Evidence based medicine

Question 1

what is the definition of a P value

Answer 1

a numerical value indicating the probability that an observation has occurred due to chance

Question 2

what scale is the p value measured on

Answer 2

0-1

Question 3

what does p0 mean

Answer 3

impossible event

Question 4

what does p1 mean

Answer 4

certain event

Question 5

what is a null hypothesis (H0)

Answer 5

a hypothesis where there is no relationship between study variables

Question 6

what is a type 1 error in a null hypothesis (H0)

Answer 6

when you reject H0 and its true (false positive)

Question 7

what is a type 2 error in a null hypothesis (H0)

Answer 7

when you accept H0 and its false (false negative)

Question 8

what are confidence values

Answer 8

a range of values which probably contain the ‘true’ value

Question 9

what is the definition of a statistically significant finding

Answer 9

when you can correctly reject a false H0

Question 10

what can affect a statistically significant finding

Answer 10

population size

Question 11

when are randomised control trials not used (3)

Answer 11

if not safe, expensive, unethical (large scale trial so need lots of people)

Question 12

what is another name for a randomised control trial

Answer 12

cohort study

Question 13

does a cohort study use a

a) population of people, some affected and some not affected followed up over time
b) a group of people all affected, then look at their habits etc

Answer 13

a

Question 14

what is it called when you do a study based on a group of people, all of whom are affected

Answer 14

a case control study

Question 15

why would a case control study be used (and example)

Answer 15

if there are no willing participants (eg mobile phones and brain cancer)

Question 16

what is the method of a cohort study/randomised control trial

Answer 16

a group of people, some exposed some not, followed up over time
incidence (WITH outcome or WITHOUT outcome) is compared between exposed and not exposed (= 4 end groups)

Question 17

what is a prospective cohort study

Answer 17

FORWARDS ARROW
look at a population, split into exposed and not exposed groups then over time compare exposure to incidence rates (WITH outcome and WITHOUT outcome)

Question 18

what is a retrospective cohort study

Answer 18

BACKWARDS ARROW

looking at the effect of exposure, but the outcomes have already occurred

Question 19

what is the benefit of a retrospective study over a cohort study

Answer 19

quicker - outcomes have already occurred

Question 20

what risk calculation is used in a cohort study (2)

Answer 20

relative risk/risk ratio (RR)

absolute risk

Question 21

what is relative risk/risk ratio (RR)

definition and calculation

Answer 21

used for a cohort study

incidence in exposed group divided by incidence in not exposed group

a/(a+b) divided by c/(c+d)

Question 22

what is absolute risk

Answer 22

used for a cohort study

risk in exposed group minus risk in non exposed group

a/(a+b) minus c/(c+d)

Question 23

what is a case control study

Answer 23

when you find a group of patients that all already have the outcome and look at their exposure

Question 24

what cohort study is a case control similar to

Answer 24

a retrospective cohort study

Question 25

what is the difference between a retrospective cohort study and a case control study

Answer 25

a case control study only looks at people WITH an outcome whereas a retrospective cohort study looks at people WITH and WITHOUT an outcome

Question 26

how are retrospective cohort studies similar to case control studies

Answer 26

they both find a group of people THEN look at their exposure = dont need to wait for the changes to happen

Question 27

which ratio is calculated for a case control study

Answer 27

odds ratio (OR)

Question 28

what is an odds ration

Answer 28

used for a case control study odds of event when exposed divided by odds of event when not exposed a/c divided by b/d

Question 29

what are the problems with a case study (3)

Answer 29

relies on peoples memory control group isn't representative of population(you only have people WITH an outcome, not WITHOUT) social acceptability bias (tell ourselves lies to make us feel/look better)

Question 30

qualitative research

Answer 30

analyze data from people about experiences and circumstances via interviews not numbers

Question 31

quantitative research

Answer 31

based on numbers

Question 32

what makes a confidence interval become more narrow

Answer 32

increase in sample size

Question 33

how does the mean and median correlate in a normal population distribution

Answer 33

same value

Question 34

statistical significance definition

Answer 34

results are unlikely to have arisen by chance

Question 35

definition of cohort study

Answer 35

measuring things at intervals over time | patients already in the same group (eg born in same hospital, at same GP practice)

Question 36

what is it called when there is no longer a need to sample more to reach new conclusions or to back up existing conclusions

Answer 36

saturation

Question 37

external validity definition

Answer 37

the extent to which one can appropriately apply the results to other populations

Question 38

what is a case control study

Answer 38

group of people with risk factor and group of people without risk factor (control group)

Question 39

does peer review give validitiy

Answer 39

no

Question 40

what % is the normal distribution of the population in what is this interval called

Answer 40

middle 95% (2.5% either side) confidence interval

Question 41

governing process for audit (collecting patient data that already exists)

Answer 41

caldicott

Question 42

governing process for collecting data from students

Answer 42

university research ethics committee

Question 43

governing process for collecting data from current patients (eg via interviews, not treatment success etc)

Answer 43

NHS research ethics committee