The AB/BA design is reviewed from a historical perspective. Particular attention is paid to the problem of carry-over and various attempts to deal with it. The two-stage procedure, an approach which was popular for many years, is shown to be unsafe. The analysis of AB/BA designs with baseline data is also considered. It is shown that such baselines do not provide a cure for the problem of carry-over; and it is concluded that any rational analysis of such trials will always be dependent on assumptions regarding carry-over, and that it is necessary to pay particular attention to washout periods. Under such circumstances analysis of covariance may be useful. In conclusion, some speculative comments about future lines of research are offered.
Get full access to this article
View all access options for this article.
References
1.
Senn SJCross-over trials in clinical research. Chichester: John Wiley, 1993.
2.
Palmqvist M. , Balder B., Löwhagen O., Melander B., Svedmyr N., Wåhlander L.Late asthmatic reaction decreased after pretreatment with salbutamol and formoterol, a new long acting β2 agonist . Journal of Allergy and Clinical Immunology1992; 89: 844-49.
3.
Senn SJThe use of baselines in clinical trials of bronchodilators. Statistics in Medicine1989; 8: 1339-50.
4.
'Student'. The probable error of a mean. Biometrika1908; 6: 1-25.
5.
Cushy AR, Peebles ARThe action of optical isomers. II. Hyoscines. Journal of Physiology1905; 32: 501-10.
6.
Preece DAT is for trouble (and textbooks): a critique of some examples of the paired-samples t-test. The Statistician1982: 31: 169-95.
7.
Senn SJ, Richardson W.The first t-test. Statistics in Medicine1994 : 13: 785-803.
8.
Marks HPThe biological assay of insulin preparations in comparison with a stable standard . British Medical Journal1925: December 12; 1102-04.
9.
Fieller ECThe biological standardization of insulin (with discussion). Supplement to the Journal of the Royal Statistical Society1940: VII: 2-64.
10.
Chassan JBOn the analysis of simple crossovers with unequal numbers of replicates. Biometrics1964: 20: 206-08.
11.
Hills M., Armitage P.The two-period crossover clinical trial. British Joumal of Clinical Pharmacology1979: 8: 7-20.
12.
Jones B., Kenward MGDesign and analysis of cross-over trials. London: Chapman and Hall, 1989.
13.
Fleiss JLLetter to the editor. Biometrics1986: 42: 449-50.
14.
Fleiss JLA critique of recent research on the two-treatment cross-over design. Controlled Clinical Trials1989: 10: 237-43.
15.
Senn SJIs the simple carry-over model useful? Statistics in Medicine : 11: 715-26.
16.
Freeman P.The performance of the two-stage analysis of two-treatment, two-period cross-over trials. Statistics in Medicine1989: 8: 1421-32.
17.
Grizzle JEThe two-period change over design and its use in clinical trials. Biometrics1965: 21: 467-80.
18.
Grizzle JECorrection to Grizzle (1965). Biometrics1974 : 30: 727.
19.
Grieve APThe two-period change over design and its use in clinical trials. Biometrics1982: 38: 517.
20.
Grieve APA Bayesian analysis of the two-period cross-over trial. Biometrics1985: 41: 979-90.
21.
Koch GGThe use of non-parametric methods in the statistical analysis of the two-period change-over design. Biometrics1972: 28: 577-84.
22.
Armitage P. , Hills M.The two-period crossover trial. The Statistician1982: 31: 119-31.
23.
Clayton D., Hills M.A two-period cross-over trial. In: The statistical consultant in action (Eds. Hand DJ, Everitt BS). Cambridge: Cambridge University Press, 1987.
24.
Dubey S.Current thoughts on cross-over designs. Clinical Research Practices and Drug Regulatory Affairs1986: 4: 127-42.
25.
Brown B.The cross-over experiment for clinical trials. Biometrics1980: 36: 69-79.
26.
Poloniecki JD, Daniel D.Further analysis of the Hills and Armitage enuresis data. The Statistician1981: 30: 225-29.
27.
Senn SJCross-over trials, carry-over effects and the art of self delusion. Statistics in Medicine1988: 8: 715-26.
28.
Senn SJProblems with the two-stage analysis of cross-over trials. British Journal of Clinical Pharmacology1991: 32: 133.
29.
Jones B., Lewis J.The case for cross-over trials in phase III. Statistics in Medicine1994 (in press).
30.
Matthews Jns.Cross-over trials, Newcastle : University of Newcastle upon Tyne, 1990.
31.
Zimmerman H. , Rahlfs VWModel building and testing for the change-over design. Biometrical Journal1980: 22: 197-220.
32.
Lehmacher W.Analysis of the cross-over design in the presence of residual effects. Statistics in Medicine1991: 10: 891-99.
33.
Willan AR, Pater JLCarry-over and the two-period crossover design. Biometrics1986: 42: 593-99.
34.
Willan ARUsing the maximum test statistic in the two-period cross-over clinical trial . Biometrics1988: 44: 211-18.
35.
Grieve APA note on the analysis of the two-period crossover design when the period-treatment interaction is significant. Biometrical Journal1987: 29: 771-75.
36.
Chassan JBA note on relative efficiency in clinical trials. Journal of Clinical Pharmacology1970: 10: 359-60.
37.
Kenward M., Jones B.The analysis of data from 2×2 cross-over trials with baseline measurements . Statistics in Medicine1987: 6: 911-26.
38.
Ratkowsky DA, Evans MA, Alldredge JRCross-over experiments: design, analysis and application. New York: Marcel Dekker, 1993.
39.
Grieve APExtending a Bayesian analysis of the two-period crossover to allow for baseline measurements. Statistics in Medicine1994: 13: 905-29.
40.
Willan AR, Pater JLUsing baseline measurements in the two period cross-over trial. Controlled Clinical Trials1986: 7: 282-89.
41.
Chi EMAnalysis of cross-over trials when within-subject errors follow an AR(1) process . Biometrical Journal1992: 34: 359-65.
42.
Senn SJLetter to the editor: a comment on Chi, EM (1992). Biometrical Journal1994: 36: 359-65.
43.
Varma AO, Chilton NWCross-over designs involving two treatments. Journal of Periodontal Research1974; 9 suppl 14; 160-70.
44.
Patel HIThe use of baseline measurements in the two-period cross-over design. Communications in Statistics (A)1983: 12: 2693-712.
45.
Matthews Jns.Book reviews. Statistics in Medicine1994: 13: 298-300.
46.
Grieve AP, Senn SJ.Estimating treatment effects in clinical cross-over trials, (in preparation).
47.
Koch MAPrecision and bias of baseline-adjusted estimators in the 2×2 cross-over. American Statistical Association, Proceedings of the Biopharmaceutical Section1992; 68-73.
