AKT Flashcards > Statistics > Flashcards

Statistics Flashcards

(86 cards)

Study These Flashcards
1
Q

Definition and how to calculate ‘mean’? And when is it used?

A

Definition - Sum of all values divided by number of values

Used when spread of data fairly similar on each side of mid-point (normally distributed)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Definition and how to calculate ‘median’? And when is it used? What is IQR?

A

Known as midpoint

Definition - Point at which has half the values above and half below

Used to represent average when data is not symmetrical, for instance skewed distribution

Interquartile Range - 3rd quartile minus 1st quartile (contains middle ½ of sample)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Definition of ‘mode’?

A

Definition - most frequently occurring event/value

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Definition of interquartile range?

A

Difference between 1st quartile (25%) and third quartile (75% below)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is standard deviation?
When is it used?
What does 1, 2 and 3 SD include (%)?

A

Indicates how much set of values spread around the mean

Used for data normally distributed

1 SD includes 68.2%
2 SDs includes 95.4%
3 SDs includes 99.7%

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is 95% confidence intervals?

What alters CI’s?

A

Range that is likely to contain true population value lies

Means we can be 95% confident that true values lies within this range

Size is related to sample size, variability of individual results

Larger studies have narrower confidence intervals

Charts e.g. forrest plot can show different studies with confidence intervals shown as horizontal lines

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are p-values?

A

Probability is a number describing how likely an event is to occur/is true, between 0-1 (0 will not happen, 1 certainly will happen)

P Value = probability of any observed difference having happened by chance

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is null hypothesis and what do p-values of 0.5 and 0.05 represent?

A

Null hypothesis = no difference between two treatments

P=0.5 - probability of difference having happened by chance is 50:50

P= 0.05 - probability of difference happened by chance 1 in 20, 5%

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Define what statistical significance is?

A

Unlikely to have happened by chance and therefore important

Usually P<0.05 considered statistically significant

P<0.01 highly significant

P<0.001 is very highly significant

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

When is sensitivity and specificity used?

A

Used to analyse value of tests or screening tests

For any disease - may be present or absent and test can be positive or negative

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the Two-Way Table for Sensitivity and Specificity

Calculate Sensitivity and Specificity?

A

Sensitivity = If patient has disease, how often test will be positive (A/A+C)

Specificity = If patient is healthy, how often test will be negative (D/D+B)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Calculate Positive Predictor Value and Negative Predictor Value

A

Positive Predictive Value = If test result positive, likelihood that the patient will have condition (A/A+B)

Negative Predictive Value = If test result negative, likelihood that the patient does not have condition (D/D+C)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Define Type 1 Error

A

Symbolised by alpha

Equivalent to false-positive result

Hypothesis that is correct, is rejected (P-value)

Generally occurs due to random chance

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Define Type 2 Error

A

Symbolised by Beta

Equivalent to false-negative results

Hypothesis that is incorrect, is accepted

Down to function of power, insufficient sample size

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

When is relative risk used?

A

Relative risk used in cohort studies, prospective studies that follow group over period of time and investigate effect of treatment/risk factor

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Define risk

A

Risk = probability that event will happen (0-1)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Define Risk Ratios

What does 1 represent? And <1 and >1?

A

Risk Ratios = dividing risk in treated groups by risk in control group

1 = no difference in risk
>1 rate event is increased compared to controls
<1 rate of event is reduced
If CI of risk ratios does not include 1 then statistically significant

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Define Odds

A

Odds = calculated by dividing number of times an event happens by number of time it does not happen (one in every two births = 1/1 = 1, one in every 100 births = 1/99 = 0.0101)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Define and how to calculate odds ratios?

What does 1 represent? And <1 and >1?

A

Odds Ratio = Dividing odds of exposed group by odds of control group

OR of 1 = no difference in risk between groups
OR >1 = rate of event increased in exposed group
OR <1 = rate of event decreased in exposed group
If CI of odds ratio does not include 1 then statistically significant

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

What is Absolute Risk?

What is Absolute Risk Reduction?

A

Absolute Risk = number of events in treated/control group / number of people in that group

Difference between absolute risk in control group and intervention group (Control Event Rate - Experimental Event Rate)

Given as percentage

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What is relative risk?

What is relative risk reduction?

A

Relative Risk (RR) = EER/CER

Relative Risk Reduction (RRR) =(EER-CER)/CER

EER: Experimental event rate (how often does an event happen in the intervention group)

CER: Control event rate (how often does an event happen in the control group)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What is Number Needed to Treat (NNT)?

A

Number of patients who need to be treated for one to get benefit

NNT = 1/ARR or 1/(EER-CER)

ARR: Absolute risk reduction (the difference between the event
rate in the intervention group compared to the control group)
EER: Experimental event rate (how often does an event happen in the intervention group)
CER: Control event rate (how often does an event happen in the control group)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

What is Numbers Needed to Harm (NNH)?

