Selection bias is a systematic error in a study that occurs from the process used to identify (select) the study participants, allocate them to study groups and from factors that influence study participation [1][2].

It leads to preferential selection or participation of subjects into the study according to their exposure status or outcome/disease status, with resulting systematic differences in the characteristics of participants between the study and control groups, i.e. the groups differ from each other by factors which may affect the outcome of the study [2]. The measurement of the association between exposure and outcome will then differ between those who are included in the study and those who were eligible but not included.

Selection bias may be due to:

  • Sampling bias
  • Ascertainment bias
    • case ascertainment (surveillance) bias
    • referral / admission bias
    • diagnostic bias
  • Participation bias
    • self-selection (volunteerism)
    • healthy worker effect
    • non-response / refusal bias
    • survival bias
    • loss to follow-up

Selection biases in case-control studies include among others: case ascertainment (surveillance) bias, referral bias, diagnostic bias, non-response bias, survival bias.

Selection biases in cohort studies include: healthy worker effect, diagnostic bias, non-response bias, loss to follow-up.

In epidemiological studies, all efforts should be made to avoid biasing the selection of study participants. By paying attention to a number of factors, it is possible to minimise selection bias.

References

1. Rothman KJ. Epidemiology - An Introduction. New York: Oxford University Press; 2002.

2. Bailey L, Vardulaki K, Langham J, Chandramohan D. Introduction to Epidemiology. Black N, Raine R, editors. London: Open University Press in collaboration with LSHTM; 2006.