Statistics definitions Module 1 Flashcards Preview

Gastro Year 4 Module 1 > Statistics definitions Module 1 > Flashcards

Flashcards in Statistics definitions Module 1 Deck (32):
1

Genotype

The genetic makeup of a cell, an individual, or an organism

2

Population Stratification

The presence of a systematic difference in allele frequency between subpopulations in a population, possibly due to different ancestry

3

Penetrance

Extent to which a particular gene or set of genes is expressed in the phenotypes of individuals carrying it (measured by the proportion of carriers showing the characteristic phenotype)

4

Screening

Application of diagnostic tests in patients at high risk for disease, but in whom there is no a priori reason to suspect that disease is present

average risk population

5

Surveillance

Repeated application of screening

high risk population

6

Sensitivity

True positives. Proportion of people with disease who have a positive test

7

Specificity

True negatives. Proportion of people without the disease who have a negative test

8

Univariate analysis

One variable is analysed at a time

9

Multivariate analysis

More than 2 variables are analysed together for any possible association or interaction

10

Hazard ratio

A comparison of probabilities of an event in a treatment arm with probabillities of an event in control group

at a given time - set end point

Chance of an event occurring in intervention arm/ Chance of an event occurring in control arm

HR<1 at any particular time less patients in the intervention group are experiencing an event compared to the control group

HR=1 same for both

HR>1 at any particular time more patients in the intervention arm are experiencing an event compared to control group

11

Confidence interval

A confidence interval tells you how much uncertainty there is with any particular statistic.

Tells you how confident you can be that the results from a poll or survey reflect what you would expect to find in the entire population.

depends on sample size and variation within the sample

12

Null hypotesis

No significant difference between specified populations
(any observed difference being due to chance)

13

Incidence rate

Number of new cases in a population, per year

14

Prevalence

Number of cases in a population at a given point in time

15

Mortality rate

Deaths in a population per time period

16

Standardise rate

Adjust numbers to account for variables, by comparing with expected numbers in normal population

17

Odds ratio

A comparison of the odds of an event after exposure to a risk factor to the odds of an event in a control or reference setting

18

RR

The relative risk of an event is the likelihood of its occurrence after exposure to a risk variable as compared to the likelihood of its occurrence in a control or reference group

19

Prospective study

Selection of cohorts and measurement of exposures or risk factors before the outcome has occurred

Advantages: high quality, can accomodate for novelty
Disadvantages: expensive and time consuming

20

Retrospective

Looks backwards at risk factors or exposures that have already led to a specific outcome.
Advantages: Fast and inexpensive
Disadvantages: rely on accurate record keeping, unmeasured confounders might be present

21

NPV

Probabitly that negative test result does not have the disease

22

PPV

Probability that positive test result truly has the disease

23

P value

proballity of calculating a difference when there really isnt one - p values smaller than significance level allow you to reject your null hypothesis
p=0.05 gives significance levels of 95% , 95% sure that the difference is not due to chance

24

ITT

Once patients are randomised, analysis includes every patient regardless of non-comliance or dropout

avoids overoptimistic estimates
but dilute effect of intervention as including those that didn't go according to plan

25

per protocol

only analyses patients that have completed the protocol
limitations: results in ideal patient, not real world
can introduce confounders by including patients were randomised, what caused them to withdraw?

26

Cofidence level 95%

inverse of p value

27

Cross sectional study

analyse data collected at a specific point in time
advantages: routine data collection so very little expense
disadvantages: open to confounding factors

28

case control

Compare cases with disease and controls with disease and look retrospectively at exposures
advantages: cost effective and quick
disadvantages: selection bias, corelation not caussation

29

cohort study

compare exposure and non exposure ove rtime, to see who will develop a specific outcome
advanatages: can measure mltiple outcomes
disadvanateges: correlation, not causation

30

RCT

experiment that introduces treatment/exposure and then measures effect - gold standars for measuring treatment effect

advantages: randomisation minimises selction bias and confounding factors

Disadvantages: expensive, time consuming

31

Systematic review

critical assessment and evaluation of all research studies addressing a particular issue
advanatges: quick cheap minimise bias
disadvantages: selction bias can be low quality

32

Meta analysis

quantitative pooling of data - in systematic reviews
Higher statistical power but can have publication bias o r selection of studies to include