Randomization-based methods for correcting for treatment changes: examples from the Concorde trial.

We develop analysis methods for clinical trials with time-to-event outcomes which correct for treatment changes during follow-up, yet are based on comparisons of randomized groups and not of selected groups. A causal model relating observed event times to event times that would have been observed under other treatment scenarios is fitted using the semi-parametric approach of Robins and Tsiatis (avoiding assumptions about the relationship between treatment changes and prognosis). The methods are applied to the Concorde trial of immediate versus deferred zidovudine, to investigate how the results would have differed if no participant randomized to deferred zidovudine had started treatment before reaching ARC or AIDS. We consider issues relating to model choice, non-constant treatment effects and censoring.

[1]  E. Goetghebeur,et al.  Clinical trials comparing two treatment policies: which aspects of the treatment policies make a difference? , 1998, Statistics in medicine.

[2]  I. White.  Letter to the Editor: Survival analysis of randomized clinical trials adjusted for patients who switch treatments by M. G. Law and J. M. Kaldor, Statistics in Medicine, 15, 2069–2076 (1996) , 1997 .

[3]  N Segnan,et al.  Adjusting for non-compliance and contamination in randomized clinical trials. , 1997, Statistics in medicine.

[4]  Ian R. White,et al.  Impact of treatment changes on the interpretation of the Concorde trial , 1996, AIDS.

[5]  J. Kaldor,et al.  Survival analyses of randomized clinical trials adjusted for patients who switch treatments. , 1996, Statistics in medicine.

[6]  I. White,et al.  Statistical reporting of clinical trials with individual changes from allocated treatment. , 1996, Statistics in medicine.

[7]  J. Phair,et al.  A comparison of immediate with deferred zidovudine therapy for asymptomatic HIV-infected adults with CD4 cell counts of 500 or more per cubic millimeter. AIDS Clinical Trials Group. , 1995, The New England journal of medicine.

[8]  J. Robins,et al.  Adjusting for differential rates of prophylaxis therapy for PCP in high- versus low-dose AZT treatment arms in an AIDS randomized trial , 1994 .

[9]  R. Tibshirani,et al.  An Introduction to the Bootstrap , 1995 .

[10]  Bernard Hirschel,et al.  Concorde trial of immediate versus deferred zidovudine , 1994, The Lancet.

[11]  Ann Marie Swart,et al.  Concorde: MRC/ANRS randomised double-blind controlled trial of immediate and deferred zidovudine in symptom-free HIV infection , 1994, The Lancet.

[12]  P. Graham.  The Gods of the Comic-book Headlines , 1994, The Lancet.

[13]  S. Gore,et al.  The consumer principle of randomisation , 1994, The Lancet.

[14]  Joshua D. Angrist,et al.  Identification of Causal Effects Using Instrumental Variables , 1993 .

[15]  J. Robins,et al.  A method for the analysis of randomized trials with compliance information: an application to the Multiple Risk Factor Intervention Trial. , 1993, Controlled clinical trials.

[16]  Terry M. Therneau,et al.  Use of Compliance Measures in an Analysis of the Effect of Diltiazem on Mortality and Reinfarction After Myocardial Infarction , 1993 .

[17]  S W Lagakos,et al.  Adjusting for early treatment termination in comparative clinical trials. , 1990, Statistics in medicine.

[18]  R. Newcombe.  Explanatory and pragmatic estimates of the treatment effect when deviations from allocated treatment occur. , 1988, Statistics in medicine.

[19]  D. Cox,et al.  Analysis of Survival Data. , 1985 .

[20]  S. Pocock,et al.  Clinical Trials: A Practical Approach , 1984 .

[21]  D. Oakes.  A Model for Association in Bivariate Survival Data , 1982 .

[22]  D. Clayton.  A model for association in bivariate life tables and its application in epidemiological studies of familial tendency in chronic disease incidence , 1978 .

[23]  J. Robins.  Correcting for non-compliance in randomized trials using structural nested mean models , 1994 .

[24]  P Peduzzi,et al.  Analysis as-randomized and the problem of non-adherence: an example from the Veterans Affairs Randomized Trial of Coronary Artery Bypass Surgery. , 1993, Statistics in medicine.

[25]  J. Robins,et al.  Correcting for non-compliance in randomized trials using rank preserving structural failure time models , 1991 .

[26]  S. Zeger,et al.  On estimating efficacy from clinical trials. , 1991, Statistics in medicine.