Speaking Notes

PADM 5324

August 18, 20009

Dr. Neubauer

CHAPTER 1

Epidemiologyis the study of the distribution and determinants of health, disease, or injury in human populations and the application of this study to the control of health problems.

From Lecture 1 slide 6

http://ocw.jhsph.edu/courses/FundEpi/PDFs/Lecture1.pdf

What is epidemiology?

• Epidemiology is about how diseases and health conditions move through populations of people.
• It includes both morbidity and mortality.
• It is the science of understanding, predicting and preventing the things that cause people to become sick and/or to die prematurely.
• It has a close association with biostatistics.

Epidemiology is not the same as medicine, but is related.

• Both are concerned both with prevention and with cure.
• Both require and understanding of SYSTEMS dynamics.
• Both require an understanding of biology.
• Both involve observation and hypothesis testing.
• In medicine the unit of care is usually individual patients.
• In epidemiology the unit of analysis is populations of people (and environmental factors).

Epidemiology is the science behind public health.

Public Health is the practical application of epidemiology.

Medical science is the science behind medicine.

Biology is the science "under" both medical science and epidemiology.

Having said that, public health clinics are certainly involved in the delivery of services to individuals and do employ medical professionals.

Giving vaccinations (for example) can be valuable both in terms of public health and in terms of public (collective) health. In other words, the treatment can have value both in terms of protecting the individual and in terms of protecting others by preventing the individual from becoming a carrier of something.

The application of medical science is the practice of medicine.

The application of epidemiology and biostatistics is public health.

The usual degree of a person who practices medicine is MD or DO.

The usual degree of a person who works in public health is MPH.

Some people have both degrees.

Both medicine and public health involve the education of patients. But in one case the education is likely to be individual and in the other case the education is likely to be through the media or social networks.

Both medicine and public health are affected by PUBLIC POLICY.

Public policy should be informed by the insights that epidemiology can provide.

QUANTITATIVE methods of research -- measures captured as numbers.

QUALITATIVE methods of research -- often, research in which the "data" is text or speech.

"Etiology" refers to cause or origin.

"Progression" refers to how diseases spread through populations.

A vector is a means of transmission.

Statisticsis the science and art of dealing with variation of data in order to obtain reliable results and conclusion.

Biostatisticsis the application of statistics to problems in the biological sciences, health, and medicine

From Lecture 1 slide 7

http://ocw.jhsph.edu/courses/FundEpi/PDFs/Lecture1.pdf

Research done for Public Health purposes using the methods of Epidemiology can be either PRACTICAL RESEARCH PROJECTS or projects driven primarily by the simple desire to know and understand.

From Lecture 1 slide 9

http://ocw.jhsph.edu/courses/FundEpi/PDFs/Lecture1.pdf

• Survey surveys
• Surveillance studies
• Observational studies
• Experimental studies

Experimental studies are the most INVASIVE.

In any case, the intent is to collect DATA that can be subjected to HYPOTHESIS TESTING.

If the UNIT OF ANALYSIS is PERSON, then it is not enough to just have rates of something in a population. If you are testing to see if people who work in a certain industry tend to have higher levels of a certain illness, each line of data must include the ID of a person and data regarding industry and presence or absence of the illness.

Lecture Slide 11

http://ocw.jhsph.edu/courses/FundEpi/PDFs/Lecture1.pdf

DESCRIPTIVE STATISTICS should be reported in the research report but cannot alone be the basis of the testing of a hypothesis. The fact that 71 females and 64 males participated in the research (alone) tells you nothing about possible relationships between variables.

Lecture Slide 12

INFERENTIAL STATISTICAL METHODS (if the sample is random and the size of the sample is adequate) can be used to test hypotheses. I prefer Chi-Sq analysis when using ordinal variables (or nominal variables with only two values).

	Male	Female
does not have the condition	82% (n=82)	90% (n=90)
has the condition	18% (n=18)	10% (n=10)

Assume the table above was derived from a random sample of 200 persons (100 males and 100 females). Assume the hypothesis is as follows:

H1: that males are more likely than females to have the specified condition

• The percentage of males in the POPULATION with the condition is a PARAMETER which is unknown.
• The percentage of females in the POPULATION with the condition is a PARAMETER which is also unknown.
• The 18% is a STATISTIC, which is an estimate of the first PARAMETER.
• The 10% is a STATISTIC, which is an estimate of the second PARAMENTER.
• The sample must be RANDOM for us to have reason to believe that each STATISTICS is a good estimate of each PARAMETER.
• Let's assume that the sample is random. Males appear to be more likely to have the condition than females, based upon the 18% and the 10% STATISTICS.
• THE QUESTION BECOMES: In the (random) sample, a higher percentage of the males had the condition. Can we conclude that IN THE POPULATION that males are more likely than females to have the condition?
• IT IS POSSIBLE TO COMPUTE A STATISTIC CALLED CHI-SQ TO ANSWER THIS QUESTION. We may get to that later.
• LET'S ASSUME THAT "SOMEHOW" WE KNOW THAT IN THE POPULATION THERE IS NO RELATIONSHIP BETWEEN GENDER AND HAVING THE CONDITION.
• COULD WE POSSIBLY draw a random sample from the population in which 18% of the males have the condition and 10% of the females have the condition? YES
• IS IT LIKELY that we would draw a random sample from the population (in which there is no relationship between the IV and the DV) such that in the sample 18% of the males have the condition and 10% of the females have the condition? NO
• IF THE RELATIONSHIP APPARENT IN THE SAMPLE (between the IV and the DV) IS STRONG (and the sample is a random sample) THERE IS A SMALL CHANCE THAT THERE IS NO RELATIONSHIP BETWEEN THE IV AND THE DV IN THE POPULATION.
• The Chi-Sq statistic has a p value (probability) associated with it. If the p value is less than .05 (and the sample is random) we can assume that there is in fact a relationship between the IV and the DV in the population.

THIS COURSE IS ABOUT APPLYING REASON AND COMMON SENSE. Consider the slide # 17.

Johns Hopkins University is in the state of Maryland. The death rate in Maryland was about 805/100,000 in 2002.

Disney World is in the state of Florida. The death rate in Florida was about 1,004/100,000 in 2002.

Does the higher death rate in Florida mean that Florida is a “riskier” place to live than Maryland?

What other factors must be considered?

What sense do you make of the statements above?

That Maryland has a lower death rate because it has a good university hospital?

That Florida has a higher death rate because it has Disney World?

That it is more dangerous to live in Florida than in Maryland?

Slides 19 and 20 --

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