How to Ensure that the Control and Treated Patients are Similar in all Important Respects

  • Christopher J. Bulpitt
Part of the Developments in Biostatistics and Epidemiology book series (DBEP, volume 1)


It is essential that the treated and control patients are similar in order that any differences in outcome can be attributed to the treatment and not to other factors. Thus far everyone agrees, but how to obtain similar groups is open to some discussion. In this chapter we discuss the two classical methods: (1) using the patient as his own control (cross-over studies) and (2) random (chance) allocation of patients to distinct and concurrently treated groups. Randomisation is expected to result in the groups being similar. The futility of using historical controls is discussed in section 8.2.3; this chapter deals mainly with the advantages, disadvantages, and methods of randomisation.


Anticoagulant Therapy Important Respect Random Number Table Nuisance Variable Restricted Randomisation 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer Science+Business Media Dordrecht 1983

Authors and Affiliations

  • Christopher J. Bulpitt
    • 1
    • 2
    • 3
  1. 1.Department of Medical Statistics and EpidemiologyLondon School of Hygiene and Tropical MedicineLondonUK
  2. 2.Royal Postgraduate Medical SchoolLondonUK
  3. 3.Hammersmith HospitalLondonUK

Personalised recommendations