A few years ago, a Japanese commercial broadcaster"Back pain and stiff shoulders are cancerred light!"I aired a program called Such information risks misleading the public.

In this program, based on the results of a survey of people with cancer, they concluded that back pain and stiff shoulders were red flags for cancer, but there is a serious error in this argument.

If a person with back pain or stiff shoulders develops cancer, it is true that the pain tends to become stronger, but 99% of common back pain and stiff shoulders are not cancer.

The conclusion that low back pain and stiff shoulders are red flags of cancer is "magic of numbers". There is an error in the way the probability is calculated.

If the population is set to cancer patients and the frequency of low back pain and stiff shoulders is calculated, it will appear as a number that cannot be ignored. In a sense, this is a natural result, as the target population is a high-risk group who should have experienced some kind of abnormality in their physical condition.

However, instead of that, according to the title (purpose) of this program, the idea is to find the probability that cancer is lurking in back pain and stiff shoulders.

BReINIn this case,The cancer incidence rate must be calculated by setting the population to patients with back pain and stiff shoulders. So that number is less than 1%.(People who have already undergone cancer treatment and pain caused by bone metastasis during cancer treatment are not included).

BReINFor example, in a survey conducted at public hospitals in the United States, 13 out of 1,975 patients who visited the hospital with low back pain reported pain due to cancer. This is data only for low back pain, ,0.66%It will be a number.

According to a survey conducted at a clinic belonging to our association, the probability that cancer was found among patients with back pain and stiff shoulders who visited us in the past three years was 0.8%.

In other words, if there are 1,000 patients with back pain or stiff shoulders, only 8 of them will be diagnosed with cancer. Based on these numbers, can we say that back pain and stiff shoulders are red flags for cancer? I would explain it like this:

“Among the physical changes that people with cancer experience are back and neck pain. but,Cancer is found in less than 1% of patients who visit medical institutions with back pain or stiff shoulders...

Therefore, 99% of back pain and stiff shoulders are not related to cancer. However, it is more accurate for those who have severe night sweats with night sweats lasting more than 3 weeks, those who have excruciating resting pain during the day for more than several hours, those who have significant weight loss for no apparent reason, and those who have a history of cancer. should be examined

[Note: A similar statistic for medical centers and university hospitals might put the figure in excess of 1%, but this is a difficult figure to adopt as a generalization. This is because there is a statistical bias in places where severe cases gather. Also, even if you have back pain or stiff shoulders, there are some people who do not seek medical attention, so even if you set such people as the population, the numbers may change.]


Slightly off topic, but what do you think when you are told that the success rate of surgery is 80%? Some people are relieved that the success rate is high, while others are worried that the probability of failure is as high as 20%. there is.

This is because in some cases the denominator is not known when calculating the probability.In this case, it is not clear who performed the surgery. I don't know if the denominator is the number for "individual surgeon" or "whole hospital". In some cases, it may be the figures announced by the “Society”.

BReINThe meaning of "80%" changes depending on what is set as the denominator.BReIN

BReINBy the way, if I were a patient, I would like to know the natural frequency of how many patients were operated on and how many of them were successful, not the number of the surgeon himself or the percentage of the probability.

The way of thinking about probability and statistics is difficult, but I think it is necessary for those who transmit information to study it properly, and for those who receive it, it is better for them to have such knowledge.

For those interested, the book "Calculated Risks" by Gerd Gigerenzer is a great reference. In Japan, it is published by Hayakawa Shobo under the Japanese title "Your Surprisingly Dangerous Life with Weakness in Numbers: How to Avoid Being Deceived by Statistics in Hospitals and Courts".


BReIN
By the way, in Japanese, the proper use of "wa" and "ga" under the subject often becomes a problem. A long time ago, a Japanese scholar named Akira Mikami explained in his book "Elephant has a long nose" that the "ha" in "Watashi wa..." means "As far as I am concerned..." .

BReINIn other words, the correct usage of "wa" in Japanese is for the explanation of the subject "watashi" to follow in the sentence. "Elephant has a long noseis the correct answer, and we do not say, "Elephants have long trunks."

BReINThe TV program of the example put out in a big telop“Back pain and stiff shoulders are red flags of cancer”When you think of the Japanese word, it means "when it comes to back pain and stiff shoulders, it's a red flag for cancer."Of course, the denominator must be "patients with back pain and stiff shoulders."

BReINIf you are dealing with the relationship between back pain/stiff shoulders and cancer, shouldn't you try to express yourself more carefully so as not to give unnecessary anxiety to the viewers?

In general, musculoskeletal problems, represented by back pain and stiff shoulders, are diagnosed by imaging tests such as X-rays and MRIs, but important facts that shake the foundation have been found. Next time, I will talk about that (about EBM).

For those interested in Cognitive Science Integrated Approach (COSIA)

The Cognitive Science Integrated Approach (COSIA) is a concept that represents the “integration of cognitive science and medicine” and has its origins in pain management in musculoskeletal primary care.

In the current situation where image labeling and diagnosis of the cause of pain diverge, the International Association for the Study of Pain (IASP) has revised the definition of pain and is shifting to the "emotional origin theory of pain."

If you are interested in COSIA, please visit the "Special Site for Medical Professionals and Therapists". We look forward to your participation.