That is one of those things that sound intuitively true, but isn't necessarily the case in practice. One example that I am more familiar with: thyroid nodules.

That is one of those things that sound intuitively true, but isn't necessarily the case in practice. One example that I am more familiar with: thyroid nodules. Recently, the medical guidelines have increased the size threshold for a biopsy, where you poke a needle in it to find out whether it is cancer or not. We could assume that diagnosing these cancers when they are smaller would always better, but that is not the case... On one end, it turns out when the nodule really is cancer, it is just as safe to wait to do the biopsy only after it has grown over the size threshold. And on the other end, the biopsies are not 100% accurate. A good percentage of the time, it can't say conclusively whether it is cancer or not. When that happens the only way to know for sure is to surgically remove the thyroid, which might result in the patient being put on lifelong hormone medication. One might think that in a perfect world we would act rationally over this uncertain information and not remove the thyroid when all it has is a small nodule with uncertain diagnosis. However, human intuition doesn't work well with percentages. When a patient is told they have a chance of having cancer, no matter how small, they will want to pursue every avenue to investigate and resolve the matter, even when the cost benefit is not worth it. When they looked into the statistics of how many people were getting thyroid surgery we could see that as imaging technology improved, more and more people had thyroid surgery, with no reduction in the number of deaths from thyroid cancer.