Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
65 Cards in this Set
- Front
- Back
|
What are the dimensions of Epidemiologic Research?
|
Population - a set of members
Time - temporal aspect |
|
|
Epidemiology is...
|
Epidemiology is the discipline concerned with the study of the frequency of occurrence of illness (disease, defect or injury)
|
|
|
John Graunt
|
Published "The Bills of Mortality" in 1662, which had compiled data on birth and age at death.
|
|
|
William Farr
|
Estimated rates of birth and deaths in 1800s using census data on the population size.
|
|
|
John Snow
|
London physician in 1800s, studied cholera outbreak and reasoned it came from contaminated water. Studied death from cholera based on company of water supply (two water supply companies in London, randomly distributed. Natural study.)
|
|
|
Define public health
|
-The health of a community of people.
-Care for the health of a community of people. -Social aspects of health and healthcare (personnel, organization, finance, program policy). |
|
|
What is descriptive Epidemiologic research?
|
Research that provides descriptive information on patterns of occurrence of illness, such as age, sex, geographic location, time, etc.
|
|
|
What is scientific Epidemiologic research?
|
Research that aims to advance theoretical knowledge about nature and pertains to causality.
|
|
|
What are the two types of Scientific Epidemiologic Research?
|
Etiologic
(The causal origin of illness) Prognostic (Prevention of illness or amelioration of future course of illness) |
|
|
What is a "cause"?
|
"An event, condition, or characteristic that preceded the illness event and without which the illness event either would not have occurred at all or would not have occurred until some later time”
(- Rothman) |
|
|
Types of source populations
|
Dynamic, Fixed
|
|
|
Describe a dynamic population
|
-Turnover of membership
-Open population -Membership defined by a state for the duration of the state -The state is changeable Ex. Residents of Boston Patients of MGH Network Blue insurance carriers |
|
|
Describe a fixed population
|
Membership is defined on the basis of some event, forever after
Ex. Babies born to pregnant women enrolled in a study People dying in a particular region in a particular span of time |
|
|
What is a cohort?
|
A population whose membership is defined on the basis of some event, (often enrollment into a particular study).
|
|
|
What is Prevalence?
|
The existence of a particular state among members of a population
(Prevailing or existing illness) The proportion on a population in the particular state during a specified time period. |
|
|
What is Incidence?
|
The appearance of an event among members of a population
(Onset/diagnosis of illness, resolution of illness) The occurrence of an illness event among members of a population in a particular time period. |
|
|
Distinguish between Point and Period Prevalence
|
Point prevalence:
Prevalence at a specific point in time Ex: At cohort enrollment, At birth, on Dec 31, 1999. Period prevalence: over an interval of time Ex: 20 to 25 years of age, Jan-Dec 2010 |
|
|
Prevalence Formula
|
P = # of cases in time period / # in population in time period
|
|
|
What units are associated with prevalence?
|
None!
Answer will be between 0-1 or 0-100% |
|
|
Is time associated with prevalence?
|
The relevant time period does not appear in the formula but MUST be specified!
|
|
|
Is a proportion a type of rate?
|
Yes
|
|
|
What is the Incidence Rate formula?
|
IR = # of individuals who develop the illness event / Total time experienced by the population followed
|
|
|
What are the two types of Incidence?
|
Cumulative Incidence and Incidence Rate
|
|
|
What is Incidence Rate?
|
The ratio of the number of new events to the corresponding population time.
|
|
|
What is included in Follow-Up time?
|
Only the time at risk of developing the event
|
|
|
What is Mortality Rate?
|
Incidence Rate in which the event of interest is death
|
|
|
When is Incidence Rate usually estimated?
|
-In dynamic populations
-In fixed populations with considerable losses to follow-up (because it takes follow-up time changes into account) |
|
|
Is the numerator included in the denominator for incidence rate?
|
No
|
|
|
Are there units for Incidence Rate and if so what are they?
|
cases/person-time
|
|
|
What is Cumulative Incidence?
|
The proportion if individuals in a population who develop the event of interest in a specified time period.
(The population is free of the event at the start of follow-up!_ |
|
|
What is the Cumulative Incidence formula?
|
# of individuals who develop the illness event during the time period / # of individuals followed during the time period
|
|
|
When is cumulative incidence usually estimated?
|
In fixed populations with relatively complete follow-up
(Assumes everyone in population is followed for specified time period) |
|
|
Are there units for Cumulative Incidence and if so what are they?
|
No units!
