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;
84 Cards in this Set
- Front
- Back
Totality of evidence
|
you can't just use one set of data in order to answer a research question
|
|
What is an example of EXPOSURE in a study?
|
the effect of EXERCISE on depression.
(exercise= exposure) |
|
What is an example of OUTCOME in a study?
|
reduced risk of DEPRESSION
(depression is the outcome) |
|
when you design a study what are 4 components that you need?
|
exposure, outcome, comparison group and hypothesis
|
|
What are the two types of epidemiology studies?
|
experimental and observational
|
|
What are the key differences between EXPERIMENTAL and OBSERVATIONAL studies?
|
In EXPERIMENTAL, the investigator ONLY OBSERVES the exposures
EXPERIMENTAL: he controls/manipulates the variables |
|
What is a RANDOMIZED CONTROL TRIAL?
|
investigator gives cohort exposure and then studies the people (used in experimental)
|
|
What are some strengths of the RCT study?
|
a) theres a comparison group
b) lots of people therefore only dif. is the exposure!! b1) and therefore you can control the exposure c) you can know if one thing caused another (temporal) |
|
What are some weaknesses of the RCT?
|
1. expensive
2. ethics 3. you can't just generalize because everyone is different |
|
What is a QUASI-EXPERIMENTAL STUDY?
|
(similar to RCT)
1. Nt controlled by investigator so you look at exposure of people but its not random and its not controlled by the investigator |
|
Review: what are the two types of experimental studies?
|
RCT (random, inves. controls exposure)
QUASIexperimental (nt random or controlled by investigator) |
|
What are key components of observational studies?
|
1. common
2. natural (we just observe ppl) 3. collect data and analyze results (its supposed to be like an experimental study) 4. DESCRIPTIVE OR ANALYTIC |
|
what are the two types of observational studies?
|
descriptive and analytic
|
|
what are the components of descriptive studies?
|
1. questions like who what where and why. (who is getting the disease and why are they getting it?)
|
|
what are the different types of descriptive studies?
|
1. case reports
2. cross sectional 3. corelational |
|
what are the components of analytic studies?
|
questions like WHY?
|
|
what are the different types of analytic studies?
|
1. cohort
2. case control |
|
what are descriptive studies used for?
|
1. set up study to then be used in analytic studies
2. DESCRIBE what exists in a POPULATION 3. Uses person, place, or time |
|
What is a CASE REPORT? (and what type of study is this?)
|
1. MOST BASIC descriptive study
2. DETAILED report on 1 patient or a few patients 3. ******** LIMITATION: THERE IS NO COMPARISON GROUP 4. can alert people as to the relationship between exposure/outcome relationship |
|
what is an example of a case report?
|
OC'S. do they cause liver cancer? maybe, but there is also NO COMPARISON GROUP!
|
|
What is a cross sectional study?
|
SNAPSHOT.
a. nt based on outcome or exposure status |
|
what are some strengths of cross-sectional studies?
|
1. less time
2. no follow up involved |
|
what are some limitations of a cross-sectional study?
|
its hard to test your hypothesis because you are not sure if the exposure causes the disease or if the disease causes the exposure (this is called a temporal relationship)
|
|
what is one way that a cross sectional study could test a hypothesis?
|
if the exposure can't be changed (example: if blue eyes causes depression)
|
|
ecological or correlational study
|
POPULATIONS!!!!!!!!!
uses data from populations to compare disease frequencies at one time or at dif. times |
|
what are some strengths of an ecological study?
|
1. it is quick and inexpensive
|
|
what are some limitations of an ecological study?
|
1. its from a population nt an individual
2. there are other things at play here 3. these are averages |
|
what is an analytical study? and what is its purpose?
|
goal: to determine an exposure outcome w.o using too many resources
|
|
An analytical study is also used to test a priori hypothesis. What is a priori hypothesis?
|
there are two types:
priori and ad-hoc. priori: is testing the hypothesis post hoc: making a hypothesis after you have already done the experiment |
|
what is a cohort study?
|
self-selected exposures.
person is free of the outcome but is studied OVER TIME |
|
What are the two types of cohort studies?
|
prospective and retrospective
in BOTH: THE INVESTIGATOR STARTS W. THE EXPOSURE |
|
what is a prospective cohort study?
|
only OUT COME FREE INDIVIDUALS WHO ARE AT RISK
limits: can take a long time |
|
what is a retrospective study?
|
exposure and outcome occur before the study
|
|
what are some strengths of a cohort study?
|
1. you can differentiate temporality
2. you don't have recall bias (b/c you are assessing patients before the outcome develops) 3. you can examine lots of outcomes |
|
what are some limitations of a cohort study?
|
1. you need lots of people
2. you need a lot of time 3. it is very expensive |
|
what is a case control study?
|
1. both exposure and outcome have occurred at the start of the study
2. you don't even know the exposure; you are just assessing the outcome 3. the people w. the outcome come from a large population heres what you do: you take a population that has the outcome and ask them what their exposure was. then you take a population that DOESN'T have the outcome and you ask what the exposure was. |
|
what are the strengths of a case control study?
|
1. its good for rare outcomes
2. you can examine lots of exposures 3. better than cohort b/c you can deal w. long induction/latent period. |
|
what are the limitations of a case control study?
