by Gordon Fox: 5th in a series
You've been told that you had a high risk - say, 5% - of something dangerous, like a lethal arrhythmia, in the next year. Twelve years later there hasn't been a hint of that event occurring. That's great! But many people understandably are annoyed: they were told they were at risk, they got the treatment (an ICD for example), but it hasn't actually been needed. Sometimes people feel as though they were misled.
Reasons this can happen
There are many reasons why expected events don't always occur. Some possibilities include these:
- The estimate was a probability, not a prediction. Doctors don't have crystal balls; they tell you that the chance of it happening is large or small, not that it
will or
won't occur. If you are told that the risk of a dangerous arrhythmia is high, and it doesn't occur, the doctors telling you that were not wrong.
Even with high risk estimates, sometimes the event never happens.
After all, if the risk of something bad is only 1% per year, about 1 out of 100 people will have that happen, each year.
- The risk estimate is for a population, but you're an individual!
If you were told that there is a 90% chance of a bad event occurring, that comes from the fact that 90% of people with similar characteristics had that event. But individuals either have the event or they don't. They don't have 90% of the event.
- The science wasn't understood as well when the estimate was made as it is today. Risk estimates are statistical, based on the science known at the time. But our knowledge grows over time. As we learn more, our estimates get modified. That is probably not what you want to hear as an individual, but it's encouraging for us in the long run!
The first two possibilities both mean that the estimate was good. It is
very
easy to misinterpret probabilities. Here's a confession from a scientist who has worked with statistics his entire career:
everyone misinterprets statistics sometimes, including me. Just as we tend to misinterpret a statement like "65% chance of rain today" as meaning that it will rain, most of us misinterpret a statement like "only 1% of people having this medical procedure have a negative outcome." We take it to mean something like “Oh, good, I won’t have this bad thing happen to me.” But some people do, and sometimes that person is you or me.
The third possibility - the science behind a risk estimate has changed - is the one that seems most disturbing. It doesn't happen that often, and because of the way science works, a bad estimate gets corrected eventually. The most important thing to realize here is that the best your doctors can do is to use the best science that's known at the time. That’s a lot better than asking your doctor to guess.
Conclusion
Most cases of "my risk estimate was wrong" are really misunderstandings about what a risk estimate is. Don't feel too bad if you make this mistake: it is very easy to do! Probability is one of the slipperiest ideas that we have to use.
Read Part 4:
Decisions in HCM: Where do Risk Estimates Come From?
HCMA Blog
