe cases reach hospital. But the numerator will also be wrong as many deaths will have been attributed to a different cause.

If you also take into account changes in both the treatment and severity of the disease, the job of assessing the odds on any particular case dying become a matter of judgment. It would help if there was decent data on what was happening to ordinary families who did not necessarily go to their GPs or even call NHS Direct, and we might ask why plans are not in place to collect such data on a proper sample — if we can collect TV viewing figures from a representative subset, why can’t we do the same for illness?

As this is not an academic article I can make some fairly conservative assumptions of what we might face if a vaccine is not made available. Suppose that we are hearing only of a fraction of cases, and that rather than 1 in 200, only 1 in 1,000 of those infected die. Suppose that only 1 in 10 seven-year-olds becomes infected in the next year, well below current projections. This would mean about 65,000 cases, of which 1 in 1,000, or 65 seven-year-olds, would die.

That would double the number of seven-year-olds who died in this country. So even by my cautious assumptions. child mortality would return to the level of 25 years ago. A seven-year-old boy would face the same risk that a 15-year-old now faces. It is hardly surprising that doctors are drawing up emergency plans to prevent intensive care being overwhelmed, and hospitals are wondering what to do when staff cannot or will not come to work.

But this gloomy projection assumes that we do nothing. In fact, the impressive effort to produce a vaccine is yielding results and the Department of Health plans to have every child between 3 and 16 vaccinated against the current strain of swine flu by Christmas. It will be interesting to see if this is accompanied by accusations that the vaccine harms children — in which case objectors might be asked if they really prefer the alternative. However no vaccine is perfect, and there is always the risk that an important mutation will occur. For example, there is evidence that the seasonal H1N1 flu virus is already becoming resistant to Tamiflu.

We have heard recently about 18-year-old soldiers killed in Afghanistan and a 19-year-old who survived 12 days in the Australian Bush, reflecting both the vulnerability and the resilience of teenagers. Gordon Brown has asked us to be ready for “difficult times”, and we may need to brace ourselves for more deaths of young people from both war and pestilence.

David Spiegelhalter is Winton Professor of the Public Understanding of Risk at Cambridge University

Times Recommends

• Farwell to the stiff upper lip
• It’s about who runs the US, not health
• The Iron Lady and her blue felt-tip sympathy