The Role of Bias

Which way does the bias possibly influence the estimated effect? Some biases may cause the study to overestimate the real risk, this often happens when...

'sicker than average' subjects are used

What are 4 sources of bias?

1. Can be caused by investigator or study participate during design or conduct of study 2. Can occur in experimental cohort, cross-sectional, and ecological studies 3. Not inherent characteristics of the study population 4. Few studies have no bias or errors

What are 2 ways you could reduce selection bias among controls?

1. Compare the prevalence of the exposure with other sources to evaluate credibility 2. Attempt to draw controls from a variety of sources - hospital based - general population - "special relationship"

When interpreting study results, what 3 questions should you ask yourself?

1. Consider the type of study 2. See what type of discussion or explanation is provided about how potential biases were anticipated and avoided by design (no mention is a 'red flag') 3. Very important to determine if study investigators assessed the impact of any potential biases and how these might have affected the estimation of the risk or benefit observed between exposure and disease

What are 2 effects of bias?

1. Create appearance of association when there is none, or mask an association that really exists 2. Selection bias and information bias cannot be fixed in the analysis; confounding can be fixed but only to a point

What are 3 techniques to reduce selection bias?

1. Develop an explicit participant definition 2. Enroll all participants in a defined time and region 3. Strive for high participation rates

What are 3 ways you could reduce the selection bias among cases?

1. Ensure that all medical facilities are thoroughly canvased 2. Develop an effective system for case ascertainment 3. Consider whether all cases require medical attention; consider possible strategies to identify where else the cases might be ascertained

What are two major sources for systematic classification (i.e. bias)?

1. Errors introduced during selection of study subjects 2. Errors occurring during ascertainment of information

List 2 examples of selection bias.

1. Healthy workers in occupational studies; sicker or healthier individuals than others with the disease due to referral techniques 2. Respondents to the Iowa Women's Healthy Study were younger, weighed less, and were more likely to live in rural, less affluent counties than nonrespondents

What are 3 types of selection bias and what studies are they associated with?

1. Inappropriate control selection (control-selection bias) - case-control 2. Differential participation - case-control or cohort 3. Differential loss to follow-up - cohort or experimental

What are 3 solutions to bias?

1. Limit in study design 2. Limit in study conduct (during data collection) 3. Critically evaluate after study has been completed, and discuss its effects

What are 3 ways to reduce information bias?

1. Little (or nothing) can be done to fix information bias once it has occurred 2. Information bias must be avoided through careful study design and conduct 3. Information bias cannot be "controlled" in the analysis

What are 2 solutions to prevent interviewer bias?

1. Mask interviewers to study hypothesis and/or disease or exposure status of subjects 2. Use standardized methods for outcomes (or exposure) ascertainment

What 4 factors contribute to systematic errors?

1. Misclassification 2. Selection bias 3. Information or measurement bias (also called observation bias) 4. Confounding

What are 2 effects of measurement (misclassification) error?

1. Non-differential - bias toward the null 2. Differential - bias towards or away from null

What are 2 solutions to selection bias?

1. Once a selection bias occurs in a study, no amount of analysis will correct it - nothing can fix its impact and influence on a study's result 2. Researchers need to avoid it when planning and conducting their studies

What are 2 other forms of information bias?

1. Prevarication bias (lying) 2. Difficult or poorly understood outcome definitions ("chronic fatigue syndrome", "GI upset", "head stinger", etc)

What are the 2 types of measurement bias?

1. Recall bias 2. Interviewer bias

What are 3 types of information bias?

1. Recall bias 2. Interviewer bias 3. Measurement error

State how misclassification can be minimized.

Use the most accurate source of information, and use sensitive and specific criteria to define the exposure and disease.

When does measurement bias occur?

after the subjects have entered the study

Defintion: Systematic error in design or conduct of study that leads to incorrect estimate of association

bias

Definition: Deviation of results or inferences from the truth, or processes leading to such deviation Any trend in the collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are systematically different from the truth (as opposed to non-systematic or random error, i.e. chance)

bias (systematic errors)

Retrospective studies are at a greater risk than prospective studies because...

both exposure and disease have already occurred

State how selection bias can be minimized.

