study designs and measures

Question 1

Define risk

Answer 1

Measure of probability of an event occurring in a defined population in a given period of time.
(events/ population at potential risk)

Question 2

Define risk difference

Answer 2

risk in exposed - risk in unexposed. Also known as attributable risk.

Question 3

Define relative risk

Answer 3

Relative risk is a ratio of the probability of an event occurring in the exposed group versus the probability of the event occurring in the non-exposed group.
2 measure: risk ratio and rate ratio.

Question 4

What is risk ratio?

Answer 4

Risk in exposed group / risk in control group (as proportions).

Interpret as: Exposed group have x times greater risk of …… compared to the control group.
RR of 1 = no difference in outcome between the groups.

Question 5

What is rate ratio?

Answer 5

Frequency of events in time period / number at risk of event during time period.

Interpreted like risk ratios,

Question 6

In what study do we use risk ratio and rate ratios?

Answer 6

Cohort and clinical trials

Question 7

Why can we not use risk ratios in case-control studies? How do we measure risk instead?

Answer 7

We do not know what the prevalence is relative to the population at risk.

Instead we use odds and compare groups using odds ratio.

Question 8

How do we calculate odds ratio?

Answer 8

Odds of exposure in cases / odds of exposure in controls (A/C) / (B/D)

Question 9

In general, how do we interpret odds ratios?

Answer 9

In the case group, the odds of exposure to suspected RF is … times greater than than the odds in the controls group. This appears strong evidence of association.

Question 10

What does an OR of 1.0 mean?

Answer 10

An odds ratio of 1.0 means that exposure to the RF is equal in both the case and control groups.

Question 11

How do you calculate the confidence interval for a risk or rate ratio?

Answer 11

1. lnRR
2. Standard error
3. Margin of error (1.96 x SE)
4. Confidence interval (RR -ME) to (RR + ME)
5. e^x

Question 12

What types of data do we use chi squared for?

Answer 12

Categorical data

Question 13

When is chi squared used?

Answer 13

For testing hypotheses comparing groups to see whether there is any association between variables.

Question 14

How does chi square test work>

Answer 14

You are comparing what you observe to an idealised prediction model and to see how much our observations differ from our predicted model.

Question 15

When do you use Fisher’s exact test?

Answer 15

Instead of chi squared, when samples are small. 80% of expected values should be greater than 5 and none less than 1.

Question 16

What is the p-value?

Answer 16

The probability that the null hypothesis is correct.

Question 17

What is a cohort study?

Answer 17

Two groups, one exposed to RF and one not. Followed over time, see the outcome. See if a RF is causally related to health. Studies development of a disease.

Question 18

What types of research questions can be asked in a cohort study?

Answer 18

What is the prevalence of a disease in a population?

What is the relative risk? What is the association between RF and outcome?

Question 19

When would you use a cohort study?

Answer 19

To see if exposure to something causes disease.

Question 20

How do you present results of a cohort study?

Answer 20

Calculate risk ratio or rate ratio.

Question 21

What are the advantages and limitations of a cohort study?

Answer 21

Advantages:

1. measure development of a disease.
2. calculate incidence (if prospective) and prevalence
3. look at multiple exposures
4. samples not biased by knowledge of outcome
5. easier to assess cause-effect

Limitations:

1. time consuming
2. behavioural changes over time
3. change in measurement/diagnostic procedures
4. selection bias
5. loss to follow-up
6. attrition bias
7. not suitable for rare conditions

Question 22

What is a cross-sectional survey?

Answer 22

Survey taken at a specific point in time to provide a ‘snapshot’ of a population. Prevalence of a disease, health, attitudes etc.

Question 23

What types of measurements can be used in a cross-sectional survey?

Answer 23

questionnaires, diagnostic samples, interviews

Question 24

What types of research questions can you use in cross-sectional studies?

Answer 24

Prevalence of ‘insert disease’ in x area.

Question 25

What are the advantages and limitations of cross-sectional surveys?

Answer 25

Advantages: 1. good for prevalence measurement 2. many variables compared at once 3. quick and easy 4. mass data-much data in records 5. good for public health planning 6. primary and secondary outcomes Limitations: 1. temporality- different results at different times e.g. disease with short duration 2. can't measure incidence 3. prone to recall, non-response bias 4. delivery pf questions different amongst administrators 5. self administered surveys 6. difficult to assign causality 7. cluster sampling requires large samples

Question 26

How would you present results of a cross sectional survey?

Answer 26

Different charts | Calculate averages

Question 27

What research questions foes an RCT address?

Answer 27

What is the effect of a specific intervention/therapy? | Establish whether a new intervention is more effective, as good as, inferior to a current one?

Question 28

What is the difference between risk and hazard?

Answer 28

Hazard=something that if you're exposed to it, increases chances of negative outcome. Risk= probability outcome will occur in a stated period of time.

Question 29

Wha is the difference between prevalence and incidence?

Answer 29

Prevalence='event states'. number of people for whom a disease has occurred at a specified point in time. Incidence='onset'. number of people for whom a disease occurs for the first time over a specific period of time.

Question 30

In a case-control study how are cases and controls defined?

Answer 30

Cases: Look for definitions of clinical diagnosis, list criteria needed, indicate time and severity. Controls: Can be compared to cases and are from same population as cases but have no disease.