Description

Ghobriel et al developed a model for predicting survival in a liver transplant patient after the transplant based on data available prior to the procedure. The model was based on analysis of the patients in the United Network for Organ Sharing (UNOS) database. The authors are from UCLA and the University of Pennsylvania.


 

Parameters:

(1) donor's age

(2) recipient's age

(3) recipient's serum creatinine in mg/dL

(4) duration of warm ischemia in minutes

(5) recipient's first vs second transplant

(6) recipient's serum total bilirubin in mg/dL

(7) recipient's prothrombin time (PT) in seconds

(8) duration of cold ischemia in hours

First vs Second Transplant

Points

first

0

second

1

more than second

NA

 

mortality score =

= (0.0084 * (donor's age in years)) + (0.019 * (recipient's age in years)) + (0.816 * (LN(serum creatinine))) + (0.0044 * (duration warm ischemia in minutes)) + (0.659 * (points for transplant time)) + (0.10 * (LN(serum bilirubin))) + (0.0087 * (prothrombin time)) + (0.01 * (cold ischemia in hours))

 

relative risk of mortality =

= EXP((mortality score) – (reference mortality score))

 

where:

• The reference mortality score in the text on page 318 was 1.9, which was the mean mortality score for the entire database and the mean mortality score for the third quintile.

• The reference mortality score in Table 3 on page 318 was 1.33, which was the mean mortality score for the lowest quintile.

 

Quintile

Mean Mortality Score

Model Relative Risk

1

1.33

1.00

2

1.64

1.37

3

1.85

1.69

4

2.08

2.14

5

2.59

3.52

from Table 3, page 318

 

Interpretation:

• The relative mortality risk predicted by the model and seen in the database were comparable.

• The prediction model works for both hepatitis C and non-hepatitis C patients.

 

Limitations:

• Laboratory values were not referenced to method or normal reference range.

 


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