carnegie institution for science 1530 p street, nw washington, dc 20005 april 18-21, 2011 teach...

Carnegie Institution for Science1530 P Street, NW

Washington, DC 20005April 18-21, 2011

Teach EpidemiologyProfessional Development Workshop

Revised

Teach Epidemiology

A picture of the occurrence, the distribution, and the types of the diseases of mankind, in distinct epochs of time and at various

points on the earth's surface, and ... render an account of the relations of those diseases to the external condition.

(Hirsch, 1883)

What is epidemiology?

The science of the mass phenomena of infectious diseases, or as the natural history of infectious diseases ... an inductive science,

concerned not merely with describing the distribution of the disease, but fitting it into a consistent philosophy.

(Frost, 1927)

Epidemiology may be defined as the study of the distribution of a disease or condition in a population and of the factors that

influence this distribution.

(Lilienfeld, 1957)

Epidemiology is defined as the factors determining the frequency and distribution of disease in human populations. For many years the

word covered only and quite specifically, the study of the spread and decline of communicable disease in human populations and

the prophylaxis and control of those diseases ... the scope includes all disease, acute or chronic, physical or mental, communicable or non-

communicable.

(Epidemiology: A Guide to Teaching Methods, 1973)

Epidemiology is a method of reasoning about disease that deals with the biologic inferences derived from observations of disease

phenomena in population groups.

(Lilienfeld, 1977)

... the study of the distribution and determinants of health-related states or events in specified populations and the application of this study to

the control of health problems.

(Gordis, 2009)

Ecologic Study

Ecological Fallacy

An error in inference due to failure to distinguish between information

obtained from groups versus individuals. An association observed

between variables at a population level does not necessarily hold true

for individual members of these populations.

Outcome

Exposure a

Ecologic Study

Random Assignment

Healthy PeopleHealthy People

Random Assignment

Blinding

Trial Investigators

Participants

Assessors

External Data Monitoring Board

Ethical Issues

Time Check

9:45 AM

Revised

Teach Epidemiology

National Research Council , Learning and Understanding

Teach Epidemiology

Enduring Epidemiological Understandings

Knowledge that “… is connected and organized, and … ‘conditionalized’ to specify the context in which it is applicable.”

Association Found Between Coffee and Pancreatic Cancer

Associated

Teach Epidemiology

What do we mean when we say that there is an association between two things?

Associated

Tied Related

Linked

Things that are associatedare linked in some way that makes them

turn up together.

Associated

Teach Epidemiology

Things that are associated are linked in some way that makes them turn up

together.

Associated

Teach Epidemiology

Suicide Higher in Areas with Guns

Smoking Linked to Youth Eating Disorders

Snacks Key to Kids’ TV- Linked Obesity: China Study

Family Meals Are Good for Mental Health

Lack of High School Diploma Tied to US Death

Study Links

Spanking to

Aggression

Breakfast Each Day May Keep Colds Away

Study Concludes: Movies Influence

Youth Smoking

Study Links Iron

Deficiency to Math

Scores

Kids Who Watch R-Rated Movies More Likely to Drink, Smoke

Pollution Linked with Birth Defects in US Study

Depressed Teens More Likely to Smoke

Associated

Teach Epidemiology

Epidemiologic studies that are concerned with characterizing the amount and distribution of

health and disease within a population.

Descriptive Epidemiology

Teach Epidemiology

Epidemiologic studies that are concerned with determinants of disease and the reasons for relatively high or low

frequencies of disease in specific population subgroups.

Analytical Epidemiology

Teach Epidemiology

Hypothesis

Formulating

Descriptive Epidemiology

Testing

Analytical Epidemiology

An unproven idea, based on observation or reasoning, that can

be supported or refuted through investigation

An educated guess

Hypothesis

Teach Epidemiology

Hypothesis:

Buprenorphine will stop heroin addicts from using heroin.

Making Group Comparisons and Identifying Associations

Teach Epidemiology

PopulationTrial 1

Population

500 Heroin Addicts

Sample 100

Heroin Addicts

10 Weeks

Trial 1

Population

500 Heroin Addicts

Sample 100

Heroin Addicts

10 Weeks

21 Heroin Addicts Tested Negative for Heroin

Trial 1

Tested Positive for Heroin Total

10021 79

Tested Negative for Heroin

Trial 1

Teach Epidemiology

When you can measure what you are speaking about, and express it in numbers,

you know something about it.

Lord Kelvin

But when you cannot measure it, when you cannot express it in

numbers, your knowledge is of a meager and unsatisfactory kind.

Teach Epidemiology

A measure of how often an outcome occurs in a

defined population in a defined period of time. It consists of a

numerator and a denominator.

The numerator is the number of people in the population or

sample who experienced the outcome and the denominator is the total number of people in the

population or sample.

Population / Sample

Outcome

Denominator

Numerator

Teach Epidemiology

… the risk of a negative heroin test was 21 / 100 in a 10-week period

21 tested negative for

heroin 100 study subjects

Numerator

Denominator

Teach Epidemiology

A measure of how often an outcome occurs in a defined group of people

in a defined period of time.

The likelihood of an outcome occurring.

Risk / Rate

Teach Epidemiology

Trial 1

Tested Positive for Heroin

10021 79

100or 21 %

Calculating Risk

Risk of

Negative Heroin

Teach Epidemiology

Process of predicting from what is observed in a sample to what is true for the entire population.

