medical decision support medical decision support
Sign In

Predictive Models of Ho et al for a Severely Burned Patients

Description

Ho et al developed models for predicting length of stay and mortality for a patient with a major burn in Hong Kong. The authors are from the Chinese University of Hong Kong and Prince of Wales Hospital.


 

Patient selection: major burn

 

Parameters:

(1) total burn surface area in percent (from 0 to 100)

(2) age in years

(3) gender

(4) inhalation injury

 

Gender

Points

female

1

male

2

 

 

Inhalation Injury

Points for Length of Stay

Points for Mortality

no

0

1

yes

1

0

 

length of stay in days =

= (2.1 * (percent burns)) - (0.0018 * ((percent burns)^2)) + (16.7 * (points for inhalation)) - (9.4 * (points for gender)) + 8.7

 

X for mortality =

= (-1.7 * (points for inhalation injury)) + (0.001 * ((age)^2)) + (0.001 * ((percent burn)^2)) - 3.6

 

percent mortality =

= 1 / (1 + EXP((-1)*X)) * 100%

 

where:

• The equation in the paper is missing the minus sign in the EXP((-1)*X). It is needed to get the values given in the paper, and is what is typical for a logistic regression model.


To read more or access our algorithms and calculators, please log in or register.

Specialty