Blinding or Masking
Published Online: 15 JUL 2005
Copyright © 2005 John Wiley & Sons, Ltd
Encyclopedia of Biostatistics
How to Cite
Day, S. J. 2005. Blinding or Masking. Encyclopedia of Biostatistics. 1.
- Published Online: 15 JUL 2005
The need for blinding in clinical trials and epidemiological studies is reviewed and its importance stressed. Various reasons for blinding include the potential for patient bias, bias in patient management and evaluation, bias in data management and study management. Blinding and unblinding interim analyses need special attention, and procedures for emergency unblinding need to be available.
Four levels of blinding are described from the slightly ambiguous term “single-blind” through to quadruple-blinding. Methods to maintain the blind at various stages such as use of placebos, “double-dummy” designs, and sham procedures are discussed.
Finally, some of the difficulties of blinding are highlighted. Often, it may be ineffective and reporting of the achieved quality of blinding is recommended. There are special issues with regard to equivalence studies, and regulatory guidance exists discussing many issues included in this article.
- case–control studies;
- clinical trials;