Lecture 12 - Key Statistics In Evidence Based Medicine

Question 1

What is the definition of incidence?

Answer 1

Measures the number of NEW cases of a particular disease arising in a population at RISK in a CERTAIN TIME PERIOD

(A rate)

You would IGNORE pre-existing cases

Question 2

Look at slide 6:

Calculate the incidence:

Answer 2

4/10 new cases of disease merge in a year

Question 3

What is prevalence?

Answer 3

ALL CASES of disease existing in a given population at a given time (Point prevalence)

Question 4

Look at slide 10:

What is the point prevalence at year 2?

Answer 4

5/10 so 50% prevalence

Question 5

What is the difference between incidence and prevalence?

Answer 5

Incidence is the number of new cases that develop in a time period

Prevalence is all the cases of a disease at a given time point

Question 6

What is meant by high disease incidence?

Answer 6

High rate of development of disease

Question 7

What is meant by risk ratio?

Answer 7

The ratio of risk in group A Vs risk in in group B

So the risk of certain outcome in exposed group compared to the risk of certain outcome in UNexposed group

Question 8

What is odds ratio?

Most commonly used in case control studies:

Answer 8

The ratio of thee odds of an Outcome in Group A (those exposed) Vs the odds of the outcome in Group B (those unexposed)

Question 9

Look at slide 23:

Calculate the risk ratio of dying between the 2 groups:

Answer 9

Outcome is death
Exposure is smoking

25/100 = 0.25
10/100 = 0.1

0.25/0.1 = 2.5

Risk ratio = 2.5
This means that those who were exposed to smoking were 2.5 times more likely to die than those who dont smoke

Question 10

Look at slide 37:

Calculate the odds ratio of dying between the 2 groups:

Answer 10

Smoke is exposure
Outcome is death

Group A= 25:75 = 1:3
Group B= 10:90 = 1:9

Odds ratio of 3

Question 11

What is absolute risk?

Answer 11

The risk of acquiring a given disease over a. Given period of time

Question 12

What is absolute risk difference?

Answer 12

Absolute risk in exposed group - absolute risk in unexposed group

Question 13

What p-value will be indicated if the result of a statistical test is significant?

Answer 13

P <0.05

Question 14

If the P value of a statistical test is significant so p < 0.05 , what does this mean for the null hypothesis?

Answer 14

There is strong evidence to suggest the null hypothesis (Ho) is not true

Larger the p-value, the weaker the evidence against the null hypothesis

Question 15

What is actually meant by the p-value?

Answer 15

The CHANCE of getting a result which supports the null hypothesis

Larger the p-value the more we believe the null hypothesis
Smaller the p-value the evidence we have AGAINST the null hypothesis

Question 16

What are the problems with p-values?

Answer 16

Gives no indication as to the size of the effect

Gives no idea as to the range of uncertainty

P value threshold is arbitrary

Question 17

What is a confidence interval?

What confidence interval is normally used?

Answer 17

A range of values in which you are confident that the true value lies

95% confidence interval

Question 18

What value of RISK RATIO needs to NOT be in the 95% confidence interval range of values for the test to be STATISTICALLY SIGNIFICANT?

Answer 18

Risk ratio of 1 needs to NOT be in the RANGE

Question 19

Look at slide 66:

Which risk ratios show a statistically significant result?

Why?

Answer 19

A
C
D

The confidence interval does not contain the RISK RATIO 1 so its statistically significant

Question 20

What value of Mean Absolute difference value needs to NOT be in the 95% confidence interval range of values for the test to be STATISTICALLY SIGNIFICANT?

Answer 20

Mean Absolute Difference of 0

Question 21

Look at slide 72:

Which Mean Absolute Difference shows a statistically significant result?

Why?

Answer 21

A
D

The range of mean absolute difference values at the 95% confidence interval doesn’t include the mean absolute difference value of 0