Common Sources of Bias in Clinical Research
Bias is a systematic error that pushes results away from the truth. Unlike random error, it does not shrink with a larger sample. The main sources are selection bias, information (measurement) bias, confounding and reporting bias, and each can be reduced by design choices.
Main types and examples
| Type | What happens | Example | Ways to reduce |
|---|---|---|---|
| Selection bias | Study groups differ systematically from the target population or from each other | Only hospitalised patients are included | Define source population; random sampling; careful control selection |
| Information bias | Exposure or outcome is measured inaccurately or differently by group | Cases recall past exposure better than controls | Standard instruments; blinding; objective measures |
| Confounding | A third factor related to exposure and outcome distorts the association | Age affects both treatment choice and outcome | Randomisation; matching; adjustment |
| Reporting bias | Results are selectively reported or published | Only "significant" outcomes reported | Pre-register; report all outcomes |
| Performance and detection bias | Differences in care or outcome assessment | Unblinded assessors rate improvement | Blinding; standardised assessment |
Bias versus chance
Chance error produces wide confidence intervals; bias produces results that are consistently wrong. A large study with serious bias can be precisely wrong.
Illustrative example
Invented for teaching. A survey about exercise habits is emailed to hospital staff; only 25% respond, and responders are more health-conscious. Prevalence of regular exercise is overestimated (non-response selection bias). The authors can reduce this by following up non-responders and by reporting the response rate and its limits.
Common pitfalls
- Believing statistical adjustment removes all bias.
- Listing "bias" in the limitations without naming which kind and its likely direction.
Key takeaways
- Think about bias when designing the study, not only when appraising it.
- Name the specific bias and its likely effect on results.
References and further reading
- Higgins JPT, Thomas J, Chandler J, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane. Link
- Hulley SB, Cummings SR, Browner WS, Grady D, Newman TB. Designing Clinical Research. 4th ed. Lippincott Williams & Wilkins; 2013. Link
- Sterne JAC, Savovic J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. doi:10.1136/bmj.l4898
- Sterne JA, Hernan MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919. doi:10.1136/bmj.i4919
Related reading
This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.