The next part we need to understand is how the results are analysed. An ideal medication would totally stop the medical condition for which it was prescribed, after a certain period of treatment. The figure on the right presents such an example: the percentage of people recovering as a function of time after starting the medication increases till it reaches 100%. Using a placebo, some people's condition improves, but the percentage of recovering people remains in the 20% region. Another aspect of medical analysis we need to understand is that of "risk reduction" (RR). Assume that in a certain population, the probability of getting heart disease is 2% per year. This means that 2 out of every 100 people, on average, will get the disease each year. After 5 years, 5 X 2=10 people (again, on average) will become ill. Now, what happens if they all take a medication that reduces their risk by 50%? Every year only 1% (1 person out of 100) will become sick, so after 5 years, 5 people will be sick. This is one way to look at risk reduction. Another way, is to say that since 1 person gets sick each year, it will now take 10 years for 10 people to get sick. So the outcome of medication could be interpreted as "time extension" as follows:
1. Without medication, 10 people get sick in 5 years.
2. With medication, 5 people get sick in 5 years, and 5 more people will get sick in the subsequent 5 years.
This has been recognized by a few in the medical establishment: Drs. Tan and Murphy (in Lancet, vol 354, p 1378, 1999) state concerning the statement that treatment saves lives - "The press are not the only culprits creating this problem because quite often, the researchers themselves condone or even initiate this perception."