Disease Screening App

Disease screening apps are neither "online miracle doctors" that can replace hospital diagnosis, nor are they worthless "IQ tax products." They are essentially early warning tools to assist health management under the premise of compliance. The core boundary is that they can never replace clinical diagnosis.

Last month, my mother came across an advertisement for AI screening for lung nodules while browsing short videos. She excitedly downloaded the app and uploaded the chest X-ray from last year's workplace physical examination. On the same day, a report of "high-risk suspected lung cancer" came out. She was so frightened that she couldn't eat. She called me for half an hour overnight and told me that she didn't have a few days to live. I hurried home and accompanied her to a tertiary hospital to get admitted to the respiratory department. The doctor looked at the film under the light for two minutes and became happy. He said that it was just an old inflammatory nodule and that annual follow-up was enough. He didn’t even need to prescribe medicine.

I specifically asked an old classmate who works in chronic disease management at the Centers for Disease Control and Prevention, and she gave me a completely different perspective. Last year, they conducted a grassroots cancer screening pilot in Nujiang, Yunnan, and equipped village doctors with an official breast cancer screening app. The village doctors only need to upload the ultrasound images taken on site, and the background AI will first conduct a round of preliminary screening. Those suspected of being high-risk can go directly to the green channel of the state hospital for further diagnosis. Over the past six months, more than 4,000 villagers were screened, and 7 early-stage breast cancer patients were identified. All of them underwent surgery in time, and the cure rate was close to 100%. She showed me statistics from the National Cancer Center. A compliant disease screening app, combined with standardized sampling and inspection/imaging data from formal institutions, can provide an early warning warning rate of 82% for high-incidence cancers such as colorectal cancer and breast cancer, which is higher than the accuracy of manual preliminary screening in many grassroots health clinics in underdeveloped areas. For people who do not have the conditions to regularly go to large hospitals for physical examinations, this type of tool is equivalent to lowering the threshold for early screening.

But on the other hand, Fa Xiao, who is the attending physician in the respiratory department of a tertiary hospital, shook his head and said that he had seen at least 20 patients who were frightened by such apps in the past six months. The most regrettable thing is that a 28-year-old girl had a breast ultrasound for a physical examination at work. She thought it would be troublesome to go to the hospital to register and queue up, so she found a small app without medical qualifications to upload the report. The app concluded that it was "benign hyperplasia, just observe it regularly." However, she really didn't take it seriously. Three months later, she felt a lump in her breast and went to the hospital. She was already in the middle stage of breast cancer and missed the best opportunity for diