Skip to main content

Table 3 Net benefit for ACCENT data.

From: Method for evaluating prediction models that apply the results of randomized trials to individual patients

T

Treat all patients

Treat by prediction

Advantage of prediction

0.5%

0.081

0.081

0.000

1.0%

0.076

0.076

0.000

2.0%

0.066

0.060

-0.007

3.0%

0.056

0.049

-0.007

4.0%

0.046

0.050

0.004

5.0%

0.036

0.044

0.008

7.5%

0.012

0.029

0.017

10.0%

-0.014

0.013

0.013

12.5%

-0.038

-0.003

-0.003

15.0%

-0.064

-0.006

-0.006

17.5%

-0.089

0.000

0.000

20.0%

-0.114

0.000

0.000

25.0%

-0.164

0.000

0.000

  1. T is the treatment threshold corresponding to the minimum reduction in absolute risk required to consider opting for treatment. The units are the equivalent of the number of events (death or recurrence) per patient. The "advantage of prediction" column shows the increment in net benefit of prediction compared to the better of "treat all" and "treat none".