Bachur and Harper reported a predictive model for identifying a febrile infant less than 3 months (<=90 days) old who has a serious bacterial infection. The authors are from Children's Hospital Boston.
Patient selection: febrile (temperature >=38.0°C) infant less than 3 months old
Parameters:
(1) urinalysis
(2) white blood cell count per microliter
(3) temperature in °C
(4) age in days
Parameters
Findings
Points
urinalysis
negative
0
positive
1
WBC count
< 4,100 per µL
1
4,100 to 20,000 per µL
0
> 20,000 per µL
1
temperature
<= 39.6°C
0
> 39.6°C
1
age
< 13 days
1
>= 13 days
0
where:
• Positive urinalysis is one or more of: positive dipstick leukocyte esterase, positive dipstick nitrite, >= 5 WBC/hpf in spun urine sample
score =
= SUM(points for all 4 parameters)
Interpretation:
• minimum score: 0
• maximum score: 4
• A score >= 1 indicates a high risk for severe bacterial infection.
Performance:
• The negative predictive value for bacteremia or meningitis was 99.6%.
• 18% of serious bacterial infections were classified as low risk; about 80% of these were UTI.
To read more or access our algorithms and calculators, please log in or register.