Inference

Teach Epidemiology

Trial 1

What does this tell you about the hypothesis?

InferenceProbe

10021 79

100or 21 %

Risk of

Negative Heroin

Teach Epidemiology

People who participate in a trial, but do not get the treatment.

People whose results are compared to the group that was treated.

Control Group

Teach Epidemiology

100or 21 %1007921

Control Group

Extend and label the table to include a control group.

Risk of

Negative Heroin

Teach Epidemiology

100or ? %No Bupe

Control Group

Making Group Comparisons

100or 21 %1007921

Risk of

Negative Heroin

Teach Epidemiology

100or ? %No Bupe

Making Group Comparisons

100or 21 %1007921

Exposure

Outcome / Disease

Risk of

Negative Heroin

Teach Epidemiology

100or 21 %

1007921Bupe

100or ? %No Bupe

Risk of

Negative Heroin

Teach Epidemiology

A cross-classification of data where categories of one variable

are presented in rows and categories of another variable

are presented in columns

The simplest contingency table is the 2x2 table.

Contingency Table

Teach Epidemiology

Population

500 Heroin Addicts

Sample 100

Heroin Addicts

10 Weeks

21 Heroin Addicts Tested Negative for Heroin

Trial 1

Trial 2

100? %

Tested Positive for

Heroin

No Bupe 100

100? %

Risk of

Negative Heroin

Teach Epidemiology

Assigned

Volunteer Heroin Addicts

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

10021%21 79 100 or

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

10021%21 79 100 or

Inference: Process of predicting from what is observed in a sample

to what is occurring in the entire population

Teach Epidemiology

When you can measure what you are speaking about, and express it in numbers, you know something about it.

Lord Kelvin

But when you cannot measure it, when you cannot express it in

numbers, your knowledge is of a meager and unsatisfactory kind.

Teach Epidemiology

The value obtained by dividing one quantity by another

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

10021%21 79 100 or

Ratio: The value obtained by dividing one quantity by another

Risk Ratio: The ratio of two risks

Risk Ratio

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

10021%21 79 100 or

Ratio: The value obtained by dividing one quantity by another

Risk Ratio: The ratio of two risks

Risk Ratio

Create a formula

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

10021%21 79 100 or

Risk Ratio

Relative Risk: The ratio of the risk of an outcome among the exposed to the risk of the outcome among the unexposed.

Relative Risk

Teach Epidemiology

10021%21 79 100 or

Trial 2

No Bupe

Risk of

Negative Heroin

10021%21 79 100 or

Risk RatioRelative Risk

The inference here is that there is no effect of Buprenorphine

Teach Epidemiology

Trial 3

100? %100 or

No Bupe

Risk of

Negative Heroin

100? %100 or

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

Trial 3

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

10062%62 38 100 or

Trial 3

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

10062%62 38 100 or

Trial 3

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

10062%62 38 100 or

Relative Risk

Trial 3

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

10062%62 38 100 or

Relative Risk

The heroin addicts who received Bupe were ___ times as likely to test negative for heroin as those who did not receive Bupe.

Trial 3

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

10062%62 38 100 or

Relative Risk

to what is occurring in the entire population.

Trial 3

Teach Epidemiology

Trial 4

100? %100 or

No Bupe

Risk of

Negative Heroin

100? %100 or

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

Trial 4

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

1006%6 94 100 or

Trial 4

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

1006%6 94 100 or

Relative Risk

Trial 4

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

1006%6 94 100 or

Relative Risk

The heroin addicts who received Bupe were ___ times as likely to test negative for heroin as those who did not receive Bupe.

Trial 4

Teach Epidemiology

10021%21 79 100 or

No Bupe

Risk of

Negative Heroin

1006%6 94 100 or

Relative Risk

to what is occurring in the entire population.

Trial 4

Teach Epidemiology

10021%21 79 100 orBupe

Trial 1

Risk of

Negative Heroin

What do the results tell us about the hypothesis that Buprenorphine will stop heroin addicts from using heroin?

Nothing

Teach Epidemiology

Trial 1 Trial 2

Trial 3 Trial 4

Teach Epidemiology

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

Teach Epidemiology

Risk of

Negative Heroin

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

No Bupe

TotalRelative

No Bupe

TotalRelative

Tested Negative

for Heroin

Tested Positive

for HeroinRelative

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

Teach Epidemiology

Risk of

Negative Heroin

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

No Bupe

TotalRelative

No Bupe

TotalRelative

Tested Negative

for Heroin

Tested Positive

for HeroinRelative

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10010021 79

90or 21%

Teach Epidemiology

Risk of

Negative Heroin

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

No Bupe

TotalRelative

No Bupe

TotalRelative

Tested Negative

for Heroin

Tested Positive

for HeroinRelative

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

Bupe is not associated with having a negative test for heroin.

Teach Epidemiology

Risk of

Negative Heroin

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

No Bupe

TotalRelative

No Bupe

TotalRelative

Tested Negative

for Heroin

Tested Positive

for HeroinRelative

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10062 3890

or 62%62

Bupe is associated with having a positive test for heroin!

Teach Epidemiology

Risk of

Negative Heroin

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

No Bupe

TotalRelative

No Bupe

TotalRelative

Tested Negative

for Heroin

Tested Positive

for HeroinRelative

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10062 3890

or 62%62

1006 9490

or 6%6

Bupe is associated with having a negative test for heroin.

