A typical controlled clinical trial results in many measurements of safety and efficacy taken at different points in time, with some planned data missing or accumulated at times and under conditions not planned for. This paper reviews how statistical techniques for handling multivariate data can be applied to such a study. It points out the weaknesses of traditional multivariate approaches and describes techniques of analysis which have been developed in the last 15 years that might be more useful.
HorwitzRIViscoliCMClemensJDSadockRI. Developing improved observational methods for evaluating therapeutic effectiveness. Am J Med.1990;89: 630–638.
2.
WeissmanA. Two demographic dichotomies: “Black vs White,” “Left vs Right.”Drug Inf J.1992;26:351–362.
3.
GreenhouseSWGeisserS. On methods in the analysis of profile data. Psychometrika. 1959;32: 92–112.
4.
HarvilleDA. Maximum likelihood approaches to variance component estimations and to related problems. J Am Stat Assn. 1977;72: 320–338.
5.
LairdNLangNStramD. Maximum likelihood computations with repeated measures: Application of the EM algorithm. J Am Stat Assn. 1987;82: 97–105.
6.
WolfingerRTobiasRSallJ. Structuring calculations for mixed linear models. Proc of the Stat Comp Sect Am Stat Assn. 1991:150–155.
7.
SalsburgD. The Use of Restricted Significance Tests in Clinical Trials. New York: Springer-Verlag; 1992.
8.
AndersonPKGillRD. Cox's regression model for counting processes: A large sample study. An Stat. 1982;10: 1100–1120.
9.
HuberPJ. Robust Statistics. New York: John Wiley; 1981.
10.
DollingerMBStaudteRG. Influence functions of iteratively reweighted least squares estimators. J Am Stat Assn. 1991;86: 709–716.
11.
MorrisonDF. Multivariate Statistical Methods. New York: McGraw-Hill; 1967.
12.
CoxDRWermuthN. Response models for mixed binary and quantitative variables. Biometrika. 1992;79: 441–462.
13.
AnscombeFJ. Computing in Statistical Science Through APL. New York: Springer-Verlag; 1988.
14.
SAS Institute.SAS User's Guide: Statistics, Version 5 Edition. Chapter 17. Cary, NC: SAS Institute; 1955.
15.
WolfowitzJ. Finitely additive probability functions. An Math Stat. 1949;20: 141–143.
16.
HuberPJ. Projection pursuit (with discussion). An Stat. 1985;13: 435–525.
17.
AsimovD. The grand tour—a tool for viewing multidimensional data. SIAM J Scien Stat Comp. 1985;6: 128–143.
18.
O'BrienPC. Procedures for comparing samples with multiple endpoints. Biometrics. 1984;40: 1679–1687.
19.
ConoverWJSalsburgDS. Locally most powerful tests for detecting treatment effects when only a subset of patients can be expected to “respond” to treatment. Biometrics. 1988;44: 198–206.
20.
JohnsonRAVerrillSMooreDH. Two sample rank tests for detecting changes that occur in a small proportion of the treated population. Biometrics. 1987;43: 641–655.
21.
HajekJSidakZ. Theory of Rank Tests. New York: Academic Press; 1969.
22.
FriedmanJHRafskyLC. Graphics for the multivariate two sample problem (with discussion). J Am Stat Assn. 1981;76: 277–295.
23.
HenzeN. A multivariate two sample test based on the number of nearest neighbor type coincidences. An Stat. 1988;16: 772–783.
24.
BriemanLFriedmanJHClassification and Regression Trees. Belmont, CA: Wads-worth; 1984.
25.
MullerHG. Nonparametric Regression Analysis of Longitudinal Data. New York: Springer-Verlag; 1988.
26.
SilvermanBW. Some aspects of the spline smoothing approach to nonparametric regression curve fitting (with discussion). J Royal Stat Soc B. 1985;47: 1–41.
27.
EfronBTibshiraniR. Bootstrap methods for standard errors, confidence intervals, and other methods of statistical accuracy. Stat Sci. 1986;1: 54–77.
