Kids and Health Care: Fevers

The question of how to treat a young infant with a high fever has been debated for decades. The American Medical Association has tended toward the invasive extreme -- issuing guidelines that encouraged hospitalization and many lab and blood tests, sometimes including a spinal tap. This conservative approach reflects the institutional fear that doctors who do less might miss a case of serious illness such as bacteria in the blood or bacterial meningitis. It has resulted in relatively healthy babies spending unnecessary days in the hospital.

But the early 2000s saw a series of published studies that questioned this response. One study, conducted by pediatrics researchers at the University of California at San Francisco, concluded that doctors in private practices and HMOs could choose not to follow the AMA guidelines without compromising their patients' health.

The UCSF researchers evaluated data from 3,066 infants aged 3 months or younger seen by 573 practitioners from all parts of the U.S. The babies all had body temperatures of at least 100.4 degrees Fahrenheit. The researchers tracked what tests were ordered and whether the children were hospitalized. Their conclusion: Following AMA guidelines strictly didn't do anything to improve the diagnoses -- but it did result in the hospitalization of 40 percent more of the babies, which cost more, of course, than following the infants via regular office visits; and it caused a lot more stress for parents.

Of the 3,066 babies, 14 had bacterial meningitis. Another 49 had bacteria, E. coli or group B strep in the blood. In two cases, doctors missed one of these serious problems -- but both babies were examined again within a few days and treated correctly.

So, why the aggressive guidelines? Cynics might snort that the guidelines assure lots of full beds in pediatrics wards. But pediatricians say that the guidelines were based on experiences in inner-city emergency departments. In those settings, doctors say, there's often no follow-up -- and a pediatrician may not see a sick baby for weeks or months after an emergency room visit.

But that situation isn't true everywhere. In the UCSF study, only 4 percent of the cases involved one-time visits.