Skip to main content

Table 1 True and false positive rates of significant subgroup effects in simulated trials under different scenarios

From: Cumulative subgroup analysis to reduce waste in clinical research for individualised medicine

Variables Input values True positive False positive
X1 X2 X3 X4 Any of X2, X3, X4
Ratio of odds ratios (ROR) 0.90 6.2% 4.9% 5.0% 5.0% 14.2%
0.80 10.3% 5.0% 5.1% 4.9% 14.2%
0.70 17.9% 4.8% 5.0% 5.0% 14.1%
No. of patients per arm 200 7.5% 4.9% 4.9% 5.1% 14.1%
400 10.1% 5.0% 5.0% 4.8% 14.2%
600 12.6% 5.1% 5.0% 4.8% 14.2%
Event rate in the control arm 0.20 8.7% 4.9% 5.1% 5.0% 14.3%
0.30 10.2% 5.2% 5.1% 5.1% 14.6%
0.40 11.3% 5.1% 4.9% 5.1% 14.4%
Proportion of patients belonging to a subgroup 20% 8.0% 4.6% 5.0% 4.9% 14.2%
30% 9.3% 5.1% 4.9% 4.9% 14.4%
40% 10.0% 5.3% 4.9% 5.1% 14.5%
  1. Except for varying input values shown in the table, the following basic input parameters were used for other variables: No. of trials = 10; No. of patients per arm = 400; OR = 0.7; event rate in the control arm = 0.3; heterogeneity variance = 0.01; ROR = 0.8; proportion of patients with X1 = 0.4. Positive rates were calculated based on 5000 simulations for each scenario (with a given set of input parameters)