Lemonade-Ed - Testing Medicines (Activity)

Activity: Testing Medicines

A New Medicine

How do scientists design new medicines? They start by looking at what causes disease and how it affects the body. This gives them clues about which chemicals might help to treat the disease.

Before they can be tested on humans, new medicines are first tested on cell cultures. These are collections of cells that have been grown in a laboratory. Successful medicines are then tested on animals to make sure that they are safe to use before trying them on humans.

Clinical trials

Healthy volunteers are given doses of the new medicine. Scientists can then study how the chemical works in the body to check that it is safe to use. After this, the Medicines Control Agency is asked for permission to run trials on patients. These trials show how well the medicine really works. Clinical trials take three to five years.

Correlation

The scientists need to see if the medicine is going to make patients get better. They want to see if there is a link, or a correlation, between the patient taking the medicine and getting better.

Scientists usually divide the patients into two groups. One group is given the new medicine, and the other group is given a placebo. A placebo acts as a control - it does not contain any medicine at all. Neither the patients nor the doctors giving the medicines know which is the real medicine and which is the placebo. This is called a double-blind trial.

There are three types of correlation:

If the medicine works then there is a positive correlation.
If the medicine does not work then there is a negative correlation.
If the numbers of people getting better and not getting better are the same then there is no correlation.

Questions

- Explain why it is necessary to give some patients a placebo.
- Why do you think a double-blind trial is important?

Clinical Trial Results

Dr Franklin was testing a new antiviral medicine to treat a type of influenza. She tested 40 patients.

Half were given medicine B182 and the other half were given a placebo. Out of the 20 patients given medicine B182, 15 showed a positive effect (an improvement) and 5 showed no effect. Out of the 20 patients given the placebo, 1 patient showed a positive effect and 19 showed no effect.

To see if there is a correlation between taking medicine B182 and an improvement in the influenza, we use a table like the one on the right.

The 15 people in box A proved that the medicine worked because most of the people treated with the medicine got better. The 19 people in box D also proved that the new medicine worked because all but one of the people who were not treated with the medicine did not get better.

To find out what the correlation is in the results, we calculate the ratio between A + D and C + B:

If the numbers in the yellow boxes add up to more than the numbers in the brown

If the numbers in the yellow boxes add up to less than the numbers in the

If the numbers are the same then there is no correlation.

Questions:

What type of correlation is there for medicine B182?
What does this test tell you about medicine B182?
What do you think the reason might be for one person having a positive effect even though they did not receive the medicine?

Further tests for reliability

Dr Franklin then set up another experiment for medicine B182 in another part of the country. She did this to check how reliable her results were.

She gave 20 people medicine B182 and another 20 people took a placebo.

Here are the results.

- Out of the 20 people using medicine B182, 2 showed signs of improvement and 18 did not.
- Out of the 20 people not using the medicine, 18 showed signs of improvement and 2 did not.

Questions:

Put these results in a table like the one above.
Calculate the ratio A + D: B + C.
What type of correlation does this show?
What does this test tell you about medicine B182?

Page updated

Google Sites

Report abuse