Teach Epidemiology

Risk of

Negative Heroin

Nothing

Trial 1 Trial 2

Trial 3 Trial 4

No Bupe

TotalRelative

No Bupe

TotalRelative

Tested Negative

for Heroin

Tested Positive

for HeroinRelative

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

Risk of

Negative Heroin

Tested Negative

for Heroin

Tested Positive

for Heroin

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10010021 79

90or 21%

10021 7990

or 21%21

10062 3890

or 62%62

1006 9490

or 6%6

Bupe is associated with having a negative test for heroin.

Nothing

Compared to what?

Teach Epidemiology

Buprenorphine

Buprenorphine & Naloxone

Placebo

Teach Epidemiology

Handout

Teach Epidemiology

Snacks Key to Kids’ TV- Linked Obesity: China Study

Study Links

Spanking to

Aggression

Breakfast Each Day May Keep Colds Away

Youth Smoking

Study Links Iron

Deficiency to Math

Scores

In the News

2 x 2 Table

People who are exposed

2 x 2 Table

Areas with Guns

No SuicideSuicide

Areas without Guns

2 x 2 Table

R-Rated Movies

TotalDrink & Smoke

No Drink & Smoke

No R-Rated

Movies

People who are exposed and have the outcome

2 x 2 Table

R-Rated Movies

TotalDrink & Smoke

No Drink & Smoke

No R-Rated

Movies

2 x 2 Table

Family Meals

TotalMental Health

No Mental Health

No Family

People who are not exposed and do not have the outcome

2 x 2 Table

Family Meals

TotalMental Health

No Mental Health

No Family

2 x 2 Table

Study Links Iron Deficiency to Math Scores

2 x 2 Table

Iron Deficiency

Poor Math

Scores

No Iron

Deficiency

Good Math

Scores Total

People who do not have the outcome and are not exposed

2 x 2 Table

Iron Deficiency

Poor Math

Scores

No Iron

Deficiency

Good Math

Scores Total

2 x 2 Table

Pollution

Birth Defects

No Pollution

No Birth

Defects Total

People who are not exposed

2 x 2 Table

Pollution

Birth Defects

No Pollution

No Birth

Defects Total

2 x 2 Table

People who do not have the outcome

2 x 2 Table

Depression

No Depression

No Smoke Total

2 x 2 Table

Eating Disorders

No Smoke

No Eating

Disorders Total

People who are exposed and do not have the outcome

2 x 2 Table

Eating Disorders

No Smoke

No Eating

Disorders Total

2 x 2 Table

Study Links Spanking to Aggression

People who have the outcome

2 x 2 Table

Study Links Spanking to Aggression

Spanking

Aggression

No Spanking

TotalNo

Aggression

2 x 2 Table

Snacks Key to Kids’ TV-Linked Obesity – China Study

2 x 2 Table

Snacks Key to Kids’ TV-Linked Obesity – China Study

Snacks

Obesity

No Snacks

No Obesity Total

People who are not exposed and have the outcome

Teach Epidemiology

Laboratory

Teach Epidemiology

Laboratory

Teach Epidemiology

Naturally occurring circumstances in which groups of people within a population have been exposed to different levels of the hypothesized cause of an

outcome.

Natural Experiment

Teach Epidemiology

An epidemiologic study of a natural experiment in which the investigator is not involved in the intervention other than to record, classify, count,

and statistically analyze results.

Observational Study

Teach Epidemiology

An epidemiologic experiment in which subjects are assigned into groups to receive or not receive

a hypothesized beneficial intervention.

Controlled Trial

Teach Epidemiology

Buprenorphine

Teach Epidemiology

Naturally occurring circumstances in which groups of people within a population have been exposed to different levels of the hypothesized cause of an

outcome.

Observational Study of a Natural Experiment

Epidemiologic studies of natural experiments in which the investigator is not involved in the

intervention other than to record, classify, count, and statistically analyze results.

Teach Epidemiology

Stephen Jay Gould (survivor of abdominal mesothelioma)

Absolutely nothing in the available arsenal of anti-emetics worked at all. I was miserable and came to dread the frequent treatments with an almost perverse intensity. I had heard that marijuana often worked well against nausea. I was reluctant to try it because I had never smoked any substance habitually (and didn’t even know how to inhale). Moreover, I had tried marijuana twice (in the 1960s) … and had hated it …. Marijuana worked like a charm …. The sheer bliss of not experiencing nausea - and not having to fear it for all the days intervening between treatments - was the greatest boost I received in all my year of treatment, and surely the most important effect upon my eventual cure.

Teach Epidemiology

A particular or detached incident or fact of an interesting nature; a biographical incident or

fragment; a single passage of private life.

Anecdote

Teach Epidemiology

Science

Transforming Anecdote to Science

Teach Epidemiology

Anecdote

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Random Assignment

Controlled Trial

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Case-Control Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Cross-Sectional Study

Teach Epidemiology

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Random Assignment

Controlled Trial

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Case-Control Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Teach Epidemiology

The goal of every epidemiological study is to harvest valid and precise information about the

relationship between an exposure and a disease in a population.

The various study designs merely represent different ways of harvesting this information.

Essentials in Epidemiology in Public HealthAnn Aschengrau and George R. Seage III

Teach Epidemiology

Ms. Wilson

After-School, Make-Up Homework Hall

Talking too much on a cell phone causes students not to do their homework.

