Case Control Flashcards
When is case study warranted?
-usually conducted before a cohort Orr an experimental study
–costs relatively less and can be done in less time
-Preferred if the outcome (disease) is rare
–ex. -adverse run; bc investigators can intentionally search for the cases
How to select cases
-criteria MUST be well-formulated and documented to ensure all included cases are based on the same diagnostic criteria
-Case sources: hospital or clinic pts, disease registries, claim data
-generalization
Selection of Controls
-should come from the same population population at risk of disease from which cases develop
–often selected to be similar to case on key factors but w/o the disease. could be other diseases, however, that may influence presence of certain exposures, etc.
-Control sources
–hospital, clinic, best friend/neighbor, population-at-large (random digit dialing)
Selection of Controls
-should come from the same population
Confounding in Case Control Studies
-confounding
-methods used to address –>matching, restricting eligibility criteria, regression analyses (Adjusted odds ratio (for Dif factors–take one variable at a time and adjust all others looking at specific exposures))
Matching
selecting controls so that they are similar to cases in potential confounding variables
Matching examples
-age, sex, race, etc.
-groupd matching: cases matched based on proportions (i.e. 25% of cases & controls are of a specific race-make sure they’re the same)
-indiv. matching: looking at each case is individually matched to a control
Problems w/ Matching
-practical
-conceptual
-Match is based on characteristics that you ALREADY know are risk factors for the disease ( thus not interested in investigating)
Recall Bias
-Occurs when cases may be morel likely to recall certain exposures than controls
-Measurement of exposure in a case-control study is established AFTER disease develops (Ex. Zika virus–babies with little beads)
KNOW RE RISK FACTORS FOR THE DISE
Other biases in Case Controls
-selection bias
-Misclassification bias
selections bias
Exposure (or factors related to the exposure) affect the selection of cases and controls differently
–exposure more freq. in cases or controls based on how they were selected
Misclassification bias
Depends on the sensitivity and specificity of the testing to determine cases and controls
Advantages of C-C
-inexpensive, efficient
-can be conducted quickly and is often used as a “first step (don’t have to wait for disease to occur)
-ideal rare diseases or those with ling latency periods
Disadvantages of C-C
-bias, especially recall bias
-confounding
-ffinding suitable controls can be challenging
-only test for associate
-no info on incidence prevalence
-must consider time between exposure and outcome