|
1. more bias
2. you are not sure of the temporal relationship 3. its not so good if the exposure is rare 4. how do you know if you are getting the accurate exposure over the certain time period? |
|
how should you choose your study design? what are the factors?
|
1. your question
2. how much you already know 3. frequency of exposure 4. frequency of outcome 5. validity and frequency considerations 6. practical and ethical concerns |
|
When should you pick an observational study?
vs. when should you pick an experimental study? |
observational: when you want to understand prevention, treatment or etiology
be aware: this is expensive and not ethical experimental: when you want to understand treatment, prevention take note: this is not expensive and it is ethical |
|
when should you pick a cohort study?
vs. when should you pick a case-control study? |
cohort: when there is a rare exposure and not so much is known about the exposure
2. when you want to evaluate the effects of exposure when should you pick case-control? rare outcome little known about outcome want to evaluate several exposures 3. long induction, latent period 4. when the data is expensive |
|
when should you pick prospective cohort?
when should you pick retrospective? |
prospective: outcome is short latent period
2. current exposure 3. you want high quality data retro: 1. long induction 2. historical exposure 3. limited time/resources |
|
MEASURES OF DISEASE
|
HOW DO YOU QUANTIFY DISEASE OCCURANCE?
|
|
what are the factors that quantify occurrence of a disease in a population?
|
1. number of people
2. size of population (This is important!! because if you don't know the size of the population the data is pretty meaningless) 3. length of time that people were followed |
|
what are some measures of morbidity? (disease burden)
|
1. incidence
2. prevalence |
|
what are the two types of incidence?
|
1. cumulative incidence
2. incidence rate |
|
how do you measure prevalence?
|
point and period
|
|
what are some measures of mortality?
|
1. mortality rates
2. case-fatality 3. proportionate mortality 4. years of life lost. |
|
Measures of morbidity:
|
1. number of cases of the disease
|
|
incidence
|
occurrence of a NEW DISEASE that develops during a specific time
|
|
prevalence
|
measures frequency of EXISTING disease during a specific time
|
|
absolute numbers
|
reporting the number of cases (people w. outcome)
this is ...limited? because you don't know the population SIZE but what you can do is compare one year to the next |
|
what are the benefits of absolute numbers?
|
1. its easy to do
2. its easy to understand |
|
what are the drawbacks of absolute numbers?
|
1. there is no sense of scale
2. its hard to use this information because you don't know if the results are larger than expected or lower than expected (aka, is this good or bad?) |
|
when should you use absolute numbers?
|
1. when populations are similar
2. when planning medical resources |
|
FORMULA for absolute numbers
|
# of cases in a TIME PERIOD
------------------------------------ population size |
|
how do you measure incidence?
|
people who don't have the disease but are AT RISK FOR DEVELOPING THE DISEASE
|
|
Why is incidence good?
|
1. because you can measure risk - the probability of an event happening
2. you can also measure cumulative incidence (proportion) or incidence rate (incidence density) |
|
what is cumulative incidence?
|
proportion of people who are disease free who become diseased by the end of a specific time period
|
|
what is the FORMULA FOR CUMULATIVE INCIDENCE?
|
CI= # new cases during time period
----------------------------------- total population AT RISK! |
|
what are some aspects of cumulative incidence?
|
1. its a proportion
2. estimate the risk of a certain population 3. everyone is followed for a certain time period (it is a closed population |
|
what is the INCIDENCE RATE?
|
the rate at which new cases occur in a population that is at risk for the disease
|
|
What is the FORMULA for in INCIDENCE RATE
|
# of new cases in a specific time
-------------------------------------- total PERSON TIME of observation |
|
associations
|
to measure association you have to rule out exposure and disease
|
|
what do you need in order to be a confounding variable?
|
1. independent cause or predictor of disease
2. associated w. exposure 3. you need to be an intermediate |
|
what is a confounding variable?
|
it a variable that explains that exposure and disease are linked through a 3rd factor
|
|
what is bias?
|
systematic error
|
|
what are two types of bias?
|
selection bias and observation bias
|
|
what is selection bias?
|
it is systematic error in the way participants are brought in to the study
|
|
what is observational bias?
|
systematic error in the way information is obtained once already in the study
|
|
what is chance?
|
an explanation for the data
|
|
what can you conclude if p < .05?
|
that it is unlikely that the results are due to chance
|
|
what about if the confidence interval includes 1?
|
that the results are not significant and you can explain results by saying that they were due to chance
|
|
how do you move from association to causation?
|
you use totality of evidence
|
|
what are some criteria that helps you know that something is causal?
|
1. temporality
2. strength of association 3. dose-response 4. replication 5. biology 6. consideration of other things 7. cessation of exposure 8. consistency |
|
temporality
|
for an exposure to cause an outcome it has to occur before the outcome occurs
|
|
strength of relationship
|
the stronger the Relative Risk, the likelier the relationship is to be causal
|
|
dose response
|
if it is causal, the more you do it the higher the risk
|
|
replication
|
it should be reported in more than one study
|
|
biology
|
if there is a biological reason it is more likely to be causal
|
|
alternative explanations
|
can anything else explain the reason for this?
|
|
cessation
|
if you stop doing it, the outcome will decrease
|
|
consistency w. other knowledge
|
what do animal studies show us?
|
|
how do you find out if something is causal (what is the sequence of events?)
|
1. clinical observations (hypothesis)
2. available data (not ideal) 3. case control 4. cohort 5 RCT |