Sources Of Variation

Question 1

Define INCIDENCE

Answer 1

Number of new cases of a disease within a population over a period of time (e.g. 1 year)

Question 2

Define PREVALENCE

Answer 2

The number of existing cases of a disease within a population

Question 2

What is the cause of variation within an epidemiological study?

Answer 2

There is a difference between the ‘observed’ and the ‘actual’ values

Question 4

How do you calculate the incidence of a disease within a population?

Answer 4

• No.of cases / (population x time period)

- x1000 to convert to per 1000 p-yrs

Question 5

How could you calculate the prevalence of a disease within a population?

Answer 5

• No. of existing cases / total population

- x1000 to convert to per 1000 p-yrs

Question 6

What is an incidence rate ratio (IRR) and how is it calculated?

Answer 6

• Relative risk of outcome between 2 populations

- IRR = observed incidence A / observed incidence B

Question 7

What does an IRR of 1 tell you?

Answer 7

• There is no difference between the incidences of the 2 populations (i.e. they have the same incidence)
• In this case the null hypothesis would be 0

Question 8

How do you calculate the Standardised Mortality Rate (SMR)?

Answer 8

No. of observed deaths within sample population / No. of expected deaths within reference population

Question 9

How do you calculate the error factor for an IRR?

Answer 9

e.f. = e^2 x ✔️(1/d1 + 1/d2)

Question 10

How does an SMR differ from an IRR?

Answer 10

SMR (standardised mortality ratio) takes into account confounding factors to produce a comparative value between the mortality rate of a sample population and the general population

Question 12

What is the 95% confidence interval?

Answer 12

Range within which we can be 95% sure that the ‘true’ value of the quantity of interest really lies

Question 14

When calculating the error factor for an IRR, what are d1 and d2?

Answer 14

• d1 is the observed number of cases in population A

- d2 is the observed number of cases in population B

Question 16

How does incidence rate differ from exposure?

Answer 16

Incidence rate includes ‘per unit time’ but exposure does not

Question 17

What is the NULL HYPOTHESIS?

Answer 17

• The null hypothesis states that there is no correlation between the 2 observed groups and that any relationships are due to chance
• Null hypothesis for a ratio = 1
• Null hypothesis for a difference = 0
• Null hypothesis for SMR = 100 (or 1)

Question 18

What is a CONFOUNDING FACTOR?

Answer 18

• Associated with both the outcome and exposure of interest
• NOT the causal pathway between exposure and outcome
• Examples include Age, Sex, Ethnicity and Socioeconomic background

Question 19

Interpret a 95% CI of 0.35 to 0.95 with an IRR of 0.6 for a study observing the incidence of infant mortality within the UK and an LEDC

Answer 19

• Observed incidence is that you are 40% less likely to experience infant mortality in the UK than in a LEDC
• True value for infant mortality in UK is likely to be between 5-65% less than that of infant mortality of LEDC
• We can reject the null hypothesis as it does not lie within the 95% CI so the data is statistically significant (p

Question 20

What is the significance of the p-value?

Answer 20

• States the PROBABILITY due to chance

- If p0.05 then the data is NOT statistically significant and we cannot reject the null hypothesis

Question 22

How do you calculate the 95% confidence interval?

Answer 22

• Upper limit = Observed IRR x e.f.

- Lower limit = Observed IRR / e.f.