Brazilian Journal of Pulmonology

ISSN (on-line): 1806-3756 | ISSN (printed): 1806-3713


Publication continuous and bimonthly

SCImago Journal & Country Rank
Advanced Search


Current Issue: 2016 - Volume 42 - Number 5 (September/October)


Randomization: beyond tossing a coin

Randomização: mais do que o lançamento de uma moeda


Juliana Carvalho Ferreira1; 2; Cecilia Maria Patino1; 3


1. Methods in Epidemiologic, Clinical and Operations Research-MECOR-program, American Thoracic Society/Asociación Latinoamericana del Tórax.
2. Divisão de Pneumologia, Instituto do Coração - InCor - Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo, São Paulo, Brasil.
3. Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA.





Randomization is a research strategy used in order to increase the validity of clinical trials evaluating the effect of interventions (e.g., drugs or exercise). It involves the random allocation of participants to either intervention or control groups and requires that participants have an equal chance of being allocated to either group. When properly implemented, randomization prevents selection bias and produces comparable study groups in terms of known and unknown baseline risk factors. For randomization to work, investigators and participants must be unable to predict to which group each of the participants will be allocated-this is called allocation concealment; in addition, investigators must be unable to change the allocation of any participant after randomization.


Simple randomization is equivalent to tossing a coin: a new participant has an equal chance of being assigned to intervention or control groups, independently of previous assignments. Instead of tossing a coin, however, a randomization list is generated by a computer and used to prepare sequentially numbered, sealed envelopes, or, preferably, that list is administered by a central telephone service or website. The ad-vantages of simple randomization are that it is inexpensive and easy to implement. The disadvantages include the risk of producing imbalances in the number of participants in the groups, as well as in the distribution of baseline risk factors, in studies with small sample sizes (N < 100; Figure 1).

In block randomization, the randomization list is a random sequence of blocks of participants instead of individual participants. The blocks have a pre-determined size; for example, four participants in one block, with six possible intervention and control sequences. This strategy ensures that intervention and control groups are balanced in terms of the number of participants (Figure 1). To ensure allocation concealment using this method, random variation of block sizes should be used (four to eight participants per block).

Stratified randomization is an alternative when balance for key baseline risk factors is desired. Each new participant is first classified into stra-ta according to baseline characteristics (e.g., age or disease severity), and each stratum has a separate randomization list. Thereafter, once the participants are categorized into their stratum, they are randomized to either the intervention or the control groups. Stratification should be car-ried out using few relevant strata in order to work well. Stratified and block randomization strategies can be combined so that patients are first categorized into a stratum and then randomized in blocks.

Adaptive randomization uses computer algorithms that take into consideration baseline risk factors and the allocation of previous participants to allocate the next participant. The advantage of this method is that it accommodates more baseline risk factors than stratification and produces optimized group balance at the same time. However, it is more complex and requires a web-based randomization center available 24 h a day.


Simple randomization is easy to implement, is inexpensive, and can be a good option for large trials (N > 200). Block randomization is a good option when balance in the number of participants in each group is desired. Stratification is a good option to provide balance for important co-variates. Adaptive randomization methods may be a good option when the trial structure includes statisticians and information technology sup-port. For all methods, adequate implementation is paramount to ensure allocation concealment and to prevent manipulation and selection bias.


1. Kang M, Ragan BG, Park JH. Issues in outcomes research: an overview of randomization techniques for clinical trials. J Athlc Train. 2008;43(2):215-21.
2. Vickers AJ. How to randomize. J Soc Integr Oncol. 2006;4(4):194-8.



The Brazilian Journal of Pulmonology is indexed in:

Latindex Lilacs SciELO PubMed ISI Scopus Copernicus


CNPq, Capes, Ministério da Educação, Ministério da Ciência e Tecnologia, Governo Federal, Brasil, País Rico é País sem Pobreza
Secretariat of the Brazilian Journal of Pulmonology
SCS Quadra 01, Bloco K, Salas 203/204 Ed. Denasa. CEP: 70.398-900 - Brasília - DF
Fone/fax: 0800 61 6218/ (55) (61) 3245 1030/ (55) (61) 3245 6218

Copyright 2020 - Brazilian Thoracic Association

Logo GN1