Let's start with a bit of recent news:
Biology and health is complicated. There are breakthroughs in understanding and diagnostic capabilities.
QUESTION: What are some of the big breakthroughs? Let's make an informal, rough timeline.
Things will continue to evolve. But not everything will be about high-tech science. Consider the issue of treating cancer. Here's an article:
QUESTIONS:
Points
Think of all the other important things that don't get studied because
these conditions don't exist.
Contrast this sort of study with:
Tables 1.1 and 1.2 from Essential Epidemiology. Note that you have to study people who don't get sick in order to figure out what is going on in the people who do get sick.
An article about BPA and its risks
Why is BPA so difficult to study?
Here's another area that's really difficult:
Points:
What's different between education and the cancer/palliative care settings?
What are the five modes of thinking that Fraser identifies as important in liberal arts?
Framing of a hypothesis and carrying out an experiment to test the hypothesis. He writes, "additional care must be taken to find convincing evidence that the hypothesized chain of causation, and not some confounding alternative, led to the observed pattern of events.”
QUESTION: What's a “confounding alternative”?
In the grippe-like outbreak he describes, they are looking for a pattern in space and time to track down the source of the illness.
Note the importance of defining what constitutes a case of the illness. The definition was general at first: “episode of fever and muscle aches after January 1, 1971”. Note that time enters into it. The definition was broadened to include a particular sign (a serumm antibody).
We'll be examining the sorts of studies that are done in epidemiology (and in social sciences) to draw plausible conclusions about causation when experiments cannot be done.
Constructing hypotheses based on perceived similarities to systems that are better understood.
This requires a strong knowledge base, since you need to be aware of what are the candidates for making an analogy.
We can't do much with this, because this course doesn't have a pre-requisite of a lot of medical knowledge. But the application of epidemiological methods to non-health issues is a form of analogic thinking. So you'll see us making analogies to non-medical issues, as with “teaching effectiveness,” above.
Working out the consequences of a set of assumptions. His example has to do with the ratio of vaccine types to non-vaccine types in people who have had the vaccine compared to people who have not. Assuming an efficacy for the vaccine lets you work out what this ratio should be; this lets you estimate the efficacy from observed ratios.
We'll be doing some of this in the estimation problems for the course. Also, the section on detection theory makes strong use of deductive thinking.
Constraints arise from trade-offs. We'll be examining this directly: trying to figure out when an intervention is worthwhile. Much of the health-care debate in Congress relates to this because money is limited.
We'll be talking about this explicitly.
I'm not so convinced about his claim that epidemiology is strongly related to teaching aesthetic values. Certainly there are some studies that are very clever and make efficient use of available resources and information.
Studying history is, I think, crucial to developing analogic thinking. By learning from the mistakes and successes of the past, by seeing how forces play out, you can get a better idea (by analogic thinking!) of how to structure things now.
It turns out that history has shaped epidemiology. Certainly, epidemics have strongly shaped history. This is pointed out in books like Guns, Germs, and Steel. In The Great Influenza, a claim is made that flu was responsible for World War II.
Break into groups of five: Introduce yourselves. Give your group a unique but short name, e.g., Aardvarks. Assign each person a letter: A, B, C, D, E. Write down a list of all the names and the letter for each person. Also, make a copy for me and hand it in to me. Some questions:
Have the D people introduce E people.
Each day in class, over the next several sessions, I'm going to ask
a person with a given letter to introduce the person with that number.
Stay in your groups of 5 for a few minutes.
What are the major causes of death today? What were they in 1900?
Report back to the class. Use this to lead into the Risk Charts.
QUESTIONS:
Show them this problem and the Guesstimation reading.
An article about waste in health care
