5: Evidence based medicine 1 Flashcards Preview

Endocrine Week 3 2017/18 > 5: Evidence based medicine 1 > Flashcards

Flashcards in 5: Evidence based medicine 1 Deck (26)
Loading flashcards...
1

What is a randomised control trial?

Comparison of intervention vs control in different randomised groups of patients

2

What is intention to treat analysis?

ALL patient's outcomes are included in the study after randomisation

e.g patients who are killed crossing the street to collect the drug

3

Why is intention to treat analysis important?

Preserves randomisation

4

What is per protocol analysis?

Only patients who received the intervention / control are studied

e.g patients who had a stroke BEFORE cerebrovascular surgery wouldn't be included in results

5

What's wrong with per protocol analysis?

Randomisation lost

specific factors will have resulted in those events occuring in the patients and you're basically ignoring them

6

What is the term used to describe the chances of an event occuring?

Probability

7

What are two ways of expressing probability in a study?

P value

Confidence interval

8

What is a p value?

Numerical value indicating the probability that an observation has occured due to chance

9

What does due to chance mean?

The observation is not statistically significant

10

What does statistically significant mean?

An observation which is probably not due to chance

11

What does a p value of 0.0014 mean?

14 observations out of 1000 will be due to chance; statistically insignificant

12

What is a confidence interval?

Range of values which probably contains the true value

13

What is the null hypothesis?

There's no difference between the two groups; observations are likely due to sampling error

14

A small p value (< 0.05) indicates that the null hypothesis is (true / false).

Small p value means FEW observations are due to chance

So the null hypothesis is probably FALSE

15

A large p value (> 0.05) indicates that the null hypothesis is (true / false).

Large number of observations are due to chance

So null hypothesis is probably TRUE

16

What is a Type 1 error?

False positive

You falsely reject the null hypothesis when it is true

17

What does a p value of 0.05 mean in terms of Type 1 error?

1 in 20 times you reject the null hypothesis when it's true

18

What is a Type 2 error?

False negative

You accept the null hypothesis when it's false

19

A Type 1 error is also known as a false ___.

False positive

20

A Type 2 error is also known as a false ___.

false negative

21

What is a confidence interval?

Range of values which is likely to contain the true value

22

What does a 95% confidence interval mean?

95% of true values are probably within the confidence interval

23

What are three possible explanations for a positive result?

Null hypothesis is false

Random chance

Confounding variable

24

What are four possible explanations for a negative result?

Null hypothesis is true

Random chance

Confounding variable

Size of the effect is too small to measure

25

What is statistical power?

Probability of correctly rejecting the null hypothesis

26

What affects the statistical power of a study?

Size of effect being measured

Sample size of study