Pineapple Says
“False Negatives” and “False Positives”
The earlier the cancer, the better the cure, of course. However, there are still no good screening methods for many cancers.
Each effective cancer screening is strictly limited in who and what type of test it uses, no screening is right for everyone, and no screening is right for two Cancers above.
Screening for liver cancer, stomach cancer, and lung cancer is limited to high-risk groups, and screening is not recommended for the general public.
Because any screening has “false negatives” and “false positives” — a “false negative” is when you have a disease but a normal test result, while a “false negative” A false positive” means no disease but an abnormal test.
In medicine, no test is 100% accurate. Various tests done by the hospital, whether it is blood glucose test or HIV test, have a certain probability of false negative and false positive, but This probability is very low.
Math in Cancer Screening
Pineapple and everyone play a simple math game.
The incidence of lung cancer in the general population under the age of 50 is less than 0.1%. Suppose there are 1 million people in a city, and 1000 people actually have early stage cancer. Assuming that the specificity and sensitivity of low-dose helical CT detection are both 99%, which means that 99% of the sick will be detected, and 99% of the unaffected will be correctly excluded. Sounds good, right? But what about the results of the medical examination? (Table 1)
The first thing you can see is that if there is no cancer, it almost certainly is. Because 989010/(10+989010)=99.999% of negative results are reliable.
The problem with this checkup is that it’s hard to find someone who actually has cancer!
Because in this case, 1000 X 99% = 990 people with cancer would be diagnosed, and 1% of people without cancer (999,000) would be misdiagnosed Without cancer, that’s 999,000 X 1% = 9,990 people. Then the whole physical examination, there will be 990 + 9990 = 10980 people diagnosed with cancer, of which 9990 are misdiagnosed, the false positive rate is as high as 91%!
That is to say, 91% of people who are tested positive for cancer are actually fine. And the 990 people who really have cancer will also be mixed in with the 9990 people, and it is impossible to distinguish. Therefore, all people need to do a second test, or even a third test, before it can be truly confirmed.
A test with 99% specificity and sensitivity, how come the final false positive and misdiagnosis rate is as high as 91%?
The root cause is that the incidence of lung cancer in the general population is very low (less than 0.1%), and there are very few people who are actually sick, so even a little false positive Probability, it will also lead to a large number of people who are not sick being misdiagnosed, far more than real patients. The result is that the real patients are submerged in a crowd of people who are scared to death but are not sick, and it does no one any good.
The specificity and sensitivity of cancer screenings currently on the market are well below 99%, even below 90%. So in the same topic, if the specificity and sensitivity are reduced to 90%, what will be the positive misdiagnosis rate?
Yes 98.3%!
Because of this, Pineapple does not recommend various cancer screenings, especially CT screenings that affect the body. But like many doctors, I strongly recommend that people at high risk of cancer (long-term smokers over 55 years old, with family history, known oncogene mutations, etc.) regularly go for reliable cancer screening. Why?
Mainly because these people are much more likely to have early-stage cancer than the general population, the false-positive rate of physical exams will drop significantly. For example, Julie, who had her breasts and ovaries removed due to BRCA1 gene mutation, has a 10% to 50% chance of having cancer in her organs. Let’s say we take an average of 30% and give 1 million images of Zhu Li. High-risk patients like Li are screened with the same false-negative rate and false-positive rate of 1%. What are the results? (Table 2)
The false-positive misdiagnosis rate is only 2%, instead of the 83% just now.
In such situations, cancer screening can be very valuable.
Summary
To increase the usefulness of cancer screening, two things are needed:
Improved specificity and sensitivity
targeted screening of high-risk groups
Cancer screening is a particularly good concept, and everyone should pay attention, but you must first know whether a certain screening method is suitable for your situation. I am not against cancer screening, but against people being fooled and overusing ineffective “screening methods”.
Salute to life!
*This article aims to popularize the science behind cancer screening, not drug promotional materials, nor treatment recommendations. For guidance on disease treatment plans, please go to a regular hospital for treatment.