Description

Wai et al developed a number of models for predicting fibrosis and cirrhosis in patients with chronic viral hepatitis C (HCV). The authors are from the University of Michigan


 

Independent predictors of fibrosis from multivariate analysis:

(1) platelet count

(2) serum AST

(3) serum alkaline phosphatase

 

Independent predictors of cirrhosis from multivariate analysis:

(1) platelet count

(2) serum AST

(3) serum alkaline phosphatase

(4) white blood cell count

(5) AST to ALT ratio

 

Models using AST and platelet count only were comparable to ones with more variables.

 

Parameters used in models:

(1) serum AST in IU/L

(2) upper limit of normal reference range for AST

(3) platelet count (as factor * 10^9/L)

 

AST to platelet ratio index = APRI =

= ((serum AST) / (upper limit of normal for AST)) / (factor from platelet count) * 100 =

= 100 * (serum AST) / ((upper limit normal AST) * (factor for platelet count))

 

risk score for significant fibrosis =

= 2.318 + (0.274 * LN((serum AST) / (upper limit of normal))) - 0.375 * LN(factor for platelet count))

 

risk score for cirrhosis =

= 2.411 + (0.1 * LN((serum AST) / (upper limit of normal))) - 0.463 * LN(factor for platelet count))

 

Interpretation:

• Different cutoff values for the APRI are used for fibrosis and cirrhosis.

• The cutoffs for the two risk scores were not given (page 521). My guess is that a negative number is against the diagnosis and a positive number if for.

APRI

Significant Fibrosis

<= 0.50

absent

0.51 to 1.50

indeterminant

> 1.50

present

 

 

APRI

Cirrhosis

<= 1.00

absent

1.01 to 2.00

indeterminant

> 2.00

present

 

Combining these:

 

APRI

Diagnosis

<= 0.50

fibrosis absent, cirrhosis absent

0.51 - 1.00

fibrosis indeterminate, cirrhosis absent

1.01 - 1.50

fibrosis indeterminate, cirrhosis indeterminant

1.51 to 2.00

fibrosis present, cirrhosis indeterminant

> 2.00

cirrhosis present

 


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