The Implication of Risk Factors II

Lives are not "saved" with medications that reduce risks! At best, a few of the medicated people have a little longer to go until they contract the illness.

We will use various terms associated with risk in the various studies reported at this site. Here are some definitions you will need in order to understand the various studies we will be discussing.

Absolute Risk" - this tells us how many people (out of a specific group of people) will get the illness in a specific period. In our site, we will quote results using a 10-year period, and a group of 100 people for easy comparison.

For example, one might read that the risk of a heart attack for an average 60-year-old male is 12%. This means that if one tracks a group of 100 60-year-old males for 10 years, on average 12 of them would be expected to get a heart attack.

Relative Risk - This number compares the risk between two different groups of people. (In some studies an "Odds Ratio" is used. For simplicity, we will only use the term "Risk Ratio").

For example, non smokers in the previous example may have a heart attack risk rate of 10%, whereas smokers have a rate of 15%. Smokers have a higher risk, relative to non-smokers by 15/10 = 1.5; thus the Relative Risk of smoking is 1.5. Often, this is quoted by stating that smokers have a 50% higher risk of heart disease.

Margin of Error - All measurements have errors associated with them. For details on errors, please see our Error tutorial. In order to interpret a risk ratio, we need to know its "Margin of Error" (aka 95% confidence level interval) - this is the range in which the correct (or "true") result is expected to be.

For example, one might find that the heart attack rate is 12%, with a margin of error from 11% to 13%. This would mean that the average number of heart attack victims for 100 people over 10 years could be anywhere between 11 and 13, with the most probable number being 12.

Another example is that people taking vitamin C might have a relative risk of 0.9 for catching colds, with a margin of error between 0.7 and 1.1. You might think that taking this vitamin reduces your risk (0.9 - 10% less chance of getting a cold), but since the margin of error includes unity (1.0), this means that it is entirely possible that there is no reduction in risk associated with vitamin taking. (All these numbers are for illustrating the issues only, and should not be taken seriously!).

Previous | Next

Last Modification - Nov 28, 2004