medical decision support 1 medical decision support 2
Subscribe Sign In

Choi Classification and Regression Tree Model

Purpose: To predict possible outcomes for a comatose patient based on clinical findings using the Choi regression tree model.

Description

Choi's classification and regression tree model uses clinical information (pupillary light response, patient age, motor component of the Glasgow Coma Scale, and presence or absence of intracerebral mass lesion) to predict the outcome for the comatose patient.


 

The initial question is whether the pupillary response to light is normal bilaterally, or if it is unilaterally or bilaterally absent.

 

If the pupillary light response was normal bilaterally

 

Age in Years

Other Findings

Outcome

<= 26

motor response GCS <=3

good, moderate or severe

<= 26

motor response GCS > 3

good or moderate

26 < age <= 61

intracerebral mass lesion absent

good or moderate

26 < age <= 61

intracerebral mass lesion present

severe, vegetative or death

> 61

 

death

 

If the pupillary light response is absent on one or both sides

 

Motor Response GCS

Age in Years

Outcome

<=2

 

severe, vegetative or dead

> 2

<= 33

good, moderate or severe

> 2

> 33

dead

 

Performance:

• The accuracy of prediction is better at extremes (good recovery vs death) than in the intermediate levels of disability.

 


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

Specialty

Objective

ICD-10