Sun et al reported models for predicting Staphylococcal aureus bloodstream infection in a hospitalized child. The authors are from Children's Hospital Zhejiang University, Xinhua Children's Hospital, Shanghai Jiao Tong University, Hangzhou Children's Hospital and affiliated institutions.
Patient selection: hospitalized child from 0.9 to 51 months of age
Parameters:
(1) hospital acquired
(2) venous catheter, presumably central
(3) underlying disease
(4) serum procalcitonin in ng/mL
points for serum procalcitonin =
= 4.545 * (procalcitonin)
|
Parameter
|
Finding
|
Points
|
|
hospital acquired
|
no
|
0
|
|
|
yes
|
2.5
|
|
venous catheter
|
no
|
0
|
|
|
yes
|
52.7
|
|
underlying disease
|
none
|
0
|
|
|
congenital heart disease
|
18
|
|
|
GI malformation
|
36
|
|
|
tumor
|
54
|
|
|
funnel chest
|
72
|
|
|
congenital diaphragmatic hernia
|
72
|
where:
• "Other" is listed as one of the 4 underlying disease categories in Table 1 but did not appear in the nomogram. It is hypothesized that "other" might be category 4 (funnal chest, CDH).
total score =
= SUM(points for all of the parameters)
Interpretation:
• minimum score: 0
• maximum score: 227.2
value of X =
= (0.0273 * (score)) - 1.963
risk of incident Staphylococcus aureus sepsis =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under the ROC curve is 0.85.
