Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

ISBM > Epidemiology > Flashcards

Epidemiology Flashcards

(100 cards)

Start Studying
1
Q

What is epidemiology

A

The study of the distribution, determinants and control of diseases of a population

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

What is clinical epidemiology

A

The study of determinants of disease outcome in individuals with disease

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

What are exposures

A 
Any factor that might be associated with outcome 
Eg external environment (pollution etc); lifestyle (diet smoking etc)
Individual characteristics (height weight etc); medical intervention (drug/ etc)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is an outcome

A

Any health related state eg disease occurrence complications and survival

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

Give pros and cons of ecological study

What is it

A

Pro: cheap and quick

Cons: limited exposures, confounding serious problem

Observational study - test for correlation of exposure and outcome at population level

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

What is a retrospective cohort study

A

Population defined after follow up time has already occurred

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

Pros and cons of cohort study

A

Pro: no bias in exposure
Exposure precedes outcome
Direct estimate of incidence
Good for rare exposures

Cons: not good for rare outcomes
Large sample size
Potential bias in outcome

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

What is important to clarify for case control population

A

Population of interest and outcome of interest

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

Odds ratio=

A

a x d
———
b x c

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

How do you decrease the effect of confounders

A

Increase the size of the population as the differences tend to even out

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

How does case control studies work

Give a negative of this study

A

It is a observational study where you pick a case group (who have outcome) and a control group and see who did and didn’t have the exposure

Negative recall bias

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

How does a cohort study work

A

Choose a cohort without knowing exposure and outcome

It is an observational study based on area or time NOT based on exposure or outcome

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

What is the issue with retrospective studies

A

Recall bias

Eg people with lung cancer tend to remember how many cigs they had in the past whereas those without cancer don’t remember usually

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

What is information bias

A

Different collection of information

Eg our assessment of exposure is affected by outcome

Say a stereotypical pub goer and athletes both said they had just had “a few drinks”
We may assume the the former has more based on appearance alone

Information bias can also be where the assessment of outcome changes based on exposure

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

How do you fully remove bias

A

YOU CANNOT

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

When do you do intention to treat?

A

If the dropout etc is linked to the drug etc

Eg if they dropped out due to side effects

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

When can one conclude that correlation = causation

A

Once chance, bias, and confounding variables have been removed

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

How can you eliminate chance

A

Use stats: find the p value and remove the chance factor

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

When do you use odds ratio

A

If it is a case control experiment

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

When do you use a 2x2 matrix for an ecological study

A

NEVER

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

Pros and cons of a clinical trial

A

Pro:
no bias in exposure
No selection bias
Blinding minimises bias in outcome assessment

Con: 
Single exposure
No good for rare outcomes
Randomisation May be difficult/ unethical
Cost 
Follow up time
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

