Tuesday, April 30, 2013

Why only 9%?

I am going to get the first mammogram. To know what kind of water I am treading into, I researched a little. Now I know 0.8% of women getting tested have cancer, and of course it is great to know whether I have cancer or not right now. However, mammogram is not a perfect indication yet. If I have cancer, there is 90% chance that the mammogram machine would give positive results. In addition to that, it could also be wrong even if I don't have cancer. So there is 7% chance that the mammogram will show positive even if I am perfectly healthy. So suppose my test result is positive, what is the chance that I have breast cancer?

At first I thought it was 90%. But actually, it is not.

Here is the math. Now suppose I was tested with 999 other women, total 1000. Now we know by average, there are eight of us currently have cancer. Now we group these eight women into cancer group, and the rest of us goes to health group, except I don't know which group I would be in. Among the cancer group, 90% will have positive mammogram, so seven women will be identified with cancer. Among the healthy group, given that the machine could be wrong at 7% of the time, 69 women will be identified as well. So if I have a report with positive result, I am among the 76 women who is identified with cancer by machine. Among us, only 7 actually are cancer patient. So, my chance of having cancer with a positive mammogram is only 7 over 76, or 9%.


No comments:

Post a Comment