Neal et al reported a model for predicting early-onset neonatal sepsis. This can be used in a low-resource setting. The authors are from UCL Great Ormond Street Institute of Child Health and University of Zimbabwe.
Patient selection: neonate
Parameters:
(1) temperature in °C
(2) respiratory rate in breaths per minute
(3) maternal fever during labor
(4) offensive liquour
(5) prolonged rupture of the membranes (PROM)
(6) neonatal lethargy
(7) neonatal irritability, seizures, or coma
(8) neonatal chest retractions
(9) neonatal grunting
|
Parameter
|
Finding
|
Points
|
|
temperature
|
|
0.99 * (temperature)
|
|
respiratory rate
|
|
0.06 / 5 * (resp rate))
|
|
maternal fever
|
no
|
0
|
|
|
yes
|
1.44
|
|
offensive liquor
|
no
|
0
|
|
|
yes
|
0.54
|
|
PROM
|
no
|
0
|
|
|
yes
|
0.36
|
|
neonatal lethargy
|
no
|
0
|
|
|
yes
|
0.59
|
|
irritability, seizures, coma
|
no
|
0
|
|
|
yes
|
0.84
|
|
chest retractions
|
no
|
0
|
|
|
yes
|
0.41
|
|
grunting
|
no
|
0
|
|
|
yes
|
0.18
|
value of X =
= SUM(points for all of the parameters) - 39.4
probability of early-onset sepsis =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under the ROC curve is 0.74.
• The score has a good sensitivity but low specificity, making it useful for screening.
