Table 4.
Comparison of the net benefits in clinical decisions regarding the diagnosis of hepatocellular carcinoma in cirrhotic patients with AFP < 400 μg/L.
Pt (%) | Net benefit | Advantage of the model | ||
---|---|---|---|---|
Intervention to all patients | Intervention based on the model | Increment of net benefit | Unnecessary intervention avoided per 100 patients | |
3 | 0.278 | 0.278 | 0.000 | 0 |
5 | 0.263 | 0.268 | 0.005 | 10 |
8 | 0.239 | 0.253 | 0.014 | 16 |
10 | 0.222 | 0.241 | 0.019 | 17 |
20 | 0.125 | 0.206 | 0.081 | 32 |
30 | 0.000 | 0.178 | 0.178 | 42 |
40 | −0.167 | 0.163 | 0.330 | 50 |
50 | −0.400 | 0.135 | 0.535 | 54 |
60 | −0.750 | 0.111 | 0.861 | 57 |
70 | −1.333 | 0.096 | 1.429 | 61 |
80 | −2.500 | 0.060 | 2.560 | 64 |
90 | −6.000 | 0.033 | 6.033 | 67 |
96 | −16.50 | 0.019 | 16.519 | 69 |
Pt, threshold probability; “Increment of net benefit” is calculated as: net benefit of the model – net benefit of intervention to all; “Unnecessary intervention avoided per 100 patients” is calculated as Vickers and Elkin (17): (net benefit of the model – net benefit of intervention to all)/(pt/(1– pt)) × 100.