Name: Biesheuvel

Firstname: Egbert

Title: MSc / PhD

Institution: Organon NV

Street: Molenstraat 110

City: Oss

Zip-Code: 5340 BH Oss

Country: The Netherlands

Phone: +31 412 663922

Fax: +31 412 662516


Authors: Biesheuvel, E.H.E. and Hothorn, L.A.

Title: Protocol-designed subgroup analysis in multi-armed clinical trials: multiplicity aspects

Abstract: Subgroup analyses are frequently performed in randomized clinical trials. Assmann et al. (2000) discussed the massive danger of misinterpretation using an un-designed post-hoc analyses. First approaches under the view of multiplicity are proposed by Proschan and Waclawiw (2001). Here we discuss a priori designed multi-armed clinical trials with a subgroup analysis, e.g. the comparison over both genders according to Rochon et al. (1998). We propose a maximum statistic, i.e. maximum over all relevant contrasts. This test is multivariate t-distributed, with a product-moment structure for simpler designs and with an arbitrary correlation structure for more complex designs (Biesheuvel and Hothorn, (2002)). This approach will be demonstrated by means of clinical dose finding studies. The advantage versus Bonferroni will be shown by means of an analytical power estimation. Moreover, related SAS programs and macros will be given.

References: Assmann SF, Pocock SJ, Enos LE, Kasten LE. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet, Mar 25; 355:1064-9 (2000) Biesheuvel E and Hothorn LA. Many-to-one comparisons in stratified designs Biometrical Journal, 44, 101-116 (2002) Proschan MA and Waclawiw MA. Practical guidelines for multiplicity adjustment in clinical trials. Controlled Clinical Trials 21: 527-539 (2001) Rochon PA, Clark JP, Binns MA, Patel V, Gurwich JH. Reporting of gender-related information in clinical trials of drug therapy for myocardial infarction CMAJ, 159(4): 321-327 (1998)