Ensure that selection of cases and controls is independent of exposure (in a case-control study) and that selection of exposed and unexposed groups is independent of outcome (in a retrospective cohort study), and obtain high follow-up and participation rates (all types of studies).

Differential misclassification: Inaccuracies with respect to disease classification are _____ of exposure.

dependent

Differential misclassification: Inaccuracies with respect to exposure are _____ of disease.

dependent

Selection bias arises when the relation between exposure and disease is...

different for those who participate and those who theoretically would be eligible for study but do not participate

Describe information bias within a cohort study.

different procedures are used to collect information from exposed and unexposed groups

Describe information bias within a case-control study.

different techniques are used to collect information from cases and controls

Defintion: Bias that can occur if study participants exit a study for reasons related to both exposure and disease

differential loss to follow-up

Which type of bias? Toward or away from the null Since outcome cannot be known without good follow-up, must maintain high participation rates

differential loss to follow-up

What is being described below? May bias towards or away from null, dependent upon direction of bias. May be impossible to determine judge impact.

differential misclassification

What is the worst type of bias that can occur for a study? Why?

differential misclassification because it's impact may not be understood nor accessible

Defintion: Bias that can occur if willingness or ability to participate is related to both exposure and disease status

differential participation

Which type of bias? Toward or away from the null Obtain high participation rates for all groups

differential participation

Hospital-based control selection: Advantages?

easily identified & available more likely to have considered life events (reduce recall bias) more likely to cooperate

General population control: Disadvantages?

expensive to find poor cooperation likely ("don't care about topic") those who do volunteer and cooperate (can be reached) may vary systematically from the "norm" quality of life recall might vary from cases (haven't been sick so having spent time wondering "why me")

Describe recall bias within a retrospective cohort study.

exposed participants are more or less likely to recall prior disease than unexposed participants

Interviewer bias can occur when...

exposure information is sought when outcome is known (as in case-control), or when outcome information is sought when exposure is known (as in cohort study) example: investigators "probe" for information differentially between study groups

True or false: Bias is introduced primarily during the analysis stage of a study

false

True or false: Including a large sample size reduces self-selection bias

false

True or false: It is possible to fix or adjust bias during the analysis phase of a study

false

True or false: Open-ended interview questions are the best way to ascertain exposure information

false

True or false: Poor recall and recall bias are synonymous terms for the same concept

false

True or false: Using an inaccurate case definition increases the likelihood of nondifferential misclassification of the disease

false

Main source of controls? Selected in a variety of ways: canvassing random-digit dialing, voter registration lists, live in the same area (census tract identification), etc.

general population

General population control: Advantages?

gets away from concern with "sick role" habit changes useful especially if trying to assess a local environmental exposure truly represents someone that would have shown up as a case if sick

Because of drawbacks to each type of possible control, one of the best things to do is...

have multiple control groups

Another selection concern in any cohort study is...

healthy worker effect

Definition: Bias that can occur if people with disease remember/report their exposure differently

recall bias

Definition: People with disease (cases) remember or report exposures differently than those without disease (controls)

recall bias

Defintion: Occurs when the level of accuracy differs between the compared groups Occurs in a case-control study when cases remember or report their exposures differently (more or less accurately) from controls It occurs in a retrospective cohort study when individuals who are exposed remember or report illnesses differently than those who are unexposed

recall bias

Which type of bias? Toward or away from the null Use sick controls Quality questionnaires Use pre-existing data

recall bias

Selection bias occurs when selection of exposed and unexposed subjects is not independent of outcome; therefore, it can only occur in...

retrospective cohort studies

Definition: Error due to differences in the characteristics between those who are selected for a study and those who are not

selection bias

Defintion: Bias that can occur if people who agree to take part in a study are different from the source population you want to study

selection bias

Describe selection bias within case-control studies.

selection/participation of cases and controls is related to exposure status

Describe selection bias within cohort/experimental study.

selection/participation of exposed and unexposed subjects is related to disease status

