Bias and Confounding

Question 1

Define bias

Answer 1

A systematic error in the design, conduct or analysis of a study that results in deviation from the truth.

Question 2

Define Selection Bias

Answer 2

An error due to systematic differences between those who are selected into a study and those who are not

OR

An error due to systematic differences in characteristic between those who are selected in one group within an analytic study.

Means study pop cannot be compared to the general pop.

Question 3

Define sampling bias. What type is it?

Answer 3

A systematic error where each potential member of the population doesn’t have an equal chance of being selected (a selection bias)

Question 4

Define allocation bias. What type is it?

Answer 4

In RCTs - a bias as a result of a systematic difference in the characteristics between those in one trial and the other arm. Participants don’t have the same probability of entering each branch of the trial. (a selection bias)

Question 5

Define spectrum bias. What type is it?

Answer 5

In DTA studies
Specific groups are excluded inappropriately - may be difficult to diagnose.
Can make index test appear more accurate. (a selection bias)

Question 6

Define responded/response/participation bias. What type is it?

Answer 6

A systematic difference due to difference in characteristics between those who choose to volunteer or choose one study group vs the other. (a selection bias)

Question 7

Define loss to follow up/follow up/attrition bias. What type is it?

Answer 7

Systematic error due to differences in characteristics of loss to follow up vs not.
E.g. a RCT where elderly pts are particularly affected by drug side effects. (a selection bias)

Question 8

Define health worker effect bias. What type is it?

Answer 8

Less mortality and morbidity when workers are compared to the general population. (a selection bias)

Question 9

Define Information/Observation/Measurement Bias

Answer 9

A systematic error in the way that data is gathered from the comparison groups that results in differences in the quality of information between the comparator groups.
Bias from the measurers or those being measured

Question 10

Define interviewer bias. What type is it?

Answer 10

Error due to the interviewers subconscious/conscious gathering of data that differs systematically between controls and cases (e.g may probe more for exposure if not properly blinded) (measurement bias)

Question 11

Define recall bias. What type is it?

Answer 11

Systematic error due to differences in accuracy or completeness of recall of prior exposure of interest between cases and controls. (measurement bias)

Question 12

Define recording bias. What type is it?

Answer 12

Error when the data of interest is systematically recorded in more detail in the cases vs controls - seen also in self-completed questionnaires. (measurement bias)

Question 13

Define social acceptability bias. What type is it?

Answer 13

Selective suppression or revealing of information most socially acceptable answer is given. Can reduce by anonymising questionnaires. (measurement bias)

Question 14

Define detection bias. What type is it?

Answer 14

Systematic differences between groups in how outcomes are determined. Blind assessors to the intervention reduces outcome based measures. (think the hFABP test) (measurement bias)

Question 15

Define performance bias. What type is it?

Answer 15

Systematic differences between groups in the care that is provided or in exposure to factors other than the intervention of interest - e.g. more attention in the case group (measurement bias)

Question 16

Define reporting bias. What type is it?

Answer 16

Systematic difference in the reported findings and unreported findings - so may report good ones in cases. Basically cherry picking. (measurement bias)

Question 17

Define verification bias. What type is it?

Answer 17

DTA studies, not all participants receive the index test and reference standard (seen when the index test is negative) (measurement bias)

Question 18

Define review bias. What type is it?

Answer 18

DTA studies where the index test is not interpreted independently of the reference standard (measurement bias). Index test will appear more accurate (measurement bias)