1 / 23

Understanding lack of validity: Bias

Understanding lack of validity: Bias. Objectives. To define and discuss the concept of bias To define and discuss selection bias and information bias. To discuss exposure and outcome identification bias To discuss the results of information bias

paul
Télécharger la présentation

Understanding lack of validity: Bias

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Understanding lack of validity: Bias

  2. Objectives • To define and discuss the concept of bias • To define and discuss selection bias and information bias. • To discuss exposure and outcome identification bias • To discuss the results of information bias • Combine selection/information bias: detection bias, incidence-prevalence bias, temporal bias

  3. Bias - Definition • Is defined as the result of a systematic error in the design or conduct of a study • Results from: • Flaws in the method of selection of study participants • Procedures for gathering relevant exposure and/or disease information

  4. Bias - definition • Bias exists when, on the average, the results of an infinite number of studies differs from the true results • Most biases related to study design and procedures can be classified into: • Selection bias • Information bias

  5. Information bias • A systematic tendency for individuals selected for inclusion in the study to be erroneously placed in different exposure/outcome categories (misclassification) • Examples: recall bias – ability to recall past exposure is dependent on case-control status.

  6. Selection bias • Systematic error in the ascertainment of study subjects • Berksonian bias – when this bias occurs in case-control studies of hospitalized patients.

  7. Recall bias • When recall of past exposure error differs between cases and controls • Methods to prevent recall bias • Review of other documentation (eg. pharmacy or hospital charts) • Using proxy respondents (spouse, parent, etc) • Using biological markers

  8. Interviewer bias • When the disease status is not masked and the interviewer differentially ascertains exposure status • Methods to reduce this error: sub-study interviews, masking of case-control status, standardization in how you ask the question

  9. Outcome identification bias • May occur in both case-control and cohort • Usually due to an imperfect definition of the outcome or errors in the data collection stage. • Observer bias • Respondent bias

  10. Results of information bias • Non-differential misclassificaton (systematic) • Differential (non-systematic)

  11. Effect of misclassification of a confounding variable • Results in an imperfect adjustment due to residual confounding • Can lead to spurious conclusions

  12. The association of IV drug use with CD4 cell counts

  13. Medical surveillance bias • Can be selection or information bias • This bias is most likely to occur when the exposure is a medical condition or therapy that leads to frequent/detailed checkups (eg:diabetes, OCP)

  14. Cross-sectional bias • Incidence-prevalence bias – results from the inclusion of prevalent cases into the study (important when duration of disease is differential) • Temporal bias – don’t know which came first exposure or disease

  15. Lead time bias • The time by which a diagnosis can be advanced by screening • Occurs when estimating survival time

  16. Publication bias • Assumption that published papers should be unbiased and represent an unbiased sample of the theoretical “population” of unbiased studies • Papers with statistically significant results are more likely to be published

More Related