Lecture 13- Evidence based medicine

Question 1

measures of risk

Answer 1

Risk ratio

Odds ratio

Question 2

risk ratio

Answer 2

or relative risk

ratio of incidence rate in group A vs incidence rate in group B

RR= I_a/I_b

Question 3

odds ratio

Answer 3

used most often in case/control studies

• ratio of odds of ‘outcomes’ in eposed group vs odds of outcome in unexposed group

OR= odds of outcome_exposed/ odds of outcomeunexposed

Question 4

risk ratio example

Answer 4

e.g. 25% chance of death

25/100

10/100

0.25/0.1

Risk ratio: 2.6

Question 5

odds ratio example

Answer 5

Odds ratio e.g. 1 in 3 chance of dying

25: 74= 1:3
10: 90 = 1:9

Odds ratio =3

Question 6

Absolute risk and absolute risk difference

Answer 6

Question 7

P-value

Answer 7

• P-value is an expression of “statistical significance”
  
  • It is the probability that the effect observed could have occurred by chance
• Thus, a small p-value implies a small ‘chance’ of that effect not being a ‘real effect’ of a given drug etc.
• Traditionally, p-values <0.05 are considered ‘statistically significant’, i.e. we are ‘happy’ to discount a 5% chance effect
• (or to put it another way, we are happy to consider that something that appears 95% likely to be a ‘true’ effect, is ‘sufficient’)

Question 8

Issues with p-values

Answer 8

Issues with p-values

Doesn’t rule out chance

Give no indication of the size of the effect

P- values give no idea as to the range of uncertainty around the effect you have estimated

Question 9

Confidence intervals

Answer 9

• From a study you generate an outcome (or result), for instance that your old treatment (drug A) reduces mortality by 50% (i.e. RR of 0.5) compared to your new treatment (drug B)
• Whilst your ‘best estimate’ is a RR of 0.5, there is ‘uncertainty’ around that estimate
• A 95% confidence interval ‘captures that uncertainty’. For your study you find the confidence interval is 0.2 to 0.8
  
  • i.e. there is a 95% probability that the true relative risk of mortality for drug A vs drug B may be as much as 0.2, or as little as 0.8
• This finding would be statistically significant as the confidence interval for the relative risk does not include 1

Question 10

for a result to be statistically signififcant the confidence intervals must not

Answer 10

cross 1

Question 11

Answer 11

yes- doesnt cross 1

Question 12

Answer 12

no-crosses 1

Question 13

Answer 13

Yes (doesn’t cross 1)

Statistically significant reduction

Question 14

what would be the absolute difference

Answer 14

1- 0.9= 0.1 kg

no difference would equal 0kg

Question 15

Answer 15

A and D (doesn’t cross 0)

Question 17

types of quantitative study design

Answer 17

observational and experimental

Question 18

types of observational study design

Answer 18

• cross-sectional
• case studies./series
• case-control studies
• cohort studies

Question 19

types of experimental study design

Answer 19

RCT

Question 20

hierachy of evidence

Answer 20