medical decision support medical decision support
Sign In

Predictive Models of Hu et al. for Hospital-Acquired Pneumonia (HAP) in a Patient with Aneurysmal Subarachnoid Hemorrhage (SAH): Logistic Regression Model

Description

Hu et al reported models for predicting hospital-acquired pneumonia (HAP) in patients with aneurysmal subarachnoid hemorrhage. One model uses logistic regression. The authors are from Tongji Medical College of Huazhong University of Science and Technology in Wuhan, China.


Patient selection: aneurysmal subarachnoid hemorrhage

 

Parameters:

(1) age in years

(2) blood glucose in mmol/L

(3) Glasgow coma score (GCS)

(4) standard deviation for the RDS in fL

 

Parameter

Finding

Points

age in years

 

0.057 * (age)

blood glucose

<= 7.22 mmol/L

0

 

> 7.22 mmol/L

1.023

Glasgow coma scale

 

-0.342 * (GCS)

RDW SD in fL

 

0.111 * (RDW SD)

 

value of X =

= SUM(points for all of the parameters) - 5.467

 

probability of HAP =

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

 

Performance:

• The area under the ROC curve is 0.87.

 

On page 6 a different equation is given which returns values closer to the point score model (Model-Cat):

 

value of Y =

= (0.57 * (SUM(points for all of the parameters))) - 3.2785

 

probability of HAP =

= 1 / (1 + EXP((-1) * Y))

 


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

Specialty