Potential Advantages and Disadvantages of Stratification in Methods of Randomization

Conference paper
Part of the Springer Proceedings in Mathematics & Statistics book series (PROMS, volume 114)


Randomization of patients to different therapy groups has been established to a gold standard in clinical trials. But pre-stratification of randomization has been discussed as a controversial issue in this context. To support investigator’s decision concerning stratification in the phase of planning a trial we investigated the impact of stratification with respect to the risk of prognostic imbalance between treatment groups by a simulation approach. We give a comprehensive overview of the risk for pre-defined imbalances, several trial sizes and prevalence of a prognostic factor, comparing stratified vs. unstratified randomization.We quantified the maximum risk of a prognostic imbalance due to randomization of 59 % (complete randomization CR, N = 30, prevalence of a prognostic factor 50 %). For type I error we calculated a maximum of 32 % (permuted-block randomization PBR(B), N = 100, average success rate 50 %) for a clinically relevant difference, and about 5 % for a statistically significant difference in trials with N = 100 or 400 patients. Stratification can be helpful to reduce this risk by up to 16 percentage points (pps) for clinical differences in the case of a large average success rate of 50 %, and large differences between the strata (10 % vs. 90 %) in small trials of N = 100. For statistical differences, however, the impact of stratification is rather negligible.


Prognostic Factor Success Rate Therapy Group Large Trial Small Trial 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Kernan, W.N., et al.: Stratified randomization for clinical trials. J. Clin. Epidemiol. 52(1), 19–26 (1999)CrossRefGoogle Scholar
  2. 2.
    Feinstein, A.R., Landis, J.R.: The role of prognostic stratification in preventing the bias permitted by random allocation of treatment. J. Chronic Dis. 29(4), 277–284 (1976)CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2014

Authors and Affiliations

  1. 1.Institute for Biostatistics and Informatics in Medicine and Ageing ResearchUniversity Medicine RostockRostockGermany

Personalised recommendations