Research methods1 Flashcards
(98 cards)
1
Q
What is epidemiology?
A
Epidemiology is the study of the distribution and determinants of health and disease in populations.
2
Q
What are health statistics used for in epidemiology?
A
Health statistics are used to describe the distribution of disease in time, place, and person.
3
Q
Define routine data in the context of epidemiology.
A
Routine data is information about a defined population collected in a consistent manner for administrative reasons.
4
Q
What are the two most important vital statistics for assessing a population’s health?
A
- Life expectancy
- Infant mortality
5
Q
What is life expectancy?
A
Life expectancy is the average number of years that a person can be expected to live.
6
Q
What is infant mortality?
A
Infant mortality is the number of children per 1000 live births who die in their first year of life.
7
Q
List some strengths of vital statistics.
A
- Cheap and readily available
- Almost complete data
- Can be used for ecological studies
- Recorded at regular intervals
8
Q
What are some weaknesses of vital statistics?
A
- Not 100% complete
- Potential for bias
- Data can become out of date
9
Q
What is the importance of routine data quality?
A
Routine data need to be reliable, valid, and complete for accurate epidemiological analysis.
10
Q
What are methods to improve the quality of routine data?
A
- Computerised data collation and analysis
- Better feedback of data to providers
- Meaningful presentation of data
- Improved training for data entry
11
Q
Define descriptive epidemiology.
A
Descriptive epidemiology is the study of the patterns, causes, and control of disease in groups of people.
12
Q
What are the three principal dimensions used in descriptive epidemiology?
A
- Time
- Place
- Person
13
Q
What does the numerator represent in epidemiological terms?
A
The numerator is a feature that has been counted, such as the number of deaths.
14
Q
What is a denominator in epidemiology?
A
The denominator is usually the population from which the numerator was drawn.
15
Q
What is the formula for calculating a ratio?
A
Ratio = n1 : n2
16
Q
What is a proportion in epidemiology?
A
Proportion = n1 / N, where n1 is a subpopulation of the whole study population.
17
Q
What does a rate measure in epidemiology?
A
A rate measures the frequency with which a phenomenon occurs, often expressed in person-years.
18
Q
What does ‘time at risk’ refer to?
A
Time at risk describes the total amount of time individuals within a study spend being at risk of developing the disease.
19
Q
What is right censoring in cohort studies?
A
Right censoring occurs when subjects develop the disease and stop being at risk.
20
Q
What is left censoring in cohort studies?
A
Left censoring occurs when subjects join the study after the event of interest has already occurred.
21
Q
What is total person-time at risk?
A
Total person-time at risk is the sum of all the individual times at risk in a study.
22
Q
How is the incidence rate calculated using person-time at risk?
A
Incidence rate = Number of new cases / Total person-time at risk.
23
Q
What is the importance of defining populations at risk?
A
Defining populations at risk ensures that only individuals who could potentially develop the disease are included.
24
Q
What is the preferred method for calculating the incidence rate when study duration is longer?
A
Using person-time at risk
This method accounts for the number of individuals at risk over time, especially as individuals may die, become cases, or be lost to follow-up.
25
What is the formula for calculating the mortality rate?
Mortality rate = Number of deaths in study period / Person time at risk during study period
## Footnote
An example calculation shows a mortality rate of 0.95 per person year or 95 per 1000 person years.
26
What are the two types of data mentioned in the context of summarising data?
Discrete data and continuous data
## Footnote
Discrete data have finite values, while continuous data include measurable quantities.
27
What defines nominal data?
Categorical data with no order
## Footnote
Examples include country and blood group.
28
What distinguishes ordinal data from nominal data?
Ordinal data have an order but uncertain intervals
## Footnote
Examples include Likert scales and social class.
29
What are interval data?
Quantitative data with meaningful intervals but no true zero
## Footnote
An example is temperature in degrees Celsius.
30
What are ratio data?
Data with interval properties and a true zero
## Footnote
Examples include weight and Kelvin temperature.
31
What are binary data?
A special type of discrete data with only two possible values
## Footnote
Common in epidemiology to describe states like alive/dead.
32
What does incidence measure?
The new occurrence of a disease
## Footnote
It includes incidence and cumulative incidence.
33
What does prevalence measure?
Existing occurrences of a disease
## Footnote
Point prevalence is the proportion of a population with a disease at a specific time.
