Case Control Study Design Chapter 10 Flashcards
List the titles for columns and rows in a 2 x 2 breakdown and how to assess the portion of exposed for case-control studies
Columns from left to right disease cases, no disease controls
Rows: from top to bottom exposed (those who ate chili peppers), non-expose (those who did not eat chili peppers)
Portions of the population who were exposed calculations are done from top to bottom in a column fashion
Example a/A+ c
List the order of events for case-control studies
Select two groups of individuals: those with the disease (cases) and those that do not have the disease (controls)
And ascertain exposure
Determine the portion of exposed among the cases and the portion of non-exposed him on the cases. Determine a portion of exposed him on the controls and a portion of non-exposed among the controls
The portion of cases who were exposed is compared with the portion of Controls who were exposed
Define measures of association
The chance of something happening that can be expressed as a risk or the odds
Define risk
Risk equals the chance of something happening divided by the chances of all things happening
Define odds
The chance of something happening divided by the chances of it not happening
If the probability of something happening = 1/34 = 0.29
What are the odds
Formula = probability divided it by (1 -probability)
So 0.29/0.71= 0.41
Give the formula for calculating odds ratio in a case control study
The odds of exposure among the cases/the odds of exposure among the controls
Or easy way:
The cross product ratio
A times D/B times C
How are the odds ratios expressed when explaining the results
PAY ATTENTION TO THIS THIS WILL BE ON THE EXAM
The odds of being EXPOSED to chili peppers are 5.92 times greater for gastric cancer cases as compared to controls
Interpretation of odds ratio (OR) as a measure of association in case-control studies
Exactly the same as cohort
OR = 1
Odds of exposure in the case is equal to the odds of exposure in the controls (no Association)
OR>1
Odds of exposure in the cases is greater than the odds of exposure in the controls (positive association)
OR<1
Odds of exposure in the cases is less then the odds of exposure and controls (negative association) protective effect
Explain the relationship between OR and RR
KNOW THIS
OR ~RR: when the disease is RARE
CAUTION: OR > RR WHEN THE OUTCOME IS NOT rare
This can be explained because
O= p/1-p when P (risk of diseases small), then O=p/1
Explain the process and reasoning of matching
KNOW THIS
Matching cases and controls-The process of selecting the control so that they are similar to the cases
This is done because we want the cases that controls to come from a similar population of except for the exposure were studying
Cases of controls me different certain characteristics (other than exposure) which may influence the outcome. Can control for this in the analysis or by matching
List the problems with matching
If you match on too many characteristics it may be difficult to identify an appropriate control
If you match on a given characteristic then you can no longer study that characteristic
Never match on your EXPOSURE variable
When analyzing a matched case-control study what parts of the 2 x 2 are used and why
KNOW
B/C
We do not compare a and D because they are matched and they won’t teach you anything
Expressed in sentence form as:
Uterine cases were three times more likely to be exposed to estrogen then MATCHED controls
What are the necessary elements for case selection in a match case control study
Establish a case definition and stick with it (strict diagnostic criteria)
Cases can be new cases (incidents) or existing (prevalence) cases
Cases must be representative of persons with the disease in a population
Identify the sources of your cases:
Hospitals
Public health clinics
Physicians offices
Disease registries
Explain the necessary steps for control selection in matched case controlled studies
Selection of an appropriate comparison group is the most difficult and critical issue in the design of case control study
Controls our subjects FREE of the disease (or outcomes of interest)
Control should come from the same source population is the cases
The prevalence of exposure of the control should be similar to the prevalence of exposure among the source population
Source of control:
General population
Random digit dialing
Friends relatives
Hospital or clinic-based
Advantages and disadvantages of using general population as controls
Advantages
Represent the population from which the cases were selected
Disadvantages
Costly and time-consuming
How to identify people in the population
Hard to contact people
Poor recall than hospital controls
Less motivated to participate then other controls
Advantages and disadvantages of random digit dialing for controls
Advantages
May approximate random sample from the source population. Controls are often match to cases on area code and prefix
Disadvantage
Possibility of contacting each eligible subject may differ due to time of day, number and household, color ID
Widespread cellular-phone used to reduce his value
Advantages and disadvantages of friend and relative controls
Advantages
Healthier than hospital controls
Tend to be more could a whopper lived in general population controls
Often similar cases on factors such as SES, lifestyle, and ethnic background
Disadvantages the list of potential friends/relatives control is often derived from the case, this dependence May add a potential source of bias
Friend relative controls may be TOO similar to cases regarding the exposure of interest
Advantages and disadvantages of hospital controls
Advantages
Easily identified, sufficient number, low-cost
More likely to cooperate, minimize nonresponse bias
Hospital selection factors that influence decision to come to a particular hospital maybe similar to cases
May be more aware of past exposures at events
Disadvantages
Differ from healthy people, may not represent the exposure distribution of the population where the cases were obtained
Controls should be limited to diagnoses for which there are no prior indications of a relationship with exposure
How many controls per case
The optimal case-control ratio is 1 to 1
When the number of cases is small the sampling size for the study can be increased by using more than one control case
List the criteria for accessing exposure for those that occurred before the onset of disease
Study subjects-self reporting questionnaires
Particularly vulnerable to recall bias as cases may recall their exposure history more thoroughly than controls
Existing records (completed before the occurrence of outcome events)
Work history-industrial hygiene
Hospital-birth, diagnostic
Prescriptions
Interviews with Sarah get (spouse a sibling)
Advantages to case-control studies
Relatively quick and inexpensive
Small sample sizes
Optimal for evaluation of RARE DISEASES
can examine multiple etiologic factors for a single disease
Disadvantages to case-control studies
Inefficient for evaluation of rare exposures unless the disease is common among the expose
Disease status can influence the selection of the case and control subjects
Maybe difficult to establish the temporal relationship between exposure and disease. Exposure measurements are collected after disease occurrence
Prone bias compared to other analytical study designs, in particular selection and recall bias
Explain nested case-control studies
Occur after the establishment of a cohort study
Identify cases as they develop
Choose controls as the case is selected or at random
Advantages of nested case-control studies
Decrease in recall bias-dad and I are obtained before the disease develops
Identify risk factors, rather than early or subclinical disease as the specimens were obtained years before disease occurs
Less expensive to conduct (only if someone has already initiated the cohort)
Increased comparability between cases and controls