Speaking Notes

PADM 5324

October 13, 2009

Dr. Neubauer

 

WHERE WE ARE:

 

 

THESE QUESTIONS REFLECT CONTENT IN CHAPTERS 9 AND 10 OF OUR TEXTBOOK.  Questions like these are likely to be on your final examination which I anticipate giving in the usual way.

 

1)         In epidemiology, Is there a distinction between a CAUSE and a RISK FACTOR?  Given that an exposure to A does not always lead to Y, which word seems more appropriate?

 

 

2)         In a COHORT study you need to find a control group to be compared to the

 

            A         group of those exposed to a possible cause

            B          group of those who have the disease/condition

 

3)         In a CASE-CONTROL STUDY you need to find a control group to be compared to the

 

            A         group of those exposed to a possible cause

            B          group of those who have the disease/condition

 

4)         In a sense you work backward in a CASE-CONTROL STUDY to determine which people were previously EXPOSED to something that may have caused the disease/condition.

 

            A         true

            B          false

 

5)         In a COHORT STUDY you first observe who was exposed to a possible cause and then you later observe who gets the disease/condition.

 

            A         true

            B          false

 

6)         A COHORT STUDY can be completed quickly because you don't have to ask people to remember things.

 

            A         true

            B          false

 

7)         In a COHORT STUDY if you first select a pre-exposure study population and then you wait to observe exposures and then you wait to observe instances of the disease/condition . . .

 

            A         it may take many years to have any findings to report

            B          people may drop out of the study

            C         the study may be affected by unanticipated events

            D         all of the above

 

8)         A CASE-CONTROL STUDY requires asking people to remember what things they have been exposed to in the past.

 

            A         true

            B          false

 

9)         Say your hypothesis is that A causes Y and when you do not find support for X being a cause of Y you decide to test to see if B causes Y.  In a "PURE" COHORT STUDY you would select another cohort, observe for exposure to B and then wait  to observe for incidence of Y.

 

            A         true

            B          false

 

10)       If you stated with a COHORT STUDY and your "A causes Y" hypothesis did not work out, how could you modify the study to test your new "B causes Y" hypothesis?  (Let's not get deeply into all the diagrams in the section beginning on page 191.  But basically I am asking you to think about a HYBRID DESIGN also known as a COHORT-BASED CASE-CONTROL STUDY.)

 

 

 

 

 

11)       What is a dichotomous exposure?

 

 

 

12)       Are most exposures dichotomous or not dichotomous?  (consider smoking and also consider exposure to second-hand smoke)

 

 

 

 

 

13)       One of the ethical problems in designing a study relates to how to assign participants to groups.  How does the "long form" of the COHORT STUDY methodology address the problem?

 

 

13)       When selecting a control group (using either method) the idea is for the "membership" of the two groups to be the same except in the one way they are different which is relevant to the hypothesis being tested.  Group matching is not likely to get all the proportions of cases "right."  What is individual matching?

 

 

 

 

 

 

14)       Why does it become more difficult to find participants when individual matching is based upon a relatively large number of variables?

 

 

 

 

15)       Do most people have complete and accurate memories of their exposures to things that may affect their health?

 

 

 

 

16)       When asking people about their previous exposures is there a potential for misunderstanding?

 

 

 

 

 

17)       When asking people about their previous exposures is there a potential for recall bias?

 

 

 

18)       According to our textbook is there substantial evidence that recall bias is a major problem in case control studies and has lead to erroneous conclusions?

 

 

 

19)       Is there any reason not to have multiple control groups?

 

 

 

 

20)       What is the example of the use of two control groups in Gold's study of cancerous brain tumors in children?

 

 

21)       Mesothelioma is a form of lung cancer almost always associated with exposure to asbestos. 

 

http://en.wikipedia.org/wiki/Mesothelioma

 

There is often a delay of many years between exposure and incidence.  Most people have had some exposure to asbestos.  For years the use of asbestos WAS REQUIRED in the construction of new schools.  It was used in the brakes of vehicles for years.  Some people, however, have had relatively high potential exposures. 

 

In order to study the possible relationship between asbestos exposures (from specific sources) and mesothelioma, would you propose a cohort study or a case-control study?  Why?

 

 

 

Problems

Benefits

Cohort study

 

 

 

 

 

 

 

 

Case study