34
What is standardisation in epidemiology?
A technique for comparing populations with differing demographic structures
## Footnote
It helps avoid misleading comparisons of crude rates.
35
What are the two forms of standardisation?
Direct standardisation and indirect standardisation
## Footnote
Direct standardisation uses known age-specific rates; indirect uses expected rates.
36
What is the formula for cumulative incidence?
Proportion of a population who become diseased in a defined time period
## Footnote
It is also referred to as risk.
37
What is the definition of point prevalence?
Proportion of a population with a disease at a particular point in time
## Footnote
It reflects the current state of a population regarding a disease.
38
What is the purpose of direct standardisation?
To compare two populations using age-specific mortality rates
## Footnote
It involves applying these rates to a standard population.
39
What is the formula for Standardised Mortality Ratio (SMR)?
SMR = Observed deaths / Expected deaths
## Footnote
It compares actual deaths to expected deaths based on a standard population.
40
What does Years of Life Lost (YLL) measure?
Premature mortality and its impact on society
## Footnote
It emphasizes deaths occurring in younger individuals.
41
How is YLL calculated in its simplest form?
By subtracting the age at death from an upper age limit
## Footnote
For example, a person dying at 60 contributes 15 YLL if the limit is 75.
42
What does YLL stand for?
Years of Life Lost
## Footnote
YLL is a measure used to quantify the impact of premature mortality on a population.
43
How is YLL calculated in its simplest form?
By choosing an upper age limit (e.g. 75) and subtracting the age at death from this limit.
44
What is HALE?
Health-Adjusted Life Expectancy
## Footnote
HALE accounts for the quality of life years lived, rather than just counting the years.
45
What is the formula used to calculate total YLL in a population?
Σ(Deaths at each age × (Upper age limit - Age))
46
What do event-based measures of disease burden require?
Data from routine sources such as death certificates, hospital episode data, disease registers, and statutory notifications.
47
What is a limitation of using health service data to assess disease burden?
It may underestimate the burden because most care occurs outside the formal healthcare sector.
48
What are time-based measures of disease burden?
Cross-sectional prevalence surveys used when routine event-based data are not available.
49
What factors should be considered when choosing comparison groups in public health?
* Size of population
* Area characteristics (rural/urban)
* Age structure
* Ethnicity
* Socio-economic characteristics
* Disease burden
* Health service usage
* Health service provision
50
What is sampling variability?
Differences between sample estimates and the true population value due to random chance.
51
What is the definition of bias in epidemiological studies?
Differences that cause individuals to be systematically classified differently from others, unrelated to the exposure of interest.
52
What is confounding in epidemiological research?
When an apparent association between an exposure and a disease is actually due to a third factor.
53
True or False: Reverse causation occurs when the outcome causes variation in the exposure.
True
54
What is measurement error in epidemiology?
Any mistake that occurs in applying a standard set of values to observations.
55
What are the two types of measurement errors?
* Random error
* Systematic error
56
What is the impact of random errors?
They affect all groups equally and are less threatening to a study.
57
What is systematic error?
Persistent, non-random differences that affect some groups more than others.
58
What is risk in an epidemiological context?
The likelihood of an event, calculated as the proportion of individuals who develop the disease within a study period.
59
What is attributable risk (AR)?
The difference in the risk or rate of disease between the exposed and unexposed groups.
60
What is the formula for calculating Attributable Fraction (AF)?
AF = AR / Re
## Footnote
Where Re is the risk in the exposed group.
61
What does Population Attributable Risk (PAR) measure?
The excess rate of disease in the population attributable to the exposure.
62
What is the definition of odds ratio?
The odds of exposure in cases compared to the odds of exposure in controls.
63
What is the risk ratio?
The risk of disease in the exposed group divided by the risk of disease in the unexposed group.
64
What is a rate ratio?
The incidence rate in the exposed divided by the incidence rate in the unexposed.
65
What is the formula for calculating the odds ratio in a case-control study?
OR = (a/c) / (b/d)
## Footnote
'a', 'b', 'c', and 'd' represent the counts of exposed and unexposed cases and controls.
66
How is the risk ratio calculated?
Risk ratio = (Risk of disease in exposed) ÷ (Risk of disease in unexposed)
## Footnote
This formula compares the likelihood of disease between two groups.
67
What does an odds ratio greater than 1 indicate?
It indicates that the exposure is associated with a higher odds of the outcome occurring.
