Lecture 6 and 7: Evidence for disease prevention

Question 1

What is a descriptive study?

Answer 1

Describe cases (those with the disease) 
Describe population (persons in whom cases arise) 
Does not explain causality, used to generate a hypothesis.

Question 2

What is the purpose of a descriptive study?

Answer 2

To know how much disease a region or group has and to formulate a worthy hypothesis

Question 3

what is the counterfactual; method

Answer 3

what would have happened if??? would the disease have happened in the same person at the same time? in order to attempt to prove a cause of a disease.

Question 4

Method 0:

What is an anecdote/case series?

Answer 4

Anedotal
No control/comparison group
cant be used in analytical epidemiology as there is no counterfactual group

Question 5

Method 1:What is a cross sectional survey?

Answer 5

Snapshot of people with an outcome
Not easy to have a counterfactual
looking at what is happening at the time/now

Question 6

What is method 2 measuring incidence?

Answer 6

When new cases arise

Can be used in ecological studies

Question 7

What is method 3 ecological study?

Answer 7

Interested in groups not individuals e.g country level of cancer plotted against fat intake
Observational study-can not manipulate anything
Unit of observation is a group

Question 8

What is method 4 case control studies?

Answer 8

Cases- have the disease
Controls - don’t have the disease
Ask about exposure
Can calculate odds ratio
Good representation of relative risk
Common disease, cant use odds ratio as it doesn’t properly represent risk
o Doesn’t represent a standard counterfactual – roughly the non-exposed groups in both the ‘cases’ and ‘controls’ groups

Question 9

What is method 5 cohort study?

Answer 9

Followed through time for the outcome
Nobody has the disease at the start
Measured exposure to plit cohort into exposed and non exposed- follow groups through time
Measure rate of outcome in each group
Counterfactual- people who are not exposed

Question 10

What is method 6 randomised controlled trial?

Answer 10

Seen as gold standard
Healthy people for preventative trial, sick people for therapeutic trial
randomised as to whether you get treatment or not
follow groups and measure rates of outcome

Question 11

What are the +ve of method 6?

Answer 11

strongest evidence for causality
if randomised, selection bias and confounding is removed
if blinded, less observer bias

Question 12

What are the -ve of method 6?

Answer 12

Not real life, high cost, inappropriate/unethical for many research questions

Question 13

What is blinding?

Answer 13

participants are not aware of which group they are in

Question 14

What is intention to treat analysis?

Answer 14

At the start of a RCT people are assigned to treatment or control group, data analysed as if they took their treatment even if they withdrew = represents real life

Question 15

What is error?

Answer 15

difference between estimate/measure value and the true value- impossible to not have error

Question 16

What is selection bias?

Answer 16

Arises from sample selection, the ignored should be samples e.g prisons, hospitals, babies

Question 17

What is diagnostic bias?

Answer 17

occurs when made on basis of exposure, make assumptions e.g. doctor may diagnose a mesothelioma because they’ve been exposed to asbestos when it may actually be something else

Question 18

What is self selection bias?

Answer 18

may come about when subjects make decisions- when patients come forward wanting to be part of a study

Question 19

What is participation bias?

Answer 19

Cases have a clear motivation for participating as they have disease

Question 20

What is information bias?

Answer 20

error in the information collected, wrong info on right group of people

Question 21

What is recall bias?

Answer 21

errors in recall of information, e.g. mothers of children with cancer

Question 22

What is interview bias?

Answer 22

Participants say what they think the interviewer want to hear

Question 23

What is surrogate bias?

Answer 23

have to ask a surrogate about an individual e.g. baby, dead person
May not want to answer difficult questions

Question 24

What is confounding variable?

Answer 24

variables that the researcher fails to control or eliminate, damaging the internal validity of an experiment

Question 25

Cause of a disease is?

Answer 25

not provable