(Relevant time period does not appear in formula but must be specified to be interpretable) |
|
|
What are other names for Cumulative Incidence?
|
-Incidence proportion
-Risk |
|
|
What is the mathematical relationship between Cumulative Incidence and Incidence Rate when IR is constant?
|
CI = 1-e^(-IR*time)
|
|
|
What is the mathematical relationship between Cumulative Incidence and Incidence Rate when IR is NOT constant?
|
CI = 1-e^(Sum:-IRk*timek)
across intervals k=1 to n |
|
|
What affects prevalence?
|
-Frequency of illness onset
-Duration of illness (e.g. disease that is transmitted easily might have low prevalence if recovery time is short) |
|
|
What is the mathematical relationship between Prevalence and Incidence?
|
Prevalence odds = incidence rate * average duration of illness
P=(IR*D)/(1+IR*D) |
|
|
What conditions are necessary to estimate prevalence from incidence?
|
Steady State conditions:
-Incidence rate is constant over time (stable) -Duration of illness is stable over time -Inflow of illness = outflow of illness |
|
|
What is the determinant (exposure)?
|
A characteristic of individuals on which the frequency of occurrence of illness depends either causally or non-causally.
|
|
|
What are the different types of determinants?
|
-Constitutional
-Behavioral -Environmental |
|
|
What is a constitutional determinant?
|
A determinant having to do with inherent characteristics (acquired, inherited, congenital)
E.g. Genetic polymorphism, birth defect, age |
|
|
What is a behavioral determinant?
|
A determinant having to do with the actions of the individual
E.g. Smoking, medication use, nutrition |
|
|
What is an environmental determinant?
|
A determinant having to do with the external conditions or surroundings of an individual
E.g. Air quality, in utero exposure to a particular medication |
|
|
What is an index group?
|
Individuals with a particular characteristic of interest
(Often referred to as the exposed group) |
|
|
What is a reference group?
|
Individuals without the particular characteristic
-An alternative to the characteristic of interest, often referred to as the unexposed group. |
|
|
What does a normal 2x2 table look like?
|
-Determinant (Index and Reference in that order) in columns
-Rows = cases then denominator |
|
|
How do we assess differences in the frequency of occurrence of illness in the index and reference groups?
|
-Absolute difference (subtraction)
or -Relative difference (division) |
|
|
What are absolute measures of association?
|
The difference between the measures in index and reference groups
|
|
|
What are relative measures of association?
|
The ratio of the measures in index and reference groups.
|
|
|
Absolute measures of association (mathematical formulas)
|
CI (Risk) Difference:CId=CIi-CIr
IR difference: IRd=IRi - IRr P difference: Pd=Pi-Pr |
|
|
What are the units for absolute measures of association?
|
CI diff --> none
IR diff --> Person-time P diff --> none |
|
|
What is the interpretation of absolute measures of association?
|
=0: no association between determinant and illness
>0: determinant associated with increased risk/incidence/prevalence of illness <0: determinant associated with decreased risk/incidence/prevalence of illness |
|
|
Relative Measures of Association (mathematical formulas)
|
CI ratio: CIR = CIi/CIr
Incidence Rate ratio: IRR = IRi/IRr Prevalence ratio: PR=Pi/Pr |
|
|
Relative measures of association interpretation
|
=1: No association between determinant and illness
>1:determinant associated with increased risk/incidence rate/prevalence <1: determinant associated with decreased risk/incidence rate/prevalence |
|
|
What is a descriptive association?
|
The frequency of illness is related to the determinant without any view to causal interpretation
|
|
|
What are examples of descriptive variables?
|
Age, sex, race
|
|
|
What is a causal association?
|
The frequency of illness is related to the determinant and may be viewed in causal terms
|
|
|
What is the counterfactual experience?
|
The ideal reference group (not possible): the same people at the same time without the characteristic of interest
|
|
|
What is the attributable fraction?
|
The excess illness that is caused by the determinant (exposure)
AF = (CIi-CIr)/CIi AF = (RR-1)/RR |
|
|
What is the population attributable fraction?
|
The proportion of cases in the POPULATION caused by the exposure
PAF = Pe*AF where Pe=# of exposed cases/total number of cases |
|
|
What is the sufficient-component cause model?
|
Sufficient cause = a complete causal mechanism which comprises the minimal set of conditions and events that are sufficient to cause illness.
|
|
|
What is association?
|
The relation of an occurrence parameter to one or more characteristics of persons
Association DOES NOT equal causation |
|
|
How do we assess causal association?
|
No criteria exist to establish the validity of evidence, but can assess validity through criticism with the goal of evaluating the total error that affects the study.
|
|
|
What are some considerations to assess causal inference?
|
-What is the likelihood of random error?
-What is the magnitude of estimate of effect and range of possible values? -What is the likelihood of systematic error -Ultimately rests on the acceptance of the scientific community. |
