Biases and other systematic errors

Question 1

What is healthy entrant bias?

Answer 1

People who have the disease of interest, or symptoms indicative of the start of the disease, are often excluded from cohorts at recruitment
 Participants are biased towards healthy individuals
 Initial incidence of outcome therefore might be lower in the cohort than the general population
 The disease experiences of the cohort and that of the general population may well not be comparable

Question 2

What are the two main types of bias?

Answer 2

Selection bias and information bias

Question 3

What kinds of study are associated with the healthy entrant bias?

Answer 3

Cohort studies

Question 4

What is non-response bias?

Answer 4

Non-response bias can occur if those who participated in the study were different from those who did not. Important to note that this is about PARTICIPATION and not SELECTION.

Question 5

What kinds of study design are associated with non-response bias?

Answer 5

Cohort studies

Question 6

What is the kind of systematic error most closely associated with cohort studies, and why is it important in the context of the study?

Answer 6

Loss to follow up. It is important bc if there are high rates of loss to follow up it raises serious questions about the validity of findings. This is amplified if there are differential rates in where the loss to follow up occurs - amongst cases or controls.

Question 7

What kinds of systematic error are most closely associated with case control studies?

Answer 7

confounding, reverse causality, information bias (especially recall and observer), selection bias (especially with controls and/or matches)

Question 8

What is an “ecological fallacy”?

Answer 8

The concept that associations found on a group level (be it community, neighbourhood, population) may not be consistent with associations found on an individual level

Question 9

What kind of systematic error are cross sectional studies particularly liable to, and why is it important?

Answer 9

Selection bias; if the sample that is chosen is not representative of the target population, then the results will not be generalizable. Reverse causality; because we collect data on exposure and outcome at the same time, it can be difficult to establish whether the exposure occurred before or after the outcome. Information bias; “recall bias” - when there is a systematic difference in the way different groups recall events (eg exposures)

Question 10

Within cross section studies in particular, what is the only time that we can be sure that the exposure came before the outcome?

Answer 10

If the exposure is a fixed characteristic of an individual - eg their sex, blood type

Question 11

What kinds of systematic error are RCTs liable to? Address and explain

Answer 11

Loss to follow up if there is not proper tracing involved. Information bias if: 1) participants not blinded (performance bias) 2) assessors not blinded (observer bias and/or interviewers not blinded). And finally, selection bias (allocation bias) if participants are not properly randomized

Question 12

What is the prevention strategy for observer bias?

Answer 12

The masking of participants from the identity of exposures

Question 13

What is a prevention strategy for interviewer bias?

Answer 13

Concealment of identity of exposures from interviewers

Question 14

What is a prevention strategy for recall bias?

Answer 14

Collection of data from medical records (or another objective source of information)

Question 15

What is a prevention strategy for reporting bias?

Answer 15

The masking or blinding of participants

Question 16

What is a prevention strategy for performance bias?

Answer 16

The masking or blinding of assessments

Question 17

What is a prevention strategy for detection bias?

Answer 17

The masking or blinding of assessments

Question 18

What is a prevention strategy for sampling bias?

Answer 18

Avoid using volunteers and use rigorous inclusion criteria

Question 19

What is a prevention strategy for reverse causality?

Answer 19

Exclude events occurring in the initial years after baseline

Question 20

What are prevention strategies for confounding?

Answer 20

At the design stage, using matching, and at the analysis stage use logistic regression

Question 21

What is detection bias?

Answer 21

When the way in which outcome information is collected differs between groups

Question 22

What is observer bias?

Answer 22

when the investigator is aware of the disease status, treatment group or outcome of the subject and their ability to interview the subject, collect or analyse the data in an unbiased manner is compromised

Question 23

What is reporting bias?

Answer 23

When participants “collaborate” with researchers and give answers in the direction they perceive are of interest

Question 24

What is interviewer bias?

Answer 24

When an interviewer takes more detailed notes/more attention towards cases than controls in a way that means unequal levels of data have been collected for both groups that calls into question the validity of findings

Question 25

What is allocation bias?

Answer 25

if investigators know or predict which intervention the next eligible participant is supposed to receive. This knowledge may influence the way investigators approach potentially eligible participants and how they are assigned to the different groups, thereby selecting participants with good prognoses

Question 26

What is loss to follow up?

Answer 26

when losses/withdrawals are uneven in both the exposure and outcome categories. The validity of the statistical results may be affected. Refers specifically to participants not being followed for the proper duration of the study/intervention.

Question 27

How can allocation bias be prevented?

Answer 27

Clear inclusion criteria and proper randomization

Question 28

How can loss to follow up be prevented?

Answer 28

Using tracing methods to minimize loss to follow up (some is unavoidable(