Table 2 Comparison of patient outcomes between the original and synthetic cohort

From: Mimicking clinical trials with synthetic acute myeloid leukemia patients using generative artificial intelligence

 

original cohort

CTAB-GAN + 

NFlow

CR after induction therapy, n (%)

1135 (70.7)

1184 (73.7)

1110 (69.1)

OR

2.41

2.81

2.24

[95%-CI]

[2.16–2.68]

[2.51–3.14]

[2.01–2.49]

p-value

 

0.059

0.356

Median EFS, months (IQR)

7.2 (6.5–8.0)

12.8 (11.8–14.1)

9.0 (8.3–9.7)

HR

1.36

0.74

0.87

[95%-CI]

[1.25–1.47]

[0.68–0.80]

[0.80–0.94]

p-value

 

<0.0001

<0.0001

Median OS, months (IQR)

17.5 (15.7–19.2)

19.5 (15.7–19.2)

16.2 (15.7–19.2)

HR

1.14

0.88

1.00

[95%-CI]

[1.04–1.24]

[0.81–0.96]

[0.92–1.09]

p-value

 

<0.0001

0.055

  1. Logistic regression and Cox proportional hazard models were used to obtain odds ratios (OR) for achievement of complete remission (CR) and hazard ratio (HR) with corresponding 95%-confidence intervals (95%-CI). Boldface indicates statistical significance (p < 0.05). p-values are calculated using two-sample comparisons between each of the synthetic cohorts and the original cohort for reference. n number.
Back to article page