In research, the term ‘bias’ is used when a particular design or analysis is likely to favour a particular outcome. In a clinical trial, if one treatment is always given to participants who have a more severe form of a disease, then this treatment will appear worse than others. Bias can also happen if a researcher knows about the treatment a participant is receiving, and this interferes with the researcher’s ability to be impartial.
It is important to avoid bias in health research, as it can distort the results and could lead to unsafe or ineffective treatments being licensed for use, or useful treatments being overlooked. Researchers try to avoid bias by using randomisation and by ‘blinding’ those assessing the results of treatments, which may be both the patient and the doctor.
Blinding / double blinding
Blinding means that whoever is receiving or assessing the effects of treatment does not know which treatment the person has received. This helps to prevent bias. Sometimes the participant will assess the effects of treatment, sometimes the researcher will, and sometimes a researcher who is independent of the trial will carry out this assessment.
In a double blind trial, neither the participant, the doctor nor the researchers running the trial will know which treatment the participant is receiving. The aim is to avoid the hopes and expectations about the treatment, as well as the hopes and expectations of the researchers, influencing the way the benefits and risks are assessed.
It is not always possible to avoid researchers, doctors and participants knowing which treatment they are having. For example, the trial may be comparing surgery with no surgery. If the researcher knows which treatment a participant is receiving, it may be necessary for an independent researcher, who has not been involved in conducting the trial, to assess the impact of the different treatments.