Label the 2x2 Table

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

Place the data into the 2x2 Table

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

Which of the following statements can be made based on the above data:

A: 15 of 20 students, who did not do their homework, had cell phones. B: 15 of 20 students, who had cell phones, did not do their homework.

Breast Implants

No Breast

Implants

No Connective Tissue Disease

3 1,180 1,183

513 85,805 86,318

Cohort Study – The Nurses’ Health Study (1976)

TotalConnective

Tissue Disease

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

A ratio of the probability of occurrence of an event to that of its nonoccurrence.

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

A ratio of the probability of occurrence

of an event to that of its nonoccurrence.

15 to 5 or 3 to 1

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

Nothing

15 to 5 or 3 to 1

OddsCompared to what?

CompareDivideCount

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

Nothing

15 to 5 or 3 to 1

OddsCompared to what?

What did Ms. Wilson do?

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

403 to 1

10 to 30 or 1

Place the data into the 2x2

What mathematical computation would allow them to complete the statement:

The odds of having a cell phone were ____ times greater among students who had not done their homework compared to students who did do their homework.

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

403 to 1

10 to 30 or 1

The odds of having a cell phone were ____ times greater among students who had not done their homework compared to students who did do their homework.150

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

403 to 1

3 / 1 = 3 1 / 3 = .333 / .33 = 9

10 to 30 or 1

Odds Ratio

Ratio of odds in favor of exposure among cases to the odds in favor of exposure among controls.

Relative Odds

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

Odds Ratio

dDid Not

Have Cell Phone

Did Not Do Homework

Did Homework

Had Cell Phone

Odds Ratio

Arthur Herbst, et al. “Adenocarcinoma of the Vagina,” New England Journal of Medicine, 284:16, 1971, 878-881

Case-Control Study

Authentic Assessment

Teach Epidemiology

Epi – Grades 6-12

• Are realistic; simulate the way a person’s understanding is tested in the real world

•Require judgment and innovation to address an unstructured problem, rather than following a set routine

• Ask students to “do” the subject rather than simply recall what was taught

•Replicate the context in which a person would be tested at work, in the community, or at home

• Are messy and murky

• Require a repertoire of knowledge and skill to be used efficiently and effectively

• Allow opportunities for rehearsal, practice, consultation, feedback, and refinement

Epi Challenge

http://www.teachepidemiology.org/viewSO1.php

Epi Challenge

http://www.teachepidemiology.org/viewSO2.php

Think Like an Epidemiologist ChallengeNew Jersey Science Olympiad, March 15, 2011

Thank you for competing in the 3rd Think Like an Epidemiologist Challenge.

You worked with others, developed epidemiologic knowledge and skills, and used judgment and innovation to actually "do" epidemiology under

pressure.

We hope you enjoyed the challenge.

NameSchool

Teach Epidemiology

Robert Wood Johnson Foundation

Detectives in the

Classroom

Special thanks to the Epidemiology Section of the American Public Health Association for allowing us to distribute their Section pins to the student participants in the 2011 Think Like an Epidemiologist Challenge.

Time Check

10:45 AM

Revised

Teach Epidemiology

Time Check

11:00 AM

The Journey

Detectives in the Classroom - Investigation 2-6: The Journey

The Journey from Exposure to Disease

Analogy

Epi Talk

Study Design Epi Talk

Procedures and methods, established beforehand, that are followed by the investigator conducting the study.

Timing

When are the passengers identified as exposed or unexposed?

When are the passengers identified as sick or not sick?

Timing

When does the epidemiologist start to observe the journey?

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Label the Train Tracks

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design:

Controlled Trial

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Controlled

Flow Diagram

Healthy People

Random Assignment

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design:

Cohort Study

Just as in the controlled trial, the epidemiologist is also on the train during the entire journey. But there is an important difference. The epidemiologist is not telling passengers what to do. Rather, the epidemiologist is just observing them and counting. Passengers are not being told to have or not have an exposure, they are just living their normal lives. The epidemiologist, on the ride for the whole journey, just keeps observing everyone’s exposures and whether or not they develop the disease during the journey.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design: Cohort Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort

Flow Diagram

Healthy People

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort

Flow Diagram

Healthy People

Controlled

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort

Flow Diagram

Healthy People

Controlled

Random Assignment

Review

Observational StudiesEpi Talk

Epidemiologic studies of natural experiments in which the investigator is not involved in the intervention other than to record, classify, count, and statistically analyze results.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design: Case-Control Study

The epidemiologist is not on the journey. Rather, the epidemiologist is waiting at the train station at the end of the journey. As passengers get off the train, the epidemiologist selects sick passengers for the case group and selects passengers who are similar but not sick for the control group. The epidemiologist then asks each person in the case group and control group questions about their exposures during the train ride. The epidemiologist relies on passengers’ memories of their exposures that occurred during the train ride.

Case-Control Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design: Case-Control Study

Case-Control

Observational

Flow Diagram

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design: Cross-Sectional Study

The epidemiologist, who has not been on the journey, stops the train somewhere during the trip (kind of like a train robbery) and takes a “snapshot” of all the passengers by asking them whether or not they have the exposure and whether or not they have the disease. Then the epidemiologist leaves the train and goes home to analyze the data from that particular day. The journey continues without the epidemiologist.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Study Design: Cross-Sectional Study

Cross-Sectional

Observational

Flow Diagram

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Epi Talk

Controlled TrialEpi Talk

An epidemiologic experiment in which subjects are assigned into groups to receive or not receive a hypothesized beneficial intervention.

