case-control study Flashcards
How is the study designed?
- Population from which cases and controls are identified
- Investigators identify cases of disease and select controls who represent a sample of the source population that produces the cases
- Time
- Compare exposure in cases and controls.
What study can be used to solve cohort study limitations?
case-control study
What is a case-control study
- identifies individuals with a disease (cases) and compares them to individuals without the disease (controls)
- Past exposures are examined in both groups to assess associations with the disease
- The study begins by knowing the disease status of participants.
When is it desirable to conduct a case-control?
- When exposure data are expensive or difficult to obtain
- When diseases have long latent periods, results may take decades to emerge
- When the disease is rare, a cohort study would require a sample size that is too large.
- When the population is challenging to follow, a high loss of follow-up may result in biased results.
- When little is known about the disease, it can evaluate many exposures (and because investigators evaluate more than one hypothesis)
what should you consider when defining a source population?
Consider hypothesis, person, place, and time when defining the source population.
What is essential for identifying cases in an epidemiological study?What should the criteria for case definition ensure?
- A clear case definition.
- it should lead to accurate classification of the disease
What type of sources should be used? From where can cases be identified in a study?
- efficient and accurate
- Cases presenting to a hospital or clinic
- Population registries(state cancer registries, notifiable infectious diseases registries)
What is a key similarity between cases in a case-control study and a cohort study?
Cases are the same as those that would be included in a cohort study
Why is it important to enroll as many cases as possible? Do case-control studies need to include all cases of disease within a population?
- Cases are statistically precious- enroll as many of them as possible
- No, they do not need to include all cases of disease occurring within a defined population.
What is the benefit of having restrictive case definitions? How does a restrictive case definition affect sample size?
-restrictive cases reduce the number eligible for inclusion, leading to fewer classification errors
- it results in a smaller sample size.
What are controls? other names? The primary purpose of control?
- A sample of the source population that produced cases (referent group or study base)
-To estimate the exposure distribution in the source population that produced the cases.
Why is it important to know the exposure prevalence among cases and controls?
: It allows researchers to measure the association between exposure and outcome.
What is the “would criterion” in control selection?
If a control were to develop the disease under study, they should be eligible to become a case in the study.
What are the two necessary requirements for control selection in a case-control study?
- controls must come from the same source population as the cases and random selection is necessary for representativeness
- controls must be selected independently from the exposure - meaning that their exposure status does not influence their selection
How should controls be selected regarding their exposure status
Controls must be selected independently of their exposure to avoid bias.
Where are the three places to find control?
nested control
population-based control
hospital or clinic-based controls
What are nested controls? Where do they form from?
- These are controls selected from an existing cohort population. Controls represent a sub-set of the full source population
- form a cohort population/study
What are the benefits and disadvantages of nested controls?
- Good news: controls come from a clearly defined source population; already enrolled -> willing participants
- Bad news: restricted to members of existing cohort; may limit hypothesis that can be studied
What is population-based controls? Source?
- These are controls selected from the general population, most suitable when cases are from well-defined geographic area
- random digit dialing, cell phone or internet subscribers, residence lists, tax lists
what are the benefits and disadvantages of population-based controls?
- The good news is that controls often come from well-defined source populations
= Controls are selected in a way where you know they come from teh same base population as cases do - Bad news: very time-consuming, hard to inspire participation, many may not recall past exposure as well as cases
What are hospital or clinic-based controls?
what kind of people are chosen?
- Controls selected from among patients at a hospital or clinic
- Choose control patients with diseases (more often than one) other than the cases of disease
- Typically used when cases are identified from hospitals
what is the benfits and disavatnge of using hospital based controls?
- good news: easy to identify and access - less time and money, the accuracy of exposure recall comparable to cases, more willing participants
- Bad news: these controls are not randomly selected. This means that hospital-based controls must be carefully selected to accurately represent the exposure history in the source population
Which illness makes good hospital controls?
- llness that has the same catchment areas as the cases- This is what we mean when we say controls should come from the same source population as cases
- Illness that has no relation to risk factors under study - This is what we mean when we say the selection of controls should be independent of exposure
What is the purpose of sampling controls?
- This is to determine the exposure distribution in the source population that produced the cases
- Power calculation helps determine the number of controls needed because we want to be able to detect the difference in exposure levels between cases and controls
What can you not calculate? why?
- measures of frequency: risks and rates
- No longer have the entire denominator of the population at risk (because controls represent a SAMPLE of the total population)
- Can only calculate using the ODDS RATIO ( relative measure of association )
what can you not use when measuring association?
- the margin totals on the right side but can use the bottom totals
- No absolute measures of association can be calculated (no marginal totals)
odd ratio formula? what studies do you use it in?
ad/bc
- You would calculate the exposure odds
- In cohort studies, you calculate the disease odds ratio
How do you interpret the odds ratio?
The odds of developing (health outcomes of interest) among the exposed group is 0.X or XX.XX% times the odds of developing the health outcome of interest among the unexposed group during the time
What is the highest ratio for the number of controls to a case?
- When the number of cases is limited, and controls are a lot you can increase the power of the study to detect an association by increasing the size of the control group
- Up to 4:1 ratios going more than are expensive unless it is premade data
Relative measures of association for Odds ration = ?
No association between exposure and disease the odds ration =1
If exposure is assoiated with increase ODDS of disease >1
If the exposure is assoiated with decreased odds of diseas <1
what formual can odds ration = to?
APe - e we do not have the ability to use the marginal totals for a follow-up study
design.
- APt - we do not have the ability to use the marginal totals for a follow-up study design. Also, the RR (risk ratio/rate ratio) and PR (prevalence ratio) cannot be calculated in a case- control study; therefore, the OR must be used
what are the methods for samping controls (3)?
- Select from non-cases or survivors at the end of the case diagnosis - survivor sampling
- Case-base or case-cohort sampling - select from the population at risk at the beginning
- Risk set-sampling - controls are selected from the population at risk as cases are diagnosed
limitations of case control studies?
- often studying single outcome
- Inefficient for rare exposures
- More opportunity for systematic bias (recall and selection bias)
- May be unsure about temporal sequence between exposure and disease
- Cannot calculate absolute measure of association
strengtsh of case-control study?
- Fewer ethical concerns than experimental studies
- More effiecnet than cohort studies
- Requires less time and money
- Fewer subjects needed to measure an association, this is especially helpful when exposure informaion is expensive to obtain
- No need to wait for long latency diseases to devlop
- Easy to explore effect of many exposures on an outcome
- Useful for diseases with little information/unexpokired etiologies
- Important fr infections disease outbreaks
case-crossover study?
- a variant, used to study the effect of transient exposures on teh risk of acute events, cases serve as their own controls, and the exposure frequency during a hazard period is compaed to that from a control period
- Risk od motor vechile cokkiding while using phone
- Hazard period: the period of increased risk following the exposure - The exposure frequency during the hazard peirod is compared to a control period
- Cases serve as their own controls
