A simulation study on estimating biomarker–treatment interaction effects in randomized trials with prognostic variables

Haller, Bernhard; Ulm, Kurt

doi:10.1186/s13063-018-2491-0

Trials

Table 2 Proportions of rejected null hypotheses and numbers of included covariates for scenarios with K=12

From: A simulation study on estimating biomarker–treatment interaction effects in randomized trials with prognostic variables

K	Σ	β _k	Censoring	Main	True	AIC_A	AIC_B	Significance	Full
No interaction
12	Σ ₁	β _eq	Low	0.066	0.069	0.072	0.066	0.067	0.069
12	Σ ₁	β _v	Low	0.051	0.059	0.059	0.055	0.052	0.058
12	Σ ₂	β _eq	Low	0.050	0.051	0.054	0.048	0.051	0.051
12	Σ ₂	β _v	Low	0.060	0.047	0.052	0.047	0.052	0.052
12	Σ ₃	β _eq	Low	0.060	0.065	0.066	0.061	0.063	0.065
12	Σ ₃	β _v	Low	0.063	0.066	0.067	0.060	0.067	0.065
12	Σ ₁	β _eq	High	0.057	0.056	0.060	0.054	0.055	0.056
12	Σ ₁	β _v	High	0.038	0.040	0.037	0.034	0.040	0.044
12	Σ ₂	β _eq	High	0.057	0.046	0.054	0.044	0.046	0.046
12	Σ ₂	β _v	High	0.060	0.047	0.058	0.049	0.055	0.055
12	Σ ₃	β _eq	High	0.053	0.059	0.060	0.050	0.062	0.059
12	Σ ₃	β _v	High	0.057	0.056	0.064	0.055	0.061	0.065
Quantitative interaction
12	Σ ₁	β _eq	Low	0.122	0.137	0.143	0.121	0.123	0.137
12	Σ ₁	β _v	Low	0.119	0.134	0.139	0.132	0.134	0.136
12	Σ ₂	β _eq	Low	0.105	0.131	0.136	0.129	0.131	0.131
12	Σ ₂	β _v	Low	0.113	0.131	0.141	0.135	0.132	0.132
12	Σ ₃	β _eq	Low	0.129	0.146	0.148	0.146	0.141	0.146
12	Σ ₃	β _v	Low	0.121	0.121	0.127	0.120	0.116	0.120
12	Σ ₁	β _eq	High	0.098	0.109	0.120	0.109	0.100	0.109
12	Σ ₁	β _v	High	0.108	0.112	0.118	0.111	0.108	0.111
12	Σ ₂	β _eq	High	0.077	0.088	0.099	0.086	0.088	0.088
12	Σ ₂	β _v	High	0.104	0.095	0.106	0.096	0.093	0.093
12	Σ ₃	β _eq	High	0.086	0.093	0.095	0.091	0.088	0.093
12	Σ ₃	β _v	High	0.095	0.101	0.107	0.098	0.100	0.101
Qualitative interaction
12	Σ ₁	β _eq	Low	0.625	0.685	0.673	0.662	0.644	0.685
12	Σ ₁	β _v	Low	0.605	0.664	0.664	0.661	0.649	0.661
12	Σ ₂	β _eq	Low	0.517	0.641	0.641	0.634	0.641	0.641
12	Σ ₂	β _v	Low	0.521	0.646	0.648	0.643	0.644	0.644
12	Σ ₃	β _eq	Low	0.621	0.678	0.686	0.673	0.680	0.678
12	Σ ₃	β _v	Low	0.583	0.661	0.664	0.652	0.661	0.658
12	Σ ₁	β _eq	High	0.427	0.462	0.464	0.446	0.433	0.462
12	Σ ₁	β _v	High	0.424	0.438	0.447	0.432	0.440	0.443
12	Σ ₂	β _eq	High	0.338	0.403	0.410	0.389	0.403	0.403
12	Σ ₂	β _v	High	0.359	0.431	0.429	0.413	0.433	0.433
12	Σ ₃	β _eq	High	0.394	0.424	0.425	0.407	0.420	0.424
12	Σ ₃	β _v	High	0.418	0.466	0.471	0.456	0.465	0.449
Mean number of prognostic covariates included
		β _eq	Low	0	12	6.8	7.3	8.7	12
		β _v	Low	0	8	6.4	6.9	8.7	12
		β _eq	High	0	12	5.1	5.7	8.0	12
		β _v	High	0	8	5.3	5.8	8.2	12

For the scenarios with no true biomarker–treatment interaction, results for methods/strategies with an observed type I error probability above 7% are in italics. For scenarios with a true biomarker–treatment interaction, the observed power is in bold if the type I error probability did not exceed 7%

Back to article page

ISSN: 1745-6215

Contact us

Submission enquiries: Access here and click Contact Us
General enquiries: info@biomedcentral.com