Epi Talk

Cohort StudyEpi Talk

An analytical epidemiological study design in which the investigator selects a group of exposed individuals and a group of unexposed individuals and follows both groups to compare the frequency with which the disease occurs in each group.

Epi Talk

Case-Control StudyEpi Talk

An analytical epidemiological study design in which the investigator selects a group of individuals with a disease (cases) and a group of similar individuals without the disease (controls) and compares the frequency with which an exposure occurred in the cases versus the controls.

Epi Talk

Cross-Sectional StudyEpi Talk

An analytical epidemiological study design in which the investigator selects a group of individuals and determines the presence or absence of a disease and the presence or absence of an exposure at the same time.

Epi Teams

What’s My Design?

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Random Assignment

Controlled Trial

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort Study

Epi Team

Challenge

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Case-Control Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Controlled Trial

Epidemiologist is involved during the entire time from exposure and disease.

Assign treatment and control groups.

Follow through time and compare risk of disease in treatment group with risk of disease in control group.

Give exposure to treatment group, but not control group.

What’s My Design?

Observational

Flow Diagram

Case-Control Study

Observational

What’s My Design?

Cohort, Case-Control, and Cross-Sectional Study

Investigation 2-6 has ended.

The Journey

Epi Team Challenge

Epi Team ChallengeDetectives in the Classroom - Investigation 2-7: Epi Team Challenge

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Random Assignment

Controlled Trial

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Healthy People

Cohort Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Case-Control Study

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

What’s My Design?

Detectives in the Classroom - Investigation 2-7: Epi Team Challenge

Epidemiologist is involved during the entire time from exposure to disease.

What’s My Design?

Practice Clue

What’s My Design?

Clue 1

Begin Epi Team Challenge

Observational

Flow Diagram

Clue 2

Case-Control Study

What’s My Design?

Observational

Flow Diagram

Clue 3

What’s My Design?

Cohort, Case-Control, and Cross-Sectional Studies

Follow through time and compare risk of disease in exposed group with risk of disease in the unexposed group.

Clue 4

Trial and Cohort Study

What’s My Design?

Clue 5

What’s My Design?

Clue 6

What’s My Design?

Select a group of people with disease and a similar group of people without disease.

Clue 7

Case-Control Study

What’s My Design?

Healthy People

Flow Diagram

Healthy People

Random Assignment

Non-Observational

Clue 8

What’s My Design?

Observational

Flow Diagram

DZClue 9

Cross Sectional Study

What’s My Design?

Compare percent of exposed people in the two groups.

Clue 10

Case-Control Study

What’s My Design?

Healthy People

Flow Diagram

Healthy People

DZRandom

Assignment

Clue 11

What’s My Design?

Flow Diagram

Clue 12

What’s My Design?

Flow Diagram

DZObservational

Clue 13

Cohort Study

What’s My Design?

Ask each person about both exposure and disease at that point in time.

What’s My Design?

Clue 14

Ask both groups about their exposures in the past.

What’s My Design?

Clue 15

Case-Control Study

Healthy People

Flow Diagram

Healthy People

Random Assignment

Clue 16

What’s My Design?

Disease risk in exposed group is compared to disease risk in unexposed group.

Clue 17

Trial, Cohort Study, and Cross Sectional Study

What’s My Design?

Flow Diagram

DZClue 18

What’s My Design?

Select a healthy study sample.

Clue 19

Trial (?) and Cohort Study

What’s My Design?

Observe who has and has not been exposed.

Clue 20

Cohort and Cross-Sectional Studies

What’s My Design?

Clue 21

What’s My Design?

Healthy People

Flow Diagram

Healthy People

EObservational

Clue 22

Cohort Study

What’s My Design?

Epidemiologist is involved after disease has occurred and relies on subjects’ memories

to gather information about exposure.

Clue 23

Case-Control Study

What’s My Design?

Select a study sample.

Clue 24

Trial, Cohort, Case-Control, and Cross Sectional Studies

What’s My Design?

Healthy People

Flow Diagram

Healthy People

DZObservational

Clue 25

Cohort Study

What’s My Design?

Epidemiologist gathers data only at that one point in time.

Clue 26

What’s My Design?

Flow Diagram

Clue 27

What’s My Design?

Follow through time and compare risk of disease in exposed group to risk of disease in unexposed group.

Clue 28

What’s My Design?

Clue 29

What’s My Design?

Flow Diagram

DZClue 30

What’s My Design?

Flow Diagram

Clue 31

Case-Control Study

What’s My Design?

Ask each person about both exposure and disease at that point in time.

Epidemiologist gathers data only at that one point in time.

Disease risk in exposed group is compared to disease risk in unexposed group.

Select a study sample.

Clue 32

What’s My Design?

Epidemiologist is involved after disease has occurred and relies on subjects’ memories to gather information

about exposure.

Select a group of people with disease and a similar group of people without disease.

Compare percent of exposed people in the two groups.

Ask both groups about their exposures in the past.

Case-Control Study

Clue 33

What’s My Design?

Select a healthy study sample.

Follow through time and compare risk of disease in exposed group to risk of disease in unexposed group.

Observe who has and has not been exposed.

Cohort Study

Clue 34

What’s My Design?

Clue 35

What’s My Design?

End Epi Team Challenge

What’s My Design?

Detectives in the Classroom - Investigation 2-8: Which Design Is Best?

Which Design

Is Best?

Which Design Is Best?