Main source of controls? Selecting friends, next-of-kin, siblings, spouses, neighbors or work colleagues

special control

True or false: A study must be valid before its results can be generalized

true

True or false: Bias can pull an estimate of association either toward or away from the null

true

True or false: Using incentives to ensure high rates of participation in a study will decrease selection bias

true

True or false: When a positive association is biased away from the null, the true association is overestimated

true

True or false: When a positive association is biased toward the null, the true association is underestimated

true

Special control: Advantages?

want to help due to association with case share similarities (diet, habits, socio-economic statues, etc)

Random misclassification may occur...

when a survey instrument has poor reliability or poor precision (also called non-differential misclassification)

When does selection bias in a case-control study occur?

when controls or cases are more (or less) likely to be included in the study if they have been exposed - that is, inclusion in study is not independent of exposure

Which way does the bias possibly influence the estimated effect? Worst case...

when the influence of the bias cannot be determined as to how it influences the observed relationship

Systematic of differential misclassification may occur...

when there exists one or more biases in the conduct of a study

Defintion: Occurs in occupational studies when disease and death rates in a working population are compared with those in the general population The rates of disease and death among workers are typically lower than those in the general population because there is a higher proportion of ill people in the general population

healthy worker effect

Main source of controls? People from the same hospital has the cases

hospital-based

List 2 potential results of selection bias in a case-control study?

1. Relationship between exposure and disease observed among study participants is different from relationship between exposure and disease in individuals who would have been eligible but were not included 2. The odds ratio from a study that suffers from selection bias will incorrectly represent the relationship between exposure and disease in the overall study population

What are 4 sources of measurement (misclassification) error?

1. Self-reports 2. Errors on medical records, death certificates, etc. 3. Errors in how data are captured 4. Non-specific disease or exposure definition

What are 4 solutions for interviewer bias?

1. Use blinding/masking to prevent interviewer from knowing study hypothesis or know knowing whether someone is exposed or not 2. Use high quality, standardized, closed-ended questionnaires to promote consistency and specificity, and use of appropriate response categories 3. Examine pre-existing data 4. Provide adequate and rigorous training for interviewers

What are 3 solutions to recall bias?

1. Use control who are, themselves, sick 2. Use standardized questionnaires 3. Mask subjects to study hypothesis

What are 3 solutions to recall bias?

1. Use controls who are also sick to promote comparable recall 2. Use standardized, closed-ended questionnaires to promote consistency and specificity 3. Examine pre-existing data

What are 5 techniques to reduce information bias?

1. Use memory aids with subjects 2. Validate exposure with other records or biomarkers 3. Provide standardized training sessions and protocols & use standardized data collection forms 4. Blind participants as to study goals and classification status 5. Blind interviewers

What are 3 solutions to improve accuracy of collected information involving measurement (misclassification)?

1. Use most accurate source of information available 2. Multiple measurements of exposure and disease 3. Validation - corroborate the data using several sources

Recall bias is most likely to occur in: a. Case-control studies b. Prospective cohort studies c. Experimental studies d. All of the above e. None of the above

A

A prospective cohort study was conducted to determine the risk of heart attack among men with varying levels of baldness. Third-year residents in dermatology conducted visual baldness assessments at the start of the study (which was before any heart attacks had occurred). Four levels of baldness were coded: none, minimal, moderate, and severe. The follow-up rate was close to 100%. Which of the following types of bias were surely avoided in this study? a. Recall bias of the exposure information b. Differential misclassification of the exposure c. Non-differential misclassification of the exposure d. Selection bias

A. Yes B. Yes C. No D. Yes - because it is a prospective cohort study with nearly 100% follow-up

Selection bias is most likely to occur in: a. Case-control studies b. Retrospective cohort studies c. Experimental studies d. Both retrospective cohort and case-control studies e. Both retrospective cohort and experimental studies

D

State the main difference between differential and non-differential misclassification, and state which direction(s) each type of error can bias the study results.

In non-differential misclassification, inaccuracies that occur on one axis (exposure or disease) are independent of the other axis. - For example, if there is an error in exposure misclassification, it occurs with equal likelihood among diseased and non-diseased individuals In differential misclassification, inaccuracies that occur on one axis (exposure or disease) are dependent on the other axis. - For example, if there is an error in exposure misclassification, it occurs more often in the case group than the control group Non-differential misclassification results toward the null Differential misclassification results toward or away from the null

Hospital-based control selection: What diseased group would make a suitable control group?

