Public Health ESA1

Question 1

How does Epidemiology compare to Public health?

Answer 1

Epidemiology is the cause

Public Health is the consequence

Question 2

Whats the purpose of statistics?

Answer 2

To generalise/infer about, the population from a sample

Question 3

What is a sampling distribution?

Answer 3

PROBABILITY DISTRIBUTION of a number of samples. Gives you a bell shaped curve

Question 4

Whats interesting about means of the samples?

Answer 4

No matter what the population distribution .II.I.. is, the sampling distribution is always the same ..iIIi..(it will always form a bell shaped plot as you add more and more sample means)

Question 5

Whats a confidence interval?

Answer 5

The range between which you can be confident the population mean will fall

Question 6

The 95% confidence interval is wider if…?

Answer 6

1) Small sample size

2) Greater variation in the population

Question 7

What is confidence intervals a estimate of?

Answer 7

PRECISION

Question 8

Whats the difference between precision and bias?

Answer 8

Unbiased means your looking at the right target

Precision is how close to THE target (not right target) you are

Question 9

What is Selection Bias (external validity)?

Answer 9

Choosing participants from a sampling frame that is not representative of the general population

Question 10

What is Selection Bias (internal validity)?

Answer 10

ONE of the groups being studied is not representative

of the sampling frame from which it was drawn

Question 11

What type of bias? respondents to a survey with a low response rate are unlikely to be representative of the original invitees

Answer 11

Selection Bias – Leading to Errors in Comparability (Internal Validity)

Question 12

What type of bias? a study comparing the incidence of a disease
in an occupational group (exposed) with the general
population (unexposed) is more likely to find the
occupational group to be healthier than if it was
compared with another occupational group

Answer 12

Selection Bias – Leading to Errors in Comparability (Internal Validity)

Question 13

What type of bias? participants remaining at the end of follow-up in a study with high losses to follow-up are unlikely to be representative of the original participants in that
group (“Loss to Follow-Up” or “Attrition Rate”)

Answer 13

election Bias – Leading to Errors in Comparability (Internal Validity)

Question 14

What type of bias? a study comparing diabetic hospital patients (cases) with healthy GP practice patients (controls) is more likely to find smoking is associated with diabetes than a study comparing diabetic GP practice patients with healthy GP practice patients

Answer 14

Selection Bias – Leading to Errors in Comparability (Internal Validity)

Question 15

What type of bias? a study comparing hospitalised liver cirrhosis patients (cases) with other hospitalised patients (controls) is less likely to find that alcohol is
associated with liver cirrhosis than if compared with patients from the same GP practices as the
hospitalised liver cirrhosis patients

Answer 15

election Bias – Leading to Errors in Comparability (Internal Validity)

Question 16

What type of bias? a lifestyle survey of Sexual Health clinic patients is
unlikely to be representative of the general population

Answer 16

Selection Bias – Leading to Errors in Generalisability (External Validity)

Question 17

What type of bias? a study of the effect of socio-economic status
(exposure) on incidence of heart disease (outcome)
amongst fitness gym members is unlikely to be
representative of the general population

Answer 17

Selection Bias – Leading to Errors in Generalisability (External Validity)

Question 18

What is information bias?

Answer 18

bias arising from collecting information

Question 19

What are 4 example of information bias?

Answer 19

MORM
M.easurement error
O.bserver error
R.ecall error
M.is-classification

Question 20

What are two ways of stopping confounding?

Answer 20

direct standardisation → Directly Standardised Rates

Indirect standardisation → Standardised Mortality Ratios

Question 21

What are the three types of observational studies?

Answer 21

Cohort
Case controlled
Cross sectional

Question 22

Whats the difference between a cohort and a case controlled study?

Answer 22

Cohorts are usually prospective start with a patients and record the outcome over time. Case controlled are retropective

Question 23

Whats a cohort study?

Answer 23

Start with an exposure, get a non-exposed control group and monitor the outcome. Prospective study

Question 24

Whats a case controlled study?

Answer 24

Start with an outcome or disease and compare with a control group to find a retropective exposure

Question 25

Advantages of Case controlled study? 3

Answer 25

Quick
Can look at rare OUTCOMES
Look for multiple exposures

Question 26

Advantages (2) and Disadvantages (3) of Cohort study?

Answer 26

• Stonger evidence than CC and CSA
• (RARE) Single Exposure, multiple outcomes

SLOW
EXPENSIVE
Cant deduce different exposures

Question 27

What do Case control and cohort studies measure?

Answer 27

Incidence. Rate of disease

Question 28

What do crossectional analysis measure?

Answer 28

Prevalence

Question 29

What’s a cross sectional study?

Answer 29

• Meaures prevalence (portion of population that has) exposure and disease
• Looks at a single point in time
• Divide a population into expose/non-expose groups. Compare the proportion of prevalence is each group

Question 30

Advantages (3) and Disadvantages (3) of cross sectional study?

Answer 30

• Quick and Fast
• Cheaper
• used for making hypotheses

Not strong enough to show a casuality (because of not time variable). But can suggest.