Case-Control Study

Cohort Study

Main WeaknessesMain StrengthsStudy Designs

Epi Team ChallengeEpi Team Challenge

Case-Control Study

Cohort Study

Cross-Sectional

CohortCase-

Control Controlled

Case-Control Study

Cohort Study

Fastest

Which study design is the fastest?

Case-Control Study

Cohort Study

Fastest

Time Consuming

Which study designs are the most time consuming?

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Most Scientifically Sound

Which study design is the most scientifically sound?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Can Study Rare Diseases

Which study design is best for studying rare diseases?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Possible Time-Order Confusion

Which study designs do not identify the time order of exposure and disease?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Confidence in Findings

Which study design gives the least confidence in findings?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Best Measure of Exposure

Which study design provides the best measure of exposure?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Most Accurate Observational Study

Which study design is the most accurate observational study?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Which study design is the least expensive?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Unethical for Harmful Exposures

Which study design would be unethical for harmful

exposures?

Time Consuming

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Good Measure of Exposure

Which study design provides a good measure of exposure?

Time ConsumingUnethical for Harmful exposures

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Most Expensive

Which study designs are the most expensive?

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Relatively Less Expensive and Relatively Fast

Which study design is relatively less expensive and relatively fast?

Most Expensive

Most ExpensiveGood Measure of Exposure

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Possible Error in Recalling Past Exposures

Which study design creates the possibility of error in recalling past exposures?

Most Expensive

Most ExpensiveGood Measure of Exposure

Epi Team Challenge

Case-Control Study

Cohort Study

Fastest

Time Consuming

Least Expensive

Possible Error in Recalling Exposures

Which Design Is Best?

Time Consuming

Unethical for Harmful exposuresMost Expensive

Most Expensive

It depends ….

It depends on ….

• Regulations

• Time urgency

• How much is known about the association

• Money

• Whether the exposure is believed to be beneficial

Flow Diagram 2x2 Table

Detectives in the Classroom – Investigation 2-9: Designs, Diagrams, and Tables

“fit”

Designs, Diagrams, and Tables

Healthy People

2x2 Table

Where do these people “fit” in the 2x2 table?

Flow Diagram

Controlled Trial

Random Assignment

Controlled Trial

Flow Diagram 2x2 TableFlow Diagram

Random Assignment

Controlled Trial

2x2 TableFlow Diagram

Random Assignment

Controlled Trial

Where are these people in the flow diagram?

Cohort Study

2x2 Table

Flow Diagram

Cohort Study

Healthy People

Cohort Study

Healthy People

Cohort Study

Healthy People

Where do these

people go in the

2x2 table?

2x2 Table Flow Diagram

Case-Control Study

Where do these

people go in the

2x2 table?

Flow Diagram2x2 Table

Case-Control Study

Flow Diagram

2x2 Table

Where do these

people go in the

2x2 table?

Case-Control Study

Flow Diagram

2x2 Table

Where do these

people go in the

2x2 table?

Case-Control Study

E aFlow Diagram

2x2 Table

Where do these

people go in the

2x2 table?

Flow Diagram

2x2 Table

Where do these

people go in the

2x2 table?

Flow Diagram

2x2 Table

Where do these

people go in the

2x2 table?

Flow Diagram

2x2 Table

Where do these

people go in the

2x2 table?

Teach Epidemiology

The goal of every epidemiological study is to harvest valid and precise information about the

relationship between an exposure and a disease in a population.

The various study designs merely represent different ways of harvesting this information.

Essentials in Epidemiology in Public HealthAnn Aschengrau and George R. Seage III

Teach Epidemiology

Revised

Teach Epidemiology

Teach Epidemiology Workshop—Day 3Diane Marie M St. George, PhDUniversity of MD School of MedicineDept of Epidemiology and Public Health

EU7: One possible explanation for EU7: One possible explanation for finding an association is that the finding an association is that the exposure causes the outcome. exposure causes the outcome. Because studies are complicated by Because studies are complicated by factors not controlled by the observer, factors not controlled by the observer, other explanations also must be other explanations also must be considered, including confounding, considered, including confounding, chance, and bias.chance, and bias.

EU8: Judgments about whether an exposure EU8: Judgments about whether an exposure causes a disease are developed by causes a disease are developed by examining a body of epidemiologic examining a body of epidemiologic evidence, as well as evidence from other evidence, as well as evidence from other scientific disciplines.scientific disciplines.

EU9: While a given exposure may be EU9: While a given exposure may be necessary to cause an outcome, the necessary to cause an outcome, the presence of a single factor is seldom presence of a single factor is seldom sufficient. Most outcomes are caused sufficient. Most outcomes are caused by a combination of exposures that may by a combination of exposures that may include genetic make-up, behaviors, include genetic make-up, behaviors, social, economic, and cultural factors social, economic, and cultural factors and the environment. and the environment.

Reasons for associations

Confounding Bias Reverse causality Sampling error (chance) Causation

Confounding in our lives

MAP tests measure academic growth over time, independent of grade level or age.

Age- and gender-specific growth charts Age-adjusted rates of… Rates of lung cancer adjusted for smoking Computer exposure may cause illness

(absence), but it actually may just be the football game

Osteoporosis risk is higher among women who live alone than among women who live with others.

Confounding

Confounding is an alternate explanation for an observed association of interest.

Number of persons in the

homeOsteoporosis

Confounding

Confounding is an alternate explanation for an observed association of interest.

