Upgrade to remove ads
Only $35.99/year

Epi Final Exam

Terms in this set (67)

Case-control study
compare people with the disease and without disease
Odds Ratio (OR)
a measure off the strength of the relationship between 2 variables
- Used for cross sectional, case control
- Not ideal
- May create a stronger associate than actual exists
Observation Studies
no manipulation of exposure or randomization
Strength of studies order from strongest to weakest
Experimental, cohort, cross-sectional & case control
Selection of Controls
The ideal controls should have the same characteristics as the cases (except for the exposure of interest).
When to conduct a case-control study
May be an early step in the study of a disease or outcome
May be part of an outbreak investigation
Case-control study strengths
Inexpensive
Fast
Good for studying rare outcomes
Some evidence of temporal order
Case-control study limitations
Confounding
Unclear temporal order data
Memory failure
Poor estimate of risk
Unrepresentative participant group
Poor at studying rare exposures
Requires care in selecting subjects
cross-sectional studies
Exposure and disease are measured at the same time
Study prevalence not incidence
cross-sectional study strengths
Inexpensive
Fast
Provide prevalence data
Can study rare exposures
cross-sectional study limitations
Confounding
Undetermined temporal order
Poor at studying mortality
Poor at studying incidence
Poor at studying rare outcomes
Ecologic studies
a study in which the units of analysis are populations or groups of people rather than individuals
Ecologic studies strengths
Inexpensive
Fast
Simple
Easy - data are often readily available
Ecologic studies limitations
Confounding
Undetermined temporal order
Ecologic fallacy
Imprecise/inaccurate data
Limited by available data
OR = 1
implies no association. [Null value
OR > 1
1 [sick] people with the outcome were more likely than [healthy] people without the outcome to be exposed
OR < 1
◦[healthy] people without the outcome were more likely than [sick] people with the outcome to be exposed.
◦[sick] people with the outcome were less likely than [healthy] people without the outcome to be exposed.
experimental study
the researcher manipulates one of the variables and tries to determine how the manipulation influences other variables
Radomization
The best defense against bias is randomization, in which each individual is given a fair, random chance of selection.
quasi-experiment study
studies of groups of subjects who have been exposed to the conditions in the real world
advantage: can study subjects and effects can't study ethically or practically with experiments
drawback: potentially confounding variables
Blinding/Masking
A procedure in which one or more parties to the trial are kept unaware of the treatment assignment.
single-blind study
study in which the subjects do not know if they are in the experimental or the control group
double-blind study
An experiment in which neither the participant nor the researcher knows whether the participant has received the treatment or the placebo
triple blind study
A study performed where neither the treatment allocator nor the participant nor the response gatherer knows which group the participant is in
Types of experiments
clinical trials, behavioral experiments, and community interventions
confounding
occurs when two variables are associated in such a way that their effects on a response variable cannot be distinguished from each other
cohort studies
the study of a group of people who share a common characteristic and move forward in time as a unit
Basic design of cohort studies
Observational studies
A group with the exposure and a group without the exposure are followed
Incidence of disease in the groups is compared
Cohort studies baseline
◦Usually excludes people with the disease at baseline (start of the study)
temporal relationship
order in time
Cohort studies find incident (new) cases of disease
◦Exposure occurs before disease
The exposure is not a possible cause of the disease unless it occurs before the disease
◦This is a necessary but not sufficient condition
Ideal condition for a cohort study
Short period of time between exposure and disease
◦E.g., pregnancy outcomes
Cohort study limitations
Long time to see results
Expensive
Possible temporal order problem
Loss to follow up
Difficulty studying rare outcomes
Confounding
Prospective cohort study
A research study that follows over time groups of individuals who are alike in many ways but differ by a certain characteristic (for example, female nurses who smoke and those who do not smoke) and compares them for a particular outcome (such as lung cancer).
Limitations of prospective cohort studies
May take a long time for disease to develop
◦Participants may develop and die from other diseases first
◦Participants may outlive investigators
- expensive
Retrospective cohort study
A study that uses an intact cohort of known offenders and looks back into their early life experiences by checking their educational, family, police, and hospital records.
Retrospective cohort study disadvantages
Confounding
Poor control of measures
Poor control of cohort
Possible temporal order problem
Difficulty studying rare outcomes
Loss to follow up
Cohort studies investigate new cases of the disease. t/f
TRUE
Cross-sectional and case-control studies investigate all existing cases of the disease. T/F
TRUE
Incidence rate ratio
incidence rate in exposed/incidence rate in unexposed
confidence interval
Range of likely values of a population characteristic (such as mean, probability or odds ratio) for which only a sample estimate is available
95% confidence interval
the range within which the true population mean lies, with 95% certainty
bias
a systematic prejudice in one direction
Risk factors
characteristics that predict the chances for developing a certain disease
example of implicit question
what is the association between sun exposure and skin cancer?
example of research hypothesis (positive declaration)
this is an association between sun exposure and skin cancer
example of null hypothesis (negative declaration)
there is no association between sun exposure and skin cancer
how does prevalence change?
- incidence increases and duration does not change = prevalence increases
- incidence does not change and duration decreases = prevalence decreases
rate
ratio that consists of a numerator and a denominator and which TIME forms part of the denominator
incidence
new cases
example of incidence
Number of persons who died and whose deaths were attributed to Hurricane Floyd in North Carolina in 1999
prevalence
all cases
Example of prevalence
Number of Children who have immunity to measles, either because they had the disease or because they received the vaccine.
The formula for point prevalence is
# current cases / # persons in population.
incidence density
new cases / person-time risk
example of incidence density
10,000 (given) x (305/48372.5) = 63.05 per 10,000 person-years at risk
cumulative incidence
10n (given) x (new cases/number at risk)
After the second game of the college football season, 60 members of the 97-person football team developed fever, malaise, loss of appetite, and abnormal discomfort. Within a few days, 30 of the players became jaundiced. Blood samples were drawn from all members of the team to test for antibodies to hepatitis A (the presumptive diagnosis) and to test for elevation of liver enzymes.

What is the cumulative incidence of jaundice?
30/97
surveillance
Ongoing systematic collection, analysis, and interpretation of data
Decision to conduct surveillance for a conditions is based on:
- Public health importance
- Extent to which it be controlled, prevented, or treated
- Ability of the health system to control, prevent, or treat the condition
Isolation
separates people with a contagious disease from people who are not sick.
Quarantine
people who were exposed to a contagious disease are separated from others and their movement is restrict to see if they become ill
Causes of information bias
- Low sensitivity or specificity of a diagnostic procedure
- If data are collected using interviews or surveys, people correctly remember their exposure or disease history
Reporting bias
If cases (or controls) are less likely to report exposures/behaviors that are seen as social undesirable or because of the perceptions or beliefs about the behavior
A confounder is
- A risk factor for the disease
- Associated with the exposure
- Not caused by the exposure
Statistical inference
- Drawing conclusions about a population parameter (statistic) based on an estimate obtained from a sample.
- only accounts for random error (not systematic error)
Hypothesis testing
Methods to assess (test) to reject or fail to reject a null hypothesis about a population parameter (e.g., relative risk or odds ratio
p-value
how unlikely the observed results are if the null hypothesis is true
Students also viewed