News stories about the findings of medical studies often contain statements like: "Treatment X reduced the risk of heart attacks," "treatment Y causes cancer" and "treatment Z saves lives." These statements are qualitative -- they describe the effect without using numbers. But without numbers, its hard to tell how big the effect is -- and how meaningful it might be to you.
You apply this principle regularly in daily life. How you react to news that taxes are increasing, rents in your building are going up or salaries in your firm will rise depends entirely on the size of the increase. When it comes to money, it is hard to imagine anyone who would hear "going up" and not want to know what the real numbers were.
But surprisingly, many people don't demand the same kinds of numbers when judging medical findings. Yet the only way to assess the importance of medical research statements such as "treatment X reduced risk" or "saves lives" is to examine quantitative data -- the frequency of events in those receiving and not receiving treatment.
Even then, knowing the numbers is just half the story. You also need to know to what extent the numbers might apply to you. The tax increase, for instance, might apply only to people without dependents; if you're a parent of young children, you wouldn't be affected. Likewise, in the case of medical research, it is important to consider the characteristics of patients in the study -- particularly their age, sex and major risk factors or diseases.
Imagine a study designed to investigate a treatment to reduce heart attack risk. In considering how well the treatment works, you must also factor in patient characteristics that influence risk. Age, sex and whether or not a person smokes all matter a lot in determining how likely a person is to have heart problems. (See "What is the chance of having your first heart attack in the next 10 years?")
According to the table, the average 65-year-old man who smokes (10-year risk of heart attack = 16 percent) is 16 times more likely to have a heart attack in the next 10 years than is a 50-year-old woman who does not smoke (10-year risk of heart attack = 1 percent). A 65-year-old man's heart attack risk would be even higher if he had high blood pressure or if he had already had a heart attack.
Different people face very different risks of not only heart attack but also other health events. This is not only because their ages differ but also because they have different genetics and different environmental exposures. Consequently, the effectiveness of treatment will differ in different groups of people.
If you are a woman age 50 who doesn't smoke, a study showing a reduced risk of heart attack among 65-year-old men who smoke may not apply to you. Look for studies that involve patients that are similar to you. In other words, the more you resemble the patients being studied, the more likely the study findings will be relevant to you.
-- Steven Woloshin, Lisa M. Schwartz and H. Gilbert Welch