MAJOR RULE: the control group should not have a disease known or suspected to be associated with the disease of interest nor any of the risk factors associated with the disease of interest example: if a study dealt with leukemia cases, controls could be breast cancer cases, but shouldn't be bone cancer cases

State how interviewer bias can be minimized.

Mask interviewers to the study hypothesis and to the disease or exposure status of the study subjects, and carefully design the interview instrument.

State how recall bias can be minimized.

Mask study subjects to the study hypothesis, use diseased controls if conducting a case-control study, and carefully design an interview instrument.

Can biased be 'analyzed away'?

NO; although a good study should discuss the impact and direction of influence of any suspected bias

What is being described below? Will bias towards the null if the exposure has two categories. Makes the groups more similar and washes out differences.

Non-Differential misclassification:

What type of study contains both control-selection and differential participation?

case-control

Selection bias is more likely to occur in what two types of studies and why?

case-control and retrospective cohort studies because exposures and outcomes have already occurred by the time a subject is selected into a study

Describe recall bias within a case-control study.

cases are more or less likely to recall prior exposures than controls

What type of study contains both differential participation and differential loss to follow-up?

cohort

Which way does the bias possibly influence the estimated effect? _____ tend to dilute the ability to see any association, even if one truly exists. Real risk is probably underestimated Random errors cause such dilution

conservative bias - ones that make compared groups more similar

Which type of bias? Toward or away from the null Uses identical selection criteria for cases and controls (recall purpose of controls)

control-selection

Definition: A type of selection bias that occurs in case-control studies when the controls do not accurately represent the exposure distribution in the source population that produced the cases It occurs when different criteria are used to select cases and controls and these criteria are related to the exposure

control-selection bias

Definition: Bias that can occur if controls are more (or less) likely to be selected if they are exposed (or unexposed)

control-selection bias

Which type of bias? Occurs when controls fail to represent the exposure distribution in the source population from which the cases arose because controls do not accurately represent the same source population as the cases

control-selection bias

The result of bias may be an _____ or _____ estimate of the measure of association.

incorrect or invalid

Non-Differential misclassification: Inaccuracies with respect to disease classification are _____ of exposure.

independent

Non-Differential misclassification: Inaccuracies with respect to exposure are _____ of disease.

independent

Definition: Bias that can occur if the information you collect from or about study participants is erroneous

information bias

Definition: Systematic difference in soliciting, recording, interpreting information

interviewer bias

Defintion: Bias that can occur if there is a systematic difference in soliciting, recording, or interpreting information

interviewer bias

Which type of bias? Toward or away from the null Masking Quality questionnaires Train interviewers

interviewer bias

Describe interviewer bias within a case-control study.

interviewer is influenced by participants case or control status

Describe interviews bias within a cohort/experimental study.

interviewer is influenced by participants treatment or exposure status

Definition: Bias that can occur if study participants are placed into the wrong exposure of disease category

measurement (misclassification) error

Defintion: Refers to a systematic distortion in the accuracy of information generated from subjects with compared to without the exposure, or subject with or without the disease

measurement bias

Defintion: The erroneous classification of an individual, a value, or an attribute into a category other than that which it should belong

misclassification

Why are prospective studies more prone to bias?

more loss due to follow-up

Defintion: This is a type of measurement error which means that the extent of the misclassification is the same for both groups

non-differential misclassification

Which type of bias? Toward the null Accurate definition of E and D Accurate data sources Multiple measurements Validation Most common type of measurement error

non-differential misclassification

Special control: Disadvantages?

one of their shared similarities might actually be the cause of disease and these controls were "just lucky"

Recall bias can result in...

over or under estimate of measure of association

______ association is biased towards the null value. True association is underestimated.

positive or preventative

______ association is biased away from the null value. True association is overestimated.

postive or preventative

Hospital-based control selection: Disadvantages?

they are also ill and may differ in habits in ways that all ill folks do (i.e. some habits may vary due to "the sick role" & not represent what case folks were like before getting sick) not 'normal'