Exposure Outcome

Confounder

Controlling confounding

Study design phaseMatchingRestrictionRandom assignment

Study analysis phaseStratificationStatistical adjustment

Errors are mistakes that are:randomly distributednot expected to impact the MA less modifiable

Biases are mistakes that are:not randomly distributedmay impact the MAmore modifiable

Types of bias

Selection biasThe process for selecting/keeping subjects

causes mistakes Information bias

The process for collecting information from the subjects causes mistakes

Selection bias People who are working are likely to be

healthier than non-workers People who participate in a study may

be different from people who do not People who drop out of a study may be

different from those who stay in the study

Hospital controls may not represent the source population for the cases

Information bias

Misclassification, e.g. non-exposed as exposed or cases as controls

Cases are more likely than controls to recall past exposures

Interviewers probe cases more than controls (exposed more than unexposed)

Birth defects and diet

In a study of birth defects, mothers of children with and without infantile cataracts are asked about dietary habits during pregnancy.

Pesticides and cancer mortality

In a study of the relationship between home pesticide use and cancer mortality, controls are asked about pesticide use and family members are asked about their loved ones’ usage patterns.

Minimize bias

Can only be done in the planning and implementation phase

Standardized processes for data collection Masking Clear, comprehensive case definitions Incentives for participation/retention

Reverse causality

Suspected disease actually precedes suspected cause

Pre-clinical disease Exposure DiseaseFor example: Memory deficits Reading

cessation Alzheimer’s Cross-sectional study

For example: Sexual activity/Marijuana

Minimize effect of reverse causality

Done in the planning and implementation phase of a study

Pick study designs in which exposure is measured before disease onset

Assess disease status with as much accuracy as possible

Sampling error/chance

E and D are associated in a sample, but not in the population from which the sample was drawn.

RR in the populationRR in the population

D+D+ D-D-

E+E+ 5050 5050 100100

E-E- 5050 5050 100100

100100 100100 200200

RR = (24 / 718) / (281 / 8566) = 1.0

RR in sample 1RR in sample 1

D+D+ D-D-

E+E+ 2525 2525 5050

E-E- 2525 2525 5050

5050 5050 100100

RR = (24 / 718) / (281 / 8566) = 1.0

D+D+ D-D-

E+E+ 4545 55 5050

E-E- 1515 3535 5050

5050 5050 100100

RR = (24 / 718) / (281 / 8566) = 1.0

D+D+ D-D-

E+E+ 2020 3030 5050

E-E- 3030 2020 5050

5050 5050 100100

RR = (24 / 718) / (281 / 8566) = 1.0

Minimize sampling error (chance)

Random Selection

Time Check

12:30 PM

Revised

Teach Epidemiology

Time Check

1:30 PM

Revised

Teach Epidemiology

Time Check

2:00 PM

Revised

Teach Epidemiology

361Teach Epidemiology

Study Links

Spanking to

Aggression

Youth Smoking

Study Links Iron

Deficiency to Math

Scores

Ties, Links, Relationships, and Associations

Snacks Key to Kids’ TV- Linked Obesity: China

Depressed Teens More

Likely to Smoke

Study Links

Spanking to

Aggression

Youth Smoking

Study Links Iron

Deficiency to Math

Scores

Likely to Smoke

1. Cause

2. Confounding

3. Reverse Time Order

4. Chance

5. Bias

Possible Explanations for Finding an Association

Epidemiology

... the study of the distribution and determinants of health-related states or events in specified populations and the application of this study to the control of health problems.

Leon Gordis, Epidemiology, 3rd Edition, Elsevier Saunders, 2004.

1. Cause

2. Confounding

4. Chance

5. Bias

A factor that produces a change in another factor.

William A. Oleckno, Essential Epidemiology: Principles and Applications, Waveland Press, 2002.

Sample of 100

Sample of 100, 25 are Sick

Diagram

2x2 Table

a bc d

Types of Causal Relationships

a bc d

Diagram

2x2 Table

Types of Causal Relationships

Handout

X1X1X1

a bc d

Diagram

2X12 Table

Necessary and Sufficient

a bc d

X1 DZX2 X3+ +X1

X1X1X1

Diagram

2X12 Table

Necessary but Not Sufficient

a bc d

Diagram

2X12 Table

Not Necessary but Sufficient

a bc d

X1X1X1

DZX5 X6+ +

X2 X3+ +

X8 X9+ +

Not Necessary and Not Sufficient

Diagram

2X12 Table

a bc d

Diagram

2x2 Table

Necessary and Sufficient

a bc d

X DZX X+ +

Diagram

2x2 Table

Necessary but Not Sufficient

a bc d

Diagram

2x2 Table

Not Necessary but Sufficient

a bc d

DZX X+ +

X X+ +

Not Necessary and Not Sufficient

Diagram

2x2 Table

Heart Attack

NoHeart Attack

Lack of Fitness

No Lack of Fitness

Lack of fitness and physical activity causes heart attacks.

Lead Poisoning

NoLead

Poisoning

Lack of Supervision

No Lack of

Supervision

Lack of supervision of small children causes lead poisoning.

Is the association causal?

Study Links

Spanking to

Aggression

Youth Smoking

Study Links Iron

Deficiency to Math

Scores

1. Cause

2. Confounding

4. Chance

5. Bias

Likely to Smoke

1. Cause

2. Confounding

4. Chance

5. Bias

All the people in a particular group.

Population

A selection of people from a population.