A

Number of patient who need to be treated for one to be harmed, for example side effects

Calculated NNH = 1/ARI (Absolute Risk Increase)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Define Incidence

A

Number of new cases of condition over given time frame

In percentage of population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Define prevalence
Exisiting number of cases of condition in since point in time as a percentage of population
26
Define Disability-Adjusted Life Years. How to calculate DALY?
Measurement of disease burden Impact of health problem on an individual or in a population Used to compare burden of 2 diseases DALY = Years of life lost due to premature mortality (YLL) + Years lived with disability (YLD) YLL = Number of deaths at age x standard life expectancy at age YLD = Number of incident cases x disability weight x average duration of the case until remission or death in years
27
What is QALY?
Additional number of years of life x HRQL (Health Related QOL)
28
What is a Forest Plot for Meta-Analyses?
Graph that summarises results of multiple studies in meta analysis Shows individual study results, confidence intervals, weighted average of all studies, line of no effect, study weight
29
What is a scatter plot graph? What variables are suitable for scatter plot?
Graph in which values of two numeric variables are plotted along two axes, pattern reveals any correlation
30
What is L'Abbe Plot? What does it show? What does size of circle represent? What is reference line?
Scatterplot that shows results of individual studies in meta analysis that compare treatment and control interventions Shows risk on vertical axis and risk of control group on horizontal axis Size of circle represents each study precision Reference line indicates when outcomes of two groups identical Estimate effect size line shown based on model
31
What is a funnel plot? What is it used to identify?
Scatterplot shows how effect of treatment varies across studies of different sizes Used to identify publication bias, small study effects and other reporting biases Effect estimate for each study on x axis, precision of effect estimate for each study on y axis Points above or below funnel limites are considered outliers
32
What is a box and whisker plot? What does it show?
Graph showing median, range and inter-quartile range of set of values + highest and lower values excluding outliers Lower and upper quartiles shown as horizontal lines either side of rectangle
33
What is Cates Plots? What does green, yellow and red face mean?
Smiley face plots to visually communicate risks and benefits of treatment Green face = good outcome, not affected by treatment Yellow face = treatment changes category from bad to good outcome Crossed out green face = treatment causes adverse event and causes bad outcome Red face = bad outcome, not affected by treatment
34
What is bar chart? What does it represent?
Heights of bars represent number of occurrences in each categorical variable
35
What is histogram? What variables should it be used for?
Frequency distribution of continuous variables Numerical data
36
What is statistical process control charts?
Graphs that show how data changes over time
37
What is Kaplan Meier survival analysis?
Plots graphically survival of sample cohort with survival estimates re-calculated whenever there is a death
38
Types of Study - what is case study/series?
Research into single person or small group of patients
39
Types of Study - What is cross sectional survey? What kind of study? What can it not do?
Observational study where data collected at single point in time, snapshot of population to measure the prevalence of health outcome/characteristic Examines associations Cannot establish causation
40
Types of Study - What is case control study?
Retrospective study which investigates relationship between outcome and one or more risk factors Selecting patients who already have disease or outcome, matching them to patient who do not and then comparing effect of risk factors
41
Types of Study - What is cohort study?
Prospective, observational study that follows group over period of time and investigates effect of treatment or risk factor
42
Types of Study - What is randomised controlled trial?
Participants randomly selected to different groups, allowing comparison of effects of intervention while minimizing bias Gold standard of medical research
43
Types of Study - What is systematic review?
Analysis of all available research on specific clinical question, using methods to identify, select and critically appraise relevant studies Aims to provide unbiased answer to question
44
Types of Study - What is meta-analysis?
Process of combining study results using statistical methods on same topic to draw conclusions or plan new studies
45
What affects strength of evidence?
Measure of how convincing the evidence is for a hypothesis Factors affecting - study quality, study design, consistency, precision, transparency
46
Types of Statistical Tests - Define Correlation?
Correlation - Strength of relationship between two variables Denoted by letter r (can be positive or negative) R= 0 no correlation
47
Types of Statistical Tests - What is regression? What is regression line and co-efficient?
Regression - used to quantify how one variable in a data set relates to another Regression Line - line that fits best through the data point on graph, calculated using least squares between points and line Regression Co-Efficient - gradient of graph, change in value of one outcome per unit change in other,
48
Types of Statistical Tests - What are the types of regression?
Linear - line that best fits the points is straight Logistic - Where each case in sample can only belong to one of two groups, outcome as probability that case belongs to one group Poisson - Study waiting times or time between rare events Multiple - Compares effects of two or more variables on the outcome, e.g. BMI, age, gender on PEFR
