Astrid Sarvis MD, Children’s National Health System
with Karen O’Connell MD, Children’s National Health System

The Case
6 do FT F sent to the ED by her pediatrician for severe dehydration and 21% weight loss. The challenge of this case involved appropriate diagnostic workup and treatment of this patient.

Here’s How You Answered Our Questions

Denouement
The decision was made in consultation with the NICU to perform an LP. Several hours were spent trying to obtain blood for culture via venous and arterial access. The patient was ultimately transferred to the NICU where LP and blood cultures were successfully performed. The NICU started the patient on Vanc and Gent for a 48hr sepsis rule out. Blood cultures grew Strep virdans after 3 days, susceptible to Vanc, Cefotax, and Clinda so the antibiotics were changed to Cefotaxime based on the susceptibilities. ID recommended an ECHO to rule out endocarditis, given the unknown source of Strep viridans, which was negative. ID recommended IV antibiotics for a full course of 10 days and at least until the last drawn blood culture was negative for 3 days since the first culture grew strep 3 days later. The pt completed a full 10-day course of IV antibiotics. The patient’s newborn screen was confirmed normal, and her hyperglycemia, hypernatremia and hypothermia corrected. She was feeding well at discharge and gained 650 grams during the hospitalization.

Teaching Points
1) Never use a CBC or UA to screen for bacteremia or meningitis. The data concludes that peripheral WBC count alone does not predict risk of these outcomes. If you’re going to do those labs in neonates, you should do an LP.
2) Never trust a “well-appearing” neonate if there are other objective history or physical findings concerning for SBI.
3) Always include metabolic causes in your ddx for an ill-appearing neonate. With that, we should be ordering glucose, complete UA to look for ketones, ammonia and lactate at the very least.

EBM about using peripheral WBC count to screen for SBI in young infants.
1) Is the peripheral WBC count a good screen for bacteremia in young infants?
Bonsu and Harper 2003 concludes it is not.
2) Is the peripheral WBC count a good predictor of CSF pleocytosis in young infants?
Meehan et al 2008 looked at peripheral WBC, height of fever, ANC, and age of the patient to derive a prediction model in young well-appearing infants with fever. They concluded that each individual component was a poor predictor of CSF pleocytosis. However, when used as part of a decision tree, WBC count and height of fever were useful in identifying a group of patients at increased risk of having CSF pleocytosis sensitivity of 89% (95% CI, 83%-92%) and a negative predictive value of 97% (95% CI, 96%-98%).
3) Nosrati et al 2014 looked at diagnostic markers of SBI in general and WBC didn’t pan out here either.