Sample

Inference

Process of predicting from what is observed in a sample to what is not observed in a population.

To generalize back to the source population.

Sample

Population

Process of predicting from what is observed

to what is not observed.

Observed

Not Observed

Inference

Deck of

100 cards

Population

25 cards

Population

25 cards

25 cards25 cards25 cards

Even #

No Marijuana

Population

25 cards

= 2525

Even #

No Marijuana

Population

=M&M’s

No M&M’s

25 cards

Even #

No Marijuana

Population

25 cards

25 / 50 or 50%

Even #

No Marijuana

Population

25 cards

Total Risk Relative Risk

25 / 50 or 50 %

25 / 50 or 50 %50 % / 50% = = 1

____Odd #

Even #

No Marijuana

Population

25 cards

Population

To occur accidentally.

To occur without design.

Chance

A coincidence.

Chance

Population

Sample

20 cards25 cards25 cards25 cards25 cards

Sample

Population

Sample

5 5Odd #

Even #

No Marijuana

Sample

Population

Sample

5 / 10 or 50 %

5 / 10 or 50 %Odd #

Even #

No Marijuana

Sample

Population

Sample

5 / 10 or 50 %

5 / 10 or 50 %Odd #

Even #

No Marijuana

Sample

Relative Risk

50 % / 50% = = 150 %

Sample

20 cards

TotalRisk

5 / 10 = 50 %

Relative Risk

By Chance CDC

=Odd #

Even #

No Marijuana

Sample

5 / 10 or 50 %

Relative Risk

How many students picked a sample with 5 people in each cell?

= 150 %

Even #

No Marijuana

Chance

By Chance

Relative Risks

Greater than 1 Less than 1

Chance

Study Links Having an Odd Address to Marijuana Use

Relative Risks

Study Links Having an Even Address to Marijuana Use

Relative Risks

By ChanceBy Chance

25 cards25 cards25 cards25 cards

Chance

Sample

20 cards

TotalRisk

5 / 10 = 50 %

Relative Risk

=Odd #

Even #

No Marijuana

Different Sample Sizes

Relative Risks

By ChanceBy Chance

Chance

50 cards

Sample

20 cards

TotalRisk

5 / 10 = 50 %

Relative Risk

=Odd #

Even #

No Marijuana

Relative Risks

By ChanceBy Chance

Chance

75 cards

Sample

20 cards

TotalRisk

5 / 10 = 50 %

Relative Risk

=Odd #

Even #

No Marijuana

Relative Risks

By ChanceBy Chance

Chance

99 cards

Study Links

Spanking to

Aggression

Youth Smoking

Study Links Iron

Deficiency to Math

Scores

1. Cause

2. Confounding

4. Chance

5. Bias

Likely to Smoke

Association is not necessarily causation.

Revised

Teach Epidemiology

Time Check

3:30 PM

Teach Epidemiology

Explaining Associations and Judging Causation

1. Cause

2. Confounding

4. Chance

5. Bias

Teach Epidemiology

Coffee and Cancer of the Pancreas

Guilt or Innocence?Causal or Not Causal?

Does evidence from an aggregate of studies support a cause-effect relationship?

Teach Epidemiology

Sir Austin Bradford Hill “The Environment and Disease:

Association or Causation?” Proceedings of the Royal Society of Medicine

January 14, 1965

Teach Epidemiology

Handout

“In what circumstances can we pass from this observed association

to a verdict of causation?”

Teach Epidemiology

“Here then are nine different viewpoints from all of which we should study association

before we cry causation.”

Teach Epidemiology

Does evidence from an aggregate of studies support a cause-effect relationship?

1. What is the strength of the association between the risk factor and the disease?

2. Can a biological gradient be demonstrated?

3. Is the finding consistent? Has it been replicated by others in other places?

4. Have studies established that the risk factor precedes the disease?

5. Is the risk factor associated with one disease or many different diseases?

6. Is the new finding coherent with earlier knowledge about the risk factor and the m disease?

7. Are the implications of the observed findings biologically sensible?

8. Is there experimental evidence, in humans or animals, in which the disease has m been produced by controlled administration of the risk factor?

Teach Epidemiology

Handout

Teach Epidemiology

Timeline

Cohort Study

Randomized Controlled Trial

Timeline

Case-Control Study

Timeline

Random Assignment

Teach Epidemiology

Handout

Stress causes ulcers.

Helicobacter pylori causes ulcers.

Teach Epidemiology

Time Check

4:30 PM

carnegie institution for science 1530 p street, nw washington, dc 20005 april 18-21, 2011 teach...

human populations

acbdecologic study

distribution of adisease

influencethis distribution

specified populations

random assignment

external condition

relations ofthose diseases

Documents

ley 1530 de_2012

1530 mimosa flyer

man 20005 - lec 6

defter 1528-1530

proceso%20cas%20nro%20005 2013

pdf compressor · 2018. 12. 26. · 147390 27480 1952 40...

proceso%20cas%20nro%20005 1ro 2013

decreto 1530 de 1996

edição 1530

1530 oracle

102 1530-1600

suite washington, d.c. 20005-3960 perk1nsc01e

1530 xiangkun ren

prestigio nobile 1530 uputstvo za...

1530 track1 bocheva

berganza conde, periodismo especializado cap.20005

alaska 2012 bah rates - without dependents · alaska 2012...

installation and operating instructions quadratic integra...

maj 1530 ar

guia sa 8000-20005