DATA ANALYSIS

Question 1

Prevalence =

Answer 1

No. Existing Cases / Total Population

Question 2

Incidence =

Answer 2

No. New Cases / Total Population at risk

Question 3

Prevalence is useful to define the burden of disease and health service needs, what is incidence useful for?

Answer 3

-to study disease trends/causes

Question 4

What 2 things do prevalence depend on?

Answer 4

• recovery

- death

Question 5

Survival analysis is used to analyse “time to event data” to estimates what?

Answer 5

Survival Probabilities

Question 6

What is censored data in terms of survival data?

Answer 6

-not all patients will have had the “event of interest” so their time of survival can only be known up to a point

Question 7

How is the cumulative probability in terms of survival data calculated?

Answer 7

• survival prob yr 1 X survival prob yr2…X yr3

e. g. 0.9 x 1 x 0.66

Question 8

What plot can cumulative survival prob be plotted on, with points joined in a step-wise manner? Used for?

Answer 8

Kaplan-Meir Plot

-estimate survival at any point in time

Question 9

What test is used to see if survival in one group (e.g. dementia) is statistically different to another (e.g. no dementia) ?

Answer 9

Log Rank Test

Question 10

If we want to estimate the RISK DIFFERENCE between 2 groups e.g. dementia vs no dementia, what test do we use? (produces a hazard ratio)

Answer 10

Cox Regression Test

Question 11

What is a Hazard Ratio?..the relative risk of..

Answer 11

*the relative risk of an endpoint occuring at any time over follow up

Question 12

Hazard ratio of 1 means what? 1+?

Answer 12

1=no difference in the relative risk of an endpoint occuring over follow up between 2 groups
1+ = increased risk of the endpoint vs other group

Question 13

Interpret a hazard ratio of 1.66 for dementia vs. no dementia over 3 yr follow up looking at deaths (endpoint)

Answer 13

-dementia group had 66% increase in mortality risk at any time in the 3yrs