68
What is the relationship between odds ratio and risk ratio when the disease is rare?
The odds ratio provides a good approximation of the risk ratio.
69
Define 'association' in the context of epidemiology.
Association is a link between two variables.
70
What are the potential reasons for an observed association?
* Chance
* Bias
* Confounding
* Reverse causation
71
What is chance in epidemiological studies?
Chance refers to the possibility that an observed association in a sample does not reflect the wider population.
72
What is bias in epidemiological studies?
Bias is a systematic error that leads to differences between comparison groups.
73
What is selection bias?
Selection bias occurs when there are systematic differences between study participants and non-participants.
74
What is reverse causality?
Reverse causality occurs when an outcome causes a change in exposure.
75
What are Bradford Hill's criteria for determining causality?
* Strength of association
* Biological credibility
* Consistency of findings
* Temporal sequence
* Dose–response
* Specificity
* Coherence
* Reversibility
* Analogy
76
What is confounding?
Confounding is a special form of bias where the association between an exposure and an outcome is influenced by another variable.
77
What types of bias fall under information bias?
* Instrument bias
* Responder bias
* Observer bias
78
How can bias be minimised in studies?
* Randomisation
* Blinding
* Training of study personnel
* Written protocols
* Choice of controls
79
What is a mediating factor?
A mediating factor is associated with both exposure and outcome but is not independently associated with the outcome.
80
What does positive confounding do to an association?
Positive confounding makes the association appear more pronounced.
81
What is residual confounding?
Residual confounding occurs from unknown confounding factors or inaccurately measured confounders.
82
What does a risk ratio of 2.86 indicate in a clinical context?
Patients who had a coronary bypass operation were 2.86 times as likely to have hypertension compared to those who did not.
83
What is the role of p-values in assessing chance?
P-values help determine the likelihood that an observed association is due to chance.
84
What is the healthy worker effect?
This bias occurs when employed individuals are generally healthier than the general population, affecting study outcomes.
85
Fill in the blank: The relationship between alcohol and cardiovascular disease can show a ________ curve.
J-shaped
86
What is a mediating factor in the relationship between diet and coronary heart disease?
Serum cholesterol
## Footnote
Serum cholesterol is a key factor that mediates the causal relationship between diet and coronary heart disease.
87
What does residual confounding refer to?
Residual confounding occurs due to unknown confounding factors or inaccurately measured confounding factors
## Footnote
It happens even after accounting for known confounders.
88
How can randomisation help in research studies?
Randomisation distributes both known and unknown confounding factors equally between groups
## Footnote
This effectively eliminates residual confounding.
89
What is effect modification?
Effect modification occurs when the effect on the outcome of one causal factor varies according to the level of a third variable
## Footnote
This third variable is termed an effect modifier.
90
What is an example of an effect modifier in myocardial infarction risk?
Age
## Footnote
The effect of smoking on the risk of myocardial infarction is stronger in young people, indicating age as an effect modifier.
91
What is the chi-squared test for interaction used for?
To test for heterogeneity between stratum-specific estimates
## Footnote
It provides a p-value indicating the likelihood that observed heterogeneity is due to chance.
92
What are the main techniques for dealing with confounding at the design stage?
* Randomisation
* Restriction
* Matching
## Footnote
Each technique has its advantages and disadvantages.
93
What is stratification in the context of confounding?
Stratification divides confounding variables into strata and provides stratum-specific relative estimates
## Footnote
It offers a weighted-average overall single estimate of the confounding effect.
94
What is a descriptive study?
A descriptive study describes patterns of disease with regard to time, person, and place
## Footnote
It includes case reports and case series.
95
What characterizes ecological studies?
The unit of observation is a group rather than an individual
## Footnote
They use group-level data and are useful for hypothesis formulation.
96
What is a major weakness of case reports in descriptive studies?
Case reports cannot be used to test for a valid statistical association
## Footnote
The observation may just be a coincidence.
97
What does ecological fallacy refer to?
An error of logic that occurs when inferences about individuals are made based on aggregate population data
## Footnote
It can lead to incorrect conclusions.
98
What is the relationship between new immigrants and literacy as described in the ecological fallacy?
New immigrants were found to be less literate than citizens, but they settled in states with higher average literacy
## Footnote
This illustrates the ecological fallacy where aggregate data misrepresents individual characteristics.
