Case-Control Study Design

In a 2x2 table, how do you calculate the odds ratio?

(a/c)/(b/d)= ad/bc

What are common sources of cases for case-control studies?

-Hospitals -Community/Population -Registries -Large pre-paid insurance plans -Retirement communities

What are the types of matching?

-Individual -Frequency (group) matching

What are the advantages of a case-control design?

-May be quicker and less expensive than cohort studies -Well suited for rare disease and diseases with long latent periods -Fewer subjects required than for cohort studies -Multiple etiologic factors can be evaluated

What are the disadvantages of a case-control design?

-More potential sources of bias and error than cohort studies -Cannot determine temporal relationship -Appropriate control group may be difficult to identify

What is the measure of association in case-control studies?

-Odds Ratio (OR)

How do you assess exposure in a case-control study?

-Timing of exposure- when did it occur in relation to disease onset or index date -Quantification of exposure- for dose response analysis; includes amount, duration, and frequency of exposure

Describe a case-control study

A case-control study begins with people who have the disease (cases) and compares them to people who don't have the disease (controls)

How do you calculate an odds ratio?

Odds the cases were exposed/ Odds the controls were exposed

What is the push and pull with selecting cases and controls?

Representativeness vs. Comparability. Is the goal to select cases and controls that are representative of those with and without the disease, or to select cases and controls so they are like each other in all ways relevant?

What is the purpose of a case-control study?

To compare the odds of past exposure to a suspected risk factor between cases and controls.

True or False: Controls should be free of the disease under study

True

True or False: Odds ratios can be a good approximation of the risk ratio.

True, if your study is properly designed.

How are controls selected in a case-control study?

Usually select controls in the same manner as cases. Any exclusion/eligibility criteria applied to the cases should also be applied to the controls.

How are cases selected in a case-control study?

Without reference to or knowledge of their exposure history. They are selected based on OPPORTUNITY for exposure, but not just exposure.

In a 2x2 table, how do you calculate the prevalence of past exposure among cases?

a/(a+c)

In a 2x2 table, how do you calculate the odds of past exposure among cases?

a/c

In a 2x2 table, how do you calculate the prevalence of past exposure among controls?

b/(b+d)

In a 2x2 table, how do you calculate the odds of past exposure among controls?

b/d

What are the pros and cons of selecting representative cases and controls?

Advantages: Increase generalizability, reduces potential for certain types of selection bias Cons: May be more difficult, time-consuming, resource-intensive

What is the purpose of the control group in case-control studies?

Controls provide a comparison group for cases. They are intended to represent the frequency of exposure in the population from which the cases arose.

Does the odds ratio in a matched case-control analysis look at exposure or disease?

Exposure

True or False: Controls should be selected from a different source population then the cases.

False

True or False: You can calculate risk ratios from case-control data.

False. Case-controls do not use incidence, so you cannot calculate a risk ratio.

True or False: You can use case-control data to calculate incidence.

False. In most cases.

How do you calculate the odds ratio in a matched case-control analysis?

It is the ratio of discordant pairs (case+control-/case-control+) B/C

Interpret: OR<1

Negative or inverse association between exposure and outcome

Interpret: OR=1

No association between exposure and outcome

Is the sampling for a case-control study based on outcome/disease status or exposure status?

Outcome/disease status

Interpret: OR>1

Positive association between exposure and outcome

Does a case-control study look at incident or prevalence cases? Why?

Prefer to study incident cases. -Reduce potential for survival bias -Diagnoses most likely to be uniform (using same criteria) -Recall of exposure may be better

What are the pros and cons of using hospital-based controls?

Pros: Captive population, clearly identified, economical method, less recall bias Cons: Potential for selection bias, less generalizable

What are the pros and cons of selecting comparable cases and controls?

Pros: Easier to detect smaller differences; reduces chances for unmeasured confounding Cons; May have select group of cases, potential for Berkson's bias (hospital-based studies only)

What are the pros and cons of using multiple types of controls?

Pros: Internal replicate of study; access certain types of bias Cons: Costly and time-consuming; if results differ, must be able to explain *Not common

What are the pros and cons of using neighborhood controls?

Pros: Proves controls of similar socioeconomic status and environment as cases Cons: Overmatching possible, low response rates

What are the pros and cons of using a community sample (or population sample) of cases?

Pros: Representative case group Cons: Costly, time-consuming, difficult without registry

What are the pros and cons of using a hospital-based source of cases?

Pros: easier, cheaper Cons: Potential for bias, referral patterns, only suitable for diseases that are usually hospitalized

What are the pros and cons of using community/general population samples for controls?

Pros: highly representative, calculate population frequency of exposure Cons: costly, problems with refusal, phone coverage, and increase in cell phone usage (for random-digit dialing)

What is individual (pair) matching?

Selecting one or more controls for each individual case

What is frequency (group) matching?

The control group is selected so that its distribution is similar to that of cases for potentially confounding variables such as age, gender and race. Requires that all of the cases be selected first or that the distribution is unknown.

How do you interpret an odds ratio?

The odds of [EXPOSURE] among [CASES] are [ODDS RATIO] to having [EXPOSURE] among [CONTROLS]

How do you interpret a matched case-control odds ratio?

The odds of [OUTCOME] are [ODDS RATIO] in persons with [EXPOSURE] than in persons [NON-EXPOSURE]

What is an "odds"? How do you calculate an odds?

The probability of an event occurring (P) divided by the probability of the event NOT occurring (1-P) Odds= P/(1-P)

What does it mean to match cases and controls?

The process of selecting controls so that they are similar to cases for characteristics, such as age, gender, race, or socioeconomic status, that might be confounding variables. It is a way to deal with confounding, by making cases and controls similar across these characteristics.