What do statistical tests do

A

Assess if an association is present or not

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

How do you measure the strength of association of a cohort study

A

Calculate relative risk

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

What is the odds ratio from a case control study approximately equal to

A

The relative risk of the disease is uncommon in the population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
What is selection bias Eg?
When a participant selection distorts the exposure - outcome relationship from that present in the target audience Those who attend health screenings often have better outcomes than others as they are more health conscious
26
What is the healthy worker effect What is a similar effect
Comparing workers in a factory etc to general population Working individuals tend to healthier than a group that includes unemployed people The healthy migrant effect where migrant groups tend to be healthier than non migrant groups
27
Are retrospective studies always biased
Nay
28
What is the primary purpose of a random clinical trial
To remove confounding as confounding variables are equalised between 2 groups
29
Are prospective and retrospective synonyms for cohort and case control respectively
Nope
30
2 ways to describe a disease in a single group
Prevalence | Incidence
31
3 | Ways to compare disease between 2 groups
Relative risk Relative risk reduction Excess risk or risk difference
32
Which of the following is can be measured as a rate and which by a proportion? Why? Total number of flights Total flying time
Total flying time= rate Total number of flights = proportion A rate has time in the denominator
33
Can event numbers without a denominator provide a useful measure of risk
No
34
What is prevalence
The proportion of a group of individuals that have the condition
35
What is incidence
The proportion of a defined group of disease free people that develop a disease over a specified time period ie incidence is a rate
36
How to calculate relative risk
Incidence in “exposed” ———————————- Incidence in “unexposed”
37
How to calculate excess risk
Incidence In exposed - incidence in unexposed
38
What kind of risk is more important
While relative risk sounds more dramatic, absolute risks are more relevant for personal decision making
39
What is relative risk reduction
1- relative risk
40
What kind of risk does benefit ratio depend on
Absolute risk
41
Give 3 measures of population health Why do we use these summaries
Life expectancy Adjusted life expectancy Life lost Compare different populations Compare same population over time Identifying health inequalities Priority debates for health service planning/ delivery and for research
42
What is life expectancy
Median expected survival if current age specific mortality rates were applied to a newborn child over its lifetime
43
What are the different kinds of adjusted life expectancy
Disability free life expectancy Disability adjusted Quality adjusted
44
What is measure “life lost”
Years of life lost Based on ideal life expectancy of 81.5 years There is also disability adjusted years of life lost
45
What is the epidemiological transition
Economic development-> Decline in communicable etc mortality rates -> People live to older ages-> Burden of non communicable disease and disability increases Age specific non communicable mortality rates may decline
46
Describe primary prevention as a societal response to chronic disease
Aim: to reduce incidence of disease 2 interventions: at population level and in at risk individuals
47
Give 7 risk factors for cardiovascular disease
``` Lipid profile (blood cholesterol level) Blood pressure Smoking Diabetes/ blood glucose Haemostatic factors like fibrinogen Inflammatory markers Homocysteine ```
48
What could be the criteria that is assessed for treatment
Those above a threshold | High risk individuals
49
How is a threshold for treatment decided?
Risk benefit balance = favourable Depend on absolute risk
50
What are the issues with a high risk approach for distributing treatment
Need to be able to identify the high risk Individual decision making required Large numbers treated for a long time to prevent one event Limited impact on total disease burden
51
What are advantages of the high risk approach
Those at high risk get most benefit | They are motivated to take treatment
52
What are the problems with a population approach to treatment
Limited knowledge of causes of risk factor | Difficult to implement changes across a population
53
Why do we need diagnostic/ screening tests
To distinguish between those who do and don’t have the disease so that appropriate action can be taken
54
Give 5 examples of screening tests and what they test for
Mammography: preclinical Breast cancer Alpha-fetoprotein: baby with neural tube defect ECG: heart disease Serological testing: infection in animals Somatic cell counts: subclinical mastitis in cattle
55
What is the perfect diagnostic test
When healthy people never trigger a response and those with the disease always trigger a response ie no false positives etc
56
Are all screening tests binary
No they may give a range of possibilities or be a continuous measure For a continuous measure a threshold is required
57
What do graphs of a perfect diagnostic test look like
No overlap
58
Where are false results found on a measure of disease vs frequency graph
In the overlap
59
What is sensitivity
Number that test positive ———————————— Number that have disease
60
Define sensitivity
How good the test is at identifying those who really have the disease
61
What is specificity
How good a test is at identifying those who do not have the disease
62
Specificity=?
Number that test negative ————————————- Number that do not have disease
63
How does an increase in sensitivity affect specificity
Reduces specificity
64
How do you balance sensitivity and specificity
Depends on context Consequences of missing disease vs consequences of saying someone has disease when they don’t
65
Positive predictive value=
Number with disease —————————— Number that test positive
66
Negative predictive value=?
Number of healthy —————————— Number that test negative
67
How does changing specificity affect PPV What about sensitivity
Has a big effect (as specificity increases PPV curve becomes steeper) Limited effect of sensitivity on POV
68
How does changing specificity and sensitivity affect NPV
Specificity: limited effect Sensitivity: big effect
69
What is the screening paradigm
To detect the disease in asymptomatic people when disease is more treatable in order to improve outcomes
70
What is PREDICT
A web interface for clinicians and patients using data from regional cancer registry to develop a multi variable model
71
Define odds ratio
The ratio of the odds of outcome A given exposure B vs outcome A without B
72
Define relative risk
Ratio of probability of an outcome in 2 groups Eg cancer is smokers vs non smokers
73
Are odds ratio and relative risk the same
Approximately equal if disease is rare Not equal if disease is uncommon
74
How do odds ratio and relative risk compare if the disease is common
Odds ratio > relative risk
75
Odds ratio of a positive outcome is the inverse of that of a negative outcome. True or false
True It is not true for relative risk
76
How should relative risk be used in a case control study
RELATIVE RISK SHOULD NOT BE USED IN CASE CONTROL This is because RR would artificially inflate prevalence compared to cohort studies
77
What is denominator data
Any general data about a population
78
What is a measure
The numerator The incidence of an event in the population
79
Prevalence=
Incidence x disease duration
80
What is relative risk compared to excess risk
Relative = relative difference between 2 populations Excess= absolute increase in incidence- this is more relevant for personal/ clinical decisions
81
What does harm:benefit ratio depend on
Absolute risk
82
How to work out the number of people needed to treat to prevent an event
1/ excess risk
83
What is population excess/ attributable risk?
How much woukd disease incidence fall if a risk factor was removed Depends on excess risk between those with and without the factor, and the prevalence of the factor
84
What is population risk calculated as
The % of incidences attributable to the risk factor in question
85
Are most diagnostic tests binary
No We define a threshold above which is positive
86
What is the Rothman component cause model
Describes how different combinations of component causes could combine to be a sufficient cause of disease but individual components are neither necessary nor sufficient
87
Define sensitivity
How good a test is at finding those who really have the disease
88
Define specificity
How good a test is at finding those who don’t have the disease
89
Define PPV
Probability that subjects with a positive result truly have the disease
90
Define NPV
Probability that subjects with a negative result truly don’t have the disease
91
What has more of an effect on PPV
Specificity
92
Can case control studies enable the prevalence of a disease to be determined
No
93
Is r important
Yes but r squared is more important
94
Define SEM
Standard deviation of the sampling distribution of the mean
95
What is not an issue with the high risk approach
The use of drugs on asymptomatic patients
96
What is better a clinical trial or systematic review
Systematic review
97
Why may a confidence interval be narrower for one of 2 groups
It may be a bigger group
98
What does it mean if the ranges of confidence intervals overlap
It is unlikely to be significantly different
99
What does it mean if 2 confidence intervals don’t overlap
The difference is likely to be statistically significant
100
Why might we fail to find a difference
There is no difference Or We can see a difference but the power/ sample size is too small to achieve statistical significance

ISBM (2 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions