Fever Under Age 3

What's the best thing to do when a healthy-appearing child who is 3 to 36 months old has a fever of 102.2 or higher, with no clear source of the fever? Is it best just to treat them all with antibiotics, to be safe? Or to do blood tests on all of them, and just treat the ones with suspicious results? Or just to watch and see if they are getting better or worse, before testing or treating? Guidelines published a dozen years ago suggested just giving all of these kids antibiotics was best, because up to 11% of them would turn out to have dangerous bacteria in their bloodstream. Because bacterial infections were so common then, that was the strategy that had the lowest mortality rate. But the world has changed since 1993. Kids are now routinely vaccinated against the two most common bacterial culprits (pneumococcus and Hib (haemophilus influenzae type b)). What's the safest strategy today? And what would you prefer? A study in the April 2006 Pediatrics looks at 3 different options and at parents' preferences. The study assumes that a urine infection has been ruled out in all of the children. Here is what the authors report:

First, some good news. In 2006, the chance of death from this type of fever is very, very low.

For otherwise healthy kids with fevers who have not been vaccinated, the safest strategy is still just to treat with antibiotics. Mortality risk is about 4 times higher if you just observe than if you treat right away. With just observing, the risk of death is still very low - about 1 in 18,800. (To put this in perspective, it is only a little more dangerous than traveling by car. The annual risk of auto death in the U.S. at this age is about 1 in 22,000).

For otherwise healthy kids today who have received these vaccines, the risk of dying is much lower, about 1 in 100,000 more or less, no matter which fever strategy is chosen. The risk of death with going straight to antibiotics is about 1 in 93,000; the risk with just waiting and seeing with no tests or treatment is even lower, at about 1 in 172,000.

The authors argue, and I agree, that parents' preferences are key in deciding what to do. Different parents have different risks tolerances. Some would rather use antibiotics than subject their child to a needle poke; for others, the momentary pain of the needle is preferable to the antibiotic. Some parents are comfortable with a 1 in 10,000 risk; others aren't comfortable until the risk drops to 1 in 100,000 or less. Some are comfortable just observing; others feel a need to know what is going on. I strongly believe that families' preferences are important, and need to be considered.

When presented with various risks for unvaccinated kids, about 84% of parents prefer to treat and about 16% prefer just to observe. For vaccinated kids, about 50% prefer observing, 42% prefer testing, and only 8% would opt to just treat. Even when given the exact same risk numbers, in a simplified scheme of possible outcomes, no single strategy was everybody's favorite. What would you prefer if it were your child with the fever?

Alan Greene MD FAAP