49
Types of Statistical Tests - When is Spearman's Rank used?
Estimate of correlation used for non-parametric variables
50
Types of Statistical Tests - When is Pearson's Correlation Used?
Method of calculating correlation coefficient if values are sampled from normal population
51
Types of Statistical Tests - When is Chi-Squared Test used?
Test of association between two categorical variables used to compare proportions or percentages e.g. compares the percentage of patients who improved following two different interventions
52
Types of Statistical Tests - When is Mann Whitney U used?
Used in non-parametric data (not normally distributed) Rank data and compare ranks in unpaired data compares ordinal, interval, or ratio scales of unpaired data
53
Types of Statistical Tests - When is Wilcoxon Signed Rank Used?
Non-parametric test for comparing difference between paired groups, e.g. before and after treatment
54
Types of Statistical Tests - When is ANOVA/Friedman's ANOVA used?
Used in parametric (normally distributed) data Used to compare the means of two or more samples to see whether they come from the same population (null hypothesis)
55
Types of Statistical Tests - When is Students T Test Used?
Used in parametric (normally distributed data) Used to compared two samples
56
Types of Data - What are the types of quantitative data?
Quantitative - numerical data Discrete - count of individual items or values in integers Continuous - measurement of non finite values, in units smaller than one
57
Types of Data - What are the types of qualitative data?
Categorical - groupings of some kind, categories Binary - Yes/No outcomes Nominal - Groups with no rank or order between them Ordinal - Groups that are ranked in specific order
58
Types of Bias - Selection Bias?
Non random assignment to study group
59
Types of Bias - Recall Bias?
Knowledge of presence of disorder alters recall by subjects
60
Types of Bias - Sampling Bias?
Subjects not representative relative to general population, not generalisable
61
Types of Bias - Procedure Bias?
Subjects in different groups not treated the same
62
Types of Bias - Lead-Time Bias?
Early detection confused with increased survival (improved screening)
63
Types of Bias - Confounding Bias?
Occurs with 2 closely associated factors, the effect of 1 factor distorts or confuses the effect of the other
64
Types of Bias - Confirmation Bias?
Tendency to interpret new evidence as confirmation of one’s existing beliefs/theories
65
Types of Bias - Congruence Bias?
tendency to over-rely on testing their initial hypothesis, neglecting to explore alternative outcomes
66
Types of Bias - Outcome Bias?
Error made in evaluating quality of a decision when outcome already known
67
Types of Bias - Publication Bias?
Failure to publish results of a study on basis of direction or strength of study findings
68
Types of Bias - Courtesy Bias?
Individuals tend not to fully state their unhappiness with service to not offend
69
Types of Bias - Hawthorne Effect?
Occurs when group being studied changes its behaviour owing to knowledge of being studied
70
Types of Bias - Pygmalien Effect?
Occurs when researcher’s belief in the efficacy of treatment changes outcome of treatment
71
Types of Bias - Berkson's Bias?
Sample taken from subpopulation (hospital) rather than general population
72
How to calculate Power in a study?
1 - Type 2 Error (Beta)
73
Phases of Trials - Phase 1?
small studies (e.g. 100) on healthy volunteers used to assess pharmacodynamics and pharmacokinetics
74
Phases of Trials - Phase 2?
small studies (e.g. 100-300) on actual patients examines efficacy, adverse effects
75
Phases of Trials - Phase 3?
larger studies (e.g. 500-5,000 patients) examines efficacy, adverse effects may compare drug with existing treatments studies of special groups e.g. renal, elderly If drug shown to be safe and effective then drugs may be approved for marketing
76
Phases of Trials - Phase 4?
post-marketing surveillance
77
What is standard error?
SE = Standard Deviation / square root of sample size
78
Warfarin targets - when is the target INR 2.5 (2-3)?
Tx for DVT/PE (APS) AF (valvular AF) Cardioversion Bioprosthetic valve Dilated cardiomyopathy Post MI
79
Warfarin targets - when is the target INR 3.5 (3-4)?
Recurrent DVT/PE in people currently receiving anticoagulation and with an INR above 2
80
Warfarin targets - mechanical heart valve low risk?
2.5-3.0
81
Warfarin targets - mechanical heart valve medium and high risk?
3.0-3.5
82
Warfarin - INR out of range - if major bleeding?
- Stop warfarin - Arrange immediate emergency admission - IV Vitamin K 5mg and dried PCC (FII, VII, IX, X) or FFP if PCC unavailable
83
Warfarin - INR out of range - Greater than 8 with minor bleeding?
- Stop warfarin - Arrange same-day hospital assessment IV Vitamin K 1-3mg (dose rpt after 24h if still too high) - Restart warfarin when INR <5
84
Warfarin - INR out of range - Greater than 8 with no bleeding?
- Stop warfarin - Arrange same-day hospital assessment for Vitamin K 1-5mg PO (dose rpt after 24h if still too high) - Restart warfarin when INR <5
85
Warfarin - INR out of range - Between 5–8 with minor bleeding?
- Stop warfarin - Arrange same-day hospital assessment for consideration of Vitamin K PO 1-3mg - Restart warfarin when INR <5
86
Warfarin - INR out of range - Between 5–8 with no bleeding?
- Withhold 1 or 2 doses of warfarin - Monitor INR closely and alter subsequent maintenance dose appropriately - Consider Vitamin K if high risk
AKT Flashcards
flashcards
Decks in class (37)
# Cards
Brainscape helps you reach your goals faster, through stronger study habits.
© 2026 Bold Learning Solutions. Terms and Conditions