AAP Voices Blog

As I first learned in my third year of medical school, fever is one of the most common reasons children end up at the emergency department. The probable causes sometimes are obvious, like when a peek through the otoscope shows the red, bulging eardrum of otitis media. But often, the cause of a child's fever isn't readily apparent.

It's especially challenging to diagnose young infants with a serious infection that can overwhelm their immature immune system and even be fatal. Babies can't tell us how bad they feel, or where their pain is. They often don't look really sick until they're near the point treatment could no longer help them. Each year, more than 75,000 children develop severe sepsis, and almost 7,000 of them will die. That's why infants with fevers of unknown cause are usually admitted for extensive and invasive workups to rule out serious infections and sepsis.

My first exposure to the management of young infants with fever with an unknown source came during my third year of medical school. As we headed to the E.R. to admit yet another baby with fever,

I asked my resident if any babies who got septic workups ever actually had an infection that required treatment. In response, he took me down the hall to see a 3-year-old patient with a G- tube and tracheostomy who was profoundly developmentally delayed.

At 2 weeks old, the child had a 101-degree fever but looked well when his mother put him to bed one night. By the next morning. E. Coli meningitis had devastated him. This gave me a (very) healthy respect for infants with fever, but my experience still made me think our approach to ruling out sepsis was overkill. We were catheterizing babies, doing spinal taps, treating them with broad spectrum antibiotics, and hospitalizing them for 48 hours, yet only a very few of them actually turned out to have a serious bacterial infection.

Over the last several decades, pediatricians have developed numerous protocols to try to better define which infants need this full sepsis evaluation and presumptive treatment with antibiotics until the tests are negative. The goal has been to identify babies who may need fewer tests and perhaps could be sent home without antibiotic treatment. The approach to management of very young infants with fever has varied widely across the country.

The fear of missing something loomed large for years, however, and the time and academic skills needed to sift through all the literature to consider making a local change, eluded me. Then I had the opportunity to join the American Academy of Pediatrics’ Reducing Excessive Variability in Infant Sepsis Evaluations (REVISE) quality improvement project. The REVISE study attempted to standardize the management of infants with fever, safely doing less for some of them while ensuring that the high-risk infants get all the testing needed to evaluate for sepsis.

As a pediatric hospitalist working in a large community hospital with a small inpatient pediatric unit, leading pediatric change can be challenging. My group is small and perpetually stretched thin just to cover all the clinical needs of our service. There is no time during a clinical shift to do administrative work, and we are working more shifts than we each want to as we try to balance the demands of our program with our families, relationships, and sanity.

That is why being part of REVISE was such a huge benefit. I do not have any kind of data analysis support, so the ability to put all my data into QIDA and get it analyzed immediately was pivotal. I reaped the benefit of countless hours of an expert panel’s work when I used the prepared Power Point slides to present this project to my department, for example, allowing me to make a convincing presentation to get my colleagues on board with this idea. I have been able to demonstrate tangible improvements in care, reductions in cost and radiation, and reduced length of stay and admission rates for this population.

"We were catheterizing babies, doing spinal taps, treating them with broad spectrum antibiotics, and hospitalizing them for 48 hours, yet only a very few of them actually turned out to have a serious bacterial infection.”

Participation in the REVISE project also gave us access to basic algorithms for babies who would fall under the project’s guidelines. The algorithms required some tweaking to fit into our environment and culture, but it was so much easier to make minor modifications to existing algorithms then it would have been to develop them from scratch.

An unanticipated benefit was that the REVISE project gave me the impetus to really look closely at what we were doing and not doing. Before I collected the baseline data I would have guessed we were close to 100 percent consistent about getting a urinalysis when evaluating possibly infected babies and not getting a chest x-ray when there were no respiratory symptoms. As it turned out, our urinalysis rate was 78 percent and unnecessary chest x-rays were being done on 40 percent of babies. Those were aspects of REVISE that we didn’t think would be a change for us, but in fact looking at our data and actually measuring our performance led to unexpected improvement.

My emergency department partner and I reviewed literature and debated the relative merits of various inflammatory markers before deciding on procalcitonin as our marker of choice. This gave us the opportunity to explore the challenges of getting authorization to routinely use a relatively new test and to define the cutoff for “normal” values for our population. In a children’s hospital the norms built into the system are pediatric, but in our world everything is geared toward adults.

Many children are treated in community hospitals outside of university-based academic centers. When smaller community programs like mine can participate in quality improvement projects like this, it gives these children access to the same high quality, current care provided in larger medical centers.

By chance, we recently saw two babies with fever, each under 3 weeks old, whose siblings had come in with fever at the same age. Although the younger babies were both admitted, their post-REVISE stays were 24 rather than 48 hours, and they weren’t immediately started on antibiotics. The families loved our new approach, saying they were glad their younger children spent less time in the hospital and weren’t exposed to the unnecessary antibiotics.

I was worried the families might be concerned because we were doing "less," but they were happy to be part of this change.​

*The views expressed in this article are those of the author, and not necessarily those of the American Academy of Pediatrics.