Frugoni et al reported a model for predicting a prolonged hospital length of stay for a surgical burn patient. The authors are from the University of California San Diego and Outcomes Research Consortium (Cleveland).
Patient selection: burn. Most of the patients are adults.
Parameters:
(1) patient age
(2) hypertension
(3) total burn surface area (TBSA) in percent, from 0 to 100 (most patients has 1-10%)
(4) involvement of perineum
(5) WBC count in 10^9/L on admission
Parameter
Finding
Points
age
0.25 * (age in decades)
hypertension
no
0
yes
1.45
burn area
1.93 * ((percent) / 10)
involvement of the perineum
no
0
yes
1.14
WBC count
< 4 * 10^9/L
0.67
(4 to 12) * 10^9/L
0
> 12 * 10^9/L
0.67
where:
• Age can be handled in 3 separate ways. (1) INT((age)/10; (2) INT((age)/10) + 1; (3) (age)/10.
• Percent TBSA burned can be handled similarly.
value of X =
= SUM(points for all of the parameters) - 2.72
probability of prolonged hospital stay =
= 1 / (1 + EXP((-1) * X))
Performance:
• The area under the ROC curve for the training cohort was 0.85.
To read more or access our algorithms and calculators, please log in or register.
