Epidemiology

Front 1

Basic Ethical Principles

Back 1

Beneficence
Respect for Persons
Justice

Front 2

Basic Ethical Principle:
-do not harm

Back 2

Beneficence

Front 3

Basic Ethical Principle:
-a person is autonomous and has the right to make his own decisions

Back 3

Respect for Persons

Front 4

Basic Ethical Principle:
-the risks and benefits of research are equally distributed, and no one population should be burdened while another receives the benefits

Back 4

Justice

Front 5

Person:
-external and internal environment explains diseases
-defines “epidemic vs endemic”

Back 5

Hippocrates

Front 6

Person:
-external and internal environment explains diseases
-defines “epidemic vs endemic”

Back 6

Farr

Front 7

Person:
-published “The Nature and Political Observations Made Upon the Bills of Mortality”
-founder of medical stats, births, deaths, by season, year and parish

Back 7

Graunt

Front 8

Person:
-first planned clinical trial
-treated sailors with scurvy

Back 8

Lind

Front 9

Person:
-first to propose and test a hypothesis: cholera is transmitted by a contaminated water supply
-father of modern epidemiology

Back 9

Snow

Front 10

Type of Epidemiology:
-hypothesis generating
-person, place, time

Back 10

Descriptive Epidemiology

Front 11

Type of Epidemiology:
-hypothesis testing
1.Cohort
2. Case-Control
3. Intervention Studies

Back 11

Analytic Epidemiology

Front 12

-grouped based on exposure status

Back 12

Cohort Studies

Front 13

-grouped based on disease status

Back 13

Case-Control

Front 14

-case-control within cohort study

Back 14

Nested Case-Control

Front 15

-cases serves as own controls

Back 15

Case-Crossover

Front 16

Advantages:
-rare diseases
-relatively fast
-relatively inexpensive
Disadvantages:
-cannot establish temporality
-subject to recall, observer and selection biases
-selection of controls difficult
-bad for rare exposures

Back 16

Case-Control Studies

Front 17

Advantages:
-can establish temporality
-direct calculation of incidence rates
-no recall bias
Disadvantages:
-subject to attrition bias, exposure, and outcome misclassification
-relatively expensive (large sample size)
-relatively time consuming (follow-up)

Back 17

Cohort Studies

Front 18

Disease Pattern
-i.e. food outbreak

Back 18

Short-term fluctuations

Front 19

Disease Pattern
-long term changes in morbidity and mortality patterns

Back 19

Secular Trends

Front 20

Disease Pattern
-periodic, often predictable, changes in frequency of morbidity/mortality

Back 20

Cyclic Patterns

Front 21

Disease Pattern
-a rise in disease that is clearly in excess of what is normally expected

Back 21

Epidemic

Front 22

Disease Pattern
-constant presence or usual frequency of a disease

Back 22

Endemic

Front 23

Using group-level associations to make conclusions about individual-level associations
(for example, association found when national salt consumption and national heart disease levels are compared across countries cannot be used to say eating more salt leads to heart disease)

Back 23

Ecological Fallacy (Aggregation Bias)

Front 24

(# death from a specific disease/ # cases over same time period) x 10^n

Back 24

Case Fatality Rate

Front 25

(# of new cases with C - # deaths from x/ # cases with the same disease) x 10^n

Back 25

Survival Rate

Front 26

(# existing cases at a point in time/# at risk at that point in time)

Back 26

Point Prevalence – “Snapshot”

Front 27

(# existing cases at beginning + new cases over time period/ # in population (or in study))

Back 27

Period Prevalence

Front 28

New cases in a population over a time period

Back 28

Incidence

Front 29

Risk of developing a disease:
(# new cases/ # at risk)
-exclude prevalent cases

Back 29

Cumulative Incidence

Front 30

Frequency of disease development
(# new cases/ total person-years at risk)

Back 30

Incidence Rate/ Incidence Density (ID)

Front 31

(CI in exposed/ CI in unexposed)

Back 31

Risk Ratio

Front 32

(ID in exposed/ ID in unexposed)

Back 32

Rate Ratio

Front 33

(Odds of disease among exposed/ Odds of disease among unexposed)

Back 33

Odds Ratio

Front 34

(Prevalence Rate among exposed/ Prevalence Rate among unexposed)

Back 34

Prevalence Rate Ratio

Front 35

Compares the observed amount of D among exposed cohort to what would be exposed in an unexposed cohort- indirect adjustment; estimates increased risk of one group over another
(observed deaths/expected deaths) x 100

Back 35

Standardized Mortality Ratio

Front 36

Proportion of deaths from a specific cause relative to all deaths in the exposed cohort is compared with the proportion of deaths from a specific cause relative to all deaths in the unexposed cohort
(prop. (%) deaths from X (cohort)/ prop. Deaths from X (comparions)

Back 36

Proportional Mortality Ratio

Front 37

Evaluating Validity:
-not due to chance alone

Back 37

Precision

Front 38

Evaluating Validity:
-study results are true, real unbiased, lack systematic error

Back 38

Internal Validity

Front 39

Evaluating Validity:
-study results are generalizable to your target population

Back 39

External Validity

Front 40

Two methods to evaluate chance

Back 40

Hypothesis Testing
Confidence Interval

Front 41

-P-value
-probability that your study result is different from null solely due to chance
-what is the chance that there is truly no association in the population but by some luck of the sample draw, you managed to see the study result you did
-Cannot tell the magnitude of the association

Back 41

Hypothesis Testing

Front 42

-relative magnitude of the association
-general precision: wider, less precise
-tells you whether the result is statistically significantly different from null/no association

Back 42

Confidence Interval

Front 43

If the CI does NOT include the measure of no association, then it is __________

Back 43

Significant

Front 44

If the CI does include the measure of no association, then it is __________

Back 44

Not Significant

Front 45

Probability of finding an association if one exist

Back 45

Power

Front 46

Bias:
-systematic error that results from the way subjects are selected, recruited or retained in a study

Back 46

Selection Bias

Front 47

Bias:
-systematic error that arises from inaccurate measurements or misclassification of subjects with regard to exposure or outcome status

Back 47

Measurement

Front 48

-systematic error that occurs because of differential recall about pass exposure status between those who have disease and those who do not
-occurs in case-control and cross-sectional studies

Back 48

Recall Bias

Front 49

-related to BOTH exposure and disease, and it is NOT on the exposure-disease casual pathway

Back 49

Confounder

Front 50

Eliminate Confounding in Design Phase

Back 50

Randomization
Restriction
Matching

Front 51

Eliminate Confounding in Analysis Phase

Back 51

Stratification
Multivariate Analysis
Direct/Indirect Standardization or Adjustment