PASS sem 2 MMF slides

Question 1

what are the four types of observational studies?

Answer 1

ecological studies
cross sectional studies
case control
cohort

Question 2

ecological study

Answer 2

Find groups in population
• Choose exposure and outcome
• Count cases NOW

Question 3

Cross-sectional Studies

Answer 3

• Number of people with the disease in a population at a certain time

(remember: a slice thru pop and finding n.o people with disease)

Question 4

problems with case control studies

Answer 4

• selection bias (representative and same community)
• recall bias
• misclassification
• interviewer bias
• confounding

Question 5

what is response bias?

Answer 5

tendency of individuals to respond inaccurately/ falsely

> problem of cohort study
also called survey bias

Question 6

Cross sectional studies problems

Answer 6

• selection bias (representative and same community)
• response bias (honesty)
• measurement bias (whether its measured in same way)
• confounding

Question 7

case control study

Answer 7

Find people with the disease NOW
• Find people with out the disease NOW
• Always look BACKWARDS to see what exposures they’ve had
• Find ASSOCIATION

(remember: those with the CASE now e.g. disease now, looking backwards to better times)

Question 8

cohort study

Answer 8

Find people have been exposed NOW
• Find people haven’t been exposed NOW
• Follow up FORWARDS
• Can be done in the past

(remember: medicine cohort is moving forwards in time thru degree)
- they don’t have the disease. just looking at the exposed or unexposed and seeing how they turn out!

Question 9

prevalence

Answer 9

proportion of people with the disease

snapshot in time

the bathtub in bath analogy > incidence is the tap adding water in > plug leaking out is death or cure.

Question 10

how does the analysis of case control study vs cohort study differ?

Answer 10

case control- odds ratio only (as its one point in time)

cohort- rate OR odds ratio (following a cohort over a period of time so can distinguish rate and the odds ratio at a specific time)

Question 11

what are the two types of experimental studies?

Answer 11

controlled trial
- people are allocated interventions based on methods that are not random.

randomised control trial
> use two groups that are exactly the same e.g. old treatment vs new treatment.

Question 12

what is the best study/ trial (hierarchy of evidence)?

Answer 12

systematic review of randomised trials = best

consensus/ expert views= worst

observational in the middle

Question 13

what is the method for describing or finding a study?

Answer 13

PICOS

• Population
• Intervention or exposure
• Comparison/control of interest
• Outcome
• Study design>how are the results going to be analysed?

Question 14

three requirements of a sample

Answer 14

representative- same pop
unbiased-
precise- values close together

Question 15

define bias (and two main types)

Answer 15

difference between true value and expected value

two types-
selection bias
information bias

Question 16

odds ratio equation

Answer 16

probability of event happening / probability of event not happening

Question 17

confounding factors definition

Answer 17

affects both the exposure and the outcome

e.g. alcohol is confounding factor as effects risk of throat cancer and men drink more alcohol

Question 18

what could effect the validity of an experiment?

Answer 18

• selection bias > not representative of population being studied
• how the sample was chosen
• data collection so whether it was self reported/ how many dropped out/ whether it relies on recall.
• confounding factors adjusted for?
• sample big enough to rule out chance

Question 19

what is survivor bias?

Answer 19

loss of follow up

seen in prospective cohort study

Question 20

what is observation bias ?

Answer 20

patients performing their own measurements at home could be inaccurate

seen in prospective cohort study