Simultaneous Confidence Intervals for Ratios with Applications to the Comparison of Several Treatments with a Control

G. Dilba; F. Bretz; V. Guiard; L. A. Hothorn

doi:10.1055/s-0038-1633899

Methods of Information in Medicine, Table of Contents

Methods Inf Med 2004; 43(05): 465-469
DOI: 10.1055/s-0038-1633899

Original Article

Schattauer GmbH

Simultaneous Confidence Intervals for Ratios with Applications to the Comparison of Several Treatments with a Control

Authors

G. Dilba

¹Bioinformatics Unit, University of Hannover, Hannover, Germany
F. Bretz

²Novartis Pharma AG, Basel, Switzerland
V. Guiard

³Research Unit Genetics and Biometry, Research Institute for the Biology of Farm Animals, Dummerstorf, Germany
L. A. Hothorn

¹Bioinformatics Unit, University of Hannover, Hannover, Germany

Abstract

Summary

Objectives: In this article, we illustrate and compare exact simultaneous confidence sets with various approximate simultaneous confidence intervals for multiple ratios as applied to many-to-one comparisons. Quite different datasets are analyzed to clarify the points.

Methods: The methods are based on existing probability inequalities (e.g., Bonferroni, Slepian and Šidàk), estimation of nuisance parameters and re-sampling techniques. Exact simultaneous confidence sets based on the multivariate t-distribution are constructed and compared with approximate simultaneous confidence intervals.

Results: It is found that the coverage probabilities associated with the various methods of constructing simultaneous confidence intervals (for ratios) in many-to-one comparisons depend on the ratios of the coefficient of variation for the mean of the control group to the coefficient of variation for the mean of the treatments. If the ratios of the coefficients of variations are less than one, the Bonferroni corrected Fieller confidence intervals have almost the same coverage probability as the exact simultaneous confidence sets. Otherwise, the use of Bonferroni intervals leads to conservative results.

Conclusions: When the ratio of the coefficient of variation for the mean of the control group to the coefficient of variation for the mean of the treatments are greater than one (e.g., in balanced designs with increasing effects), the Bonferroni simultaneous confidence intervals are too conservative. Therefore, we recommend not using Bonferroni for this kind of data. On the other hand, the plug-in method maintains the intended confidence coefficient quite satisfactorily; therefore, it can serve as the best alternative in any case.

Keywords

Bonferroni - Fieller’s theorem - many-to-one comparisons - multivariate t-distribution - simultaneous confidence sets

Full Text

References

References
1 York RG, Funk KA, Girard MF. et al. Oral (drinking water) developmental toxicity study of ammonium perchlorate in Sprague-Dawley rats. Int J Toxicol 2003; 22: 453-64.
2 Hothorn LA, Bauss F. Biostatistical design and analyses of long-term animal studies simulating human postmenopausal osteoporosis. Drug Information J 2004; 38: 47-56.
3 Furberg CD, Wright JT, Davis BR, Cutler JA, Alderman M, Black H, Cushman W. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic. JAMA 2002; 288: 2981-97.
4 Dunnett CW. A multiple comparison procedure for comparing several treatments with a control. J Amer Statist Assoc 1955; 50: 1096-121.
5 Hsu JC. Multiple Comparisons. UK: Chapman and Hall; 1996
6 Strumper D, Durieux ME, Gogarten W, van Aken H, Hartleb K, Marcus M. Fetal plasma concentrations after intra-amniotic sufentanil in chronically instrumented pregnant sheep. Anesthesiology 2003; 98: 1400-6.
7 Hauschke D, Kieser M. Multiple testing to establish non-inferiority of k treatments with a reference based on the ratio of two means. Drug Inform J 2001; 35: 1247-51.
8 Corbett Th, Valeriote F. Tumor models and the discovery and secondary evaluation of solid tumor active agents. Internat J Pharmacognosy 1995; 33: 102-22.
9 Feuerstein TJ, Rossner R, Schumacher M. How to express an effect mean as percentage of a control mean?. J Pharmacol Toxicol Methods 1997; 37: 187-190.
10 Röhmel J. Therapeutic equivalence investigations: Statistical considerations. Statist Med 1998; 17: 1703-14.
11 Pigeot I, Schäfer J, Röhmel J, Hauschke D. Assessing non-inferiority of a new treatment in a three-arm clinical trial including a placebo. Statist Med 2003; 22: 883-99.
12 Dilba G, Bretz F, Guiard V. Simultaneous confidence sets and confidence intervals for multiple ratios (submitted).
13 Fieller EC. Some problems in interval estimation. J Roy Statist Soc Ser B 1954; 6: 175-85.
14 Jensen DR. Joint confidence sets in multiple dilution assays. Biometrical J 1989; 31: 841-53.
15 Hochberg J, Tamhane A. Multiple comparison procedures. New York: Wiley; 1987: 365-9.
16 Westfall PH, Young SS. Resampling based multiple testing. New York: Wiley; 1993: 82-4.
17 Hothorn LA. Statistische Auswerteverfahren. In: Regulatorische Toxikologie Reichl FX ed. Springer Verlag: Heidelberg; 2004: 167-81.
18 Snedecor GW, Cochran WG. Statistical Methods. The Iowa State University Press; 1967: 422
19 Biesheuvel EHE, Hothorn LA. Protocol designed subgroup analyses in multiarmed clinical trials: Multiplicity aspects. J Biopharmaceutical Statistics 2003; 13: 663-73.