Table 1 Cut off points, and positive (PPV) and negative (NPV) predictive values of the indexes evaluated, prevalence of significant fibrosis or cirrhosis in the validation studies, and proportion of liver biopsies that could be prevented.
Models aimed at predicting significant fibrosis | Prevalence of significant fibrosis (%) | Liver biopsies prevented (%) | |||
---|---|---|---|---|---|
Model | Score cut off point | PPV (%) | NPV (%) | ||
Forns8 | 26 | 51 | |||
<4.2 | 40 | 96 | |||
>6.9 | 66 | 80 | |||
APRI9 | 47 | 51 | |||
<0.5 | 61 | 86 | |||
>1.5 | 88 | 64 |
Models aimed at predicting cirrhosis | Prevalence of cirrhosis (%) | Liver biopsies prevented (%) | |||
---|---|---|---|---|---|
Model | Cut off score | PPV (%) | NPV (%) | ||
APRI9 | 15 | 15 | |||
<1 | 38 | 98 | |||
>2 | 57 | 93 | |||
Saadeh13,14 | 29 | 29 | |||
<3* | — | — | |||
>7 | 100 | 73 | |||
ALT/AST ratio11,14 | 23–34 | Up to 34 | |||
<1 | 43–100 | 78–88 | |||
Platelet count 12 | 13 | 13 | |||
<150 | 93 | 99 |
AST, aspartate aminotransferase; ALT, alanine aminotransferase; APRI, AST platelet ratio index.
*The authors did not provide enough data to calculate PPV and NPV for this cut off point.