Bias

Bias is a systematic error or deviation from the truth in the results or inferences of a clinical study. It may arise from defects in study design, measurement, analysis, or interpretation and can cause an estimated effect or other study result to differ systematically from the true value.

In clinical research, bias is distinguished from random error. Bias produces systematic deviation, whereas random error produces variation around the true value.

Definitions

Several authoritative sources provide definitions of bias, with wording that varies between sources.

Cochrane Handbook

We define bias as a systematic error, or deviation from the truth, in results 1.

CDISC/NCI Controlled Terminology

Bias refers to defects in study design, measurement, analysis or interpretation such that they cause a result to depart from the true value in a consistent direction 2.

Meaning in clinical research

Bias may arise from participant selection, allocation to intervention groups, measurement of exposures or outcomes, handling of missing data, or selection and reporting of study results. Different methodological frameworks classify these sources into different bias domains, including selection bias, measurement bias, observer bias, detection bias, performance bias, attrition bias, and reporting bias.

Confounding is related to bias but is generally treated as a distinct methodological concept. Confounding occurs when an extraneous variable is associated with both an exposure or intervention and an outcome, resulting in a distorted estimate of the association or treatment effect.

Risk of bias

Risk of bias refers to the likelihood that a study result is systematically distorted because of problems in the design, conduct, analysis, or reporting of the study. Risk-of-bias assessments use domains that vary according to the study design and assessment framework.

Related Terminology

  • Blinding (masking)
  • Randomization

References

  1. Cochrane. 2024. Cochrane Handbook for Systematic Reviews of Interventions, Chapter 7: Considering bias and conflicts of interest among the included studies. Available at: https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-07 ↩︎
  2. Clinical Data Interchange Standards Consortium. 2025. CDISC Glossary Controlled Terminology.
    Available at: https://evs.nci.nih.gov/ftp1/CDISC/Glossary/CDISC%20Glossary.html ↩︎