SPHL 602 Module 8: Measurement Error Flashcards
Epidemiology
The study of the distribution and determinants of disease frequency in human populations and the application of this study to control health problems
Descriptive Epidemiology
The analysis of disease patterns according to the characteristics of person, place, and time
Who is getting the disease? Where is it occurring? How is it changing over time?
Bias
A systematic error that results in an incorrect (invalid) estimate of the measure of association between the exposure and disease
Ways that bias can be introduced
By the investigator, study participants, or the study design
Internal Validity
- There is no bias that is influencing
measurements, or - Bias is minimized
- If bias, confounding, and random error have been ruled out as alternative explanations for a study’s findings, the study is deemed internally valid
External Validity
- Internal validity must be established
- Results can be generalized beyond the study population
What are the two main types of bias in epidemiological research?
- Selection bias
2. Observation/information bias
Characteristics of bias (4)
- does not mean that the investigator is “prejudiced.”
- can arise in all study types: experimental, cohort, case-control
- occurs in the design and conduct of a study.
- can be evaluated but not fixed in the analysis phase
Selection Bias
- An error that results from procedures used
to select subjects and from factors that
influence participation in the study - Occurs when procedures used to select
study subjects lead to a result different
from what would have been obtained from
the entire population targeted for the study
In what types of studies is selection bias most likely to occur and why? (2)
In case-control or retrospective cohort studies because the exposure and outcome have already occurred at the time the subjects are selected
Where else can selection bias occur?
- In prospective cohort and experimental studies from differential loss to follow-up because this has an impact on which subjects are "selected" for analysis \+ Volunteer and non-response bias (Stigma Study) \+ Hospital patient bias \+ Healthy worker effect
Differential loss to follow-up
Drop-out related to outcome status and exposure status
Solutions to Selection Bias
- Use same criteria for selecting cases and controls
- Obtain all relevant subject records
- Obtain high participation rates
- Take into account diagnostic and referral patterns of disease
Observation Bias
An error that arises from systematic differences in the way information on exposure on or disease is obtained from the study groups
Characteristics of Observation Bias
- Results in participants who are incorrectly classified as either exposed or unexposed or as diseased or not diseased
- Occurs after the subjects have entered the study
Types of Observation Bias (4)
- Recall bias
- Interviewer bias
- Differential misclassification bias
- Non-Differential misclassification bias
Recall Bias
A type of information bias in which there are differences in the levels of accuracy in the information provided by compared groups
Possible result of recall bias
Can result in over-or-under-estimate of the measure of association
Ways to address recall bias
- Use controls who are themselves sick
- Use standardized questionnaires that
obtain complete information - Mask subjects to study hypothesis
Interviewer Bias
A systematic difference in soliciting, recording, or interpreting information that occurs in studies using in-person or telephone interviews
In what situations is interviewer bias likely to occur?
- Whenever exposure information is sought when the outcome is known, or
- When outcome information is sought when exposure is known
Ways to address interviewer bias
- Mask interviewers to study hypothesis and disease, or exposure status of subjects
- Use standardized methods of outcome (or exposure) ascertainment
Misclassification Bias (Measurement Error)
Subject’s exposure or disease status is erroneously classified
Types of Misclassification Bias
- Non-Differential misclassification bias
2. Differential misclassification bias
Non-Differential Misclassification Bias
- inaccuracies with respect to disease classification are independent of exposure, or
- inaccuracies with respect to exposure are independent of disease
Problem caused by Non-Differential Misclassification Bias
Will bias towards the null if the exposure has two categories; makes the groups more similar
Ways to Address Non-Differential Misclassification Bias
- Use multiple measurements
2. Most accurate source of information
Random Error
the probability that the observed result is due to “chance,” an uncontrollable force that seems to have no assignable cause
Loss to Follow-Up
A type of selection bias that occurs when subjects whose outcomes are unknown drop out of a cohort study because they can no longer be located or because they no longer want to participate
Healthy Worker Effect
A type of selection bias that occurs when the general population is chosen as a comparison (control) group in an occupational cohort study
In what types of studies can recall bias occur?
1) In a cohort study if cases are more or less likely than
controls to recall and report prior exposures
2) In a cohort study if exposed subjects are more or
less likely than unexposed subjects to recall and
report subsequent diseases
Control Selection Bias
Selection of an inappropriate control group in a case-control study
Self-Selection Bias
1) An error occurring in a case-control study arising from
refusal or nonresponse by participants related to both
exposure and disease, or
2) Agreement to participate that is related to both the
exposure and disease
Differential Surveillance, Diagnosis, or Referral
A type of selection bias that can result from a tendency to hospitalize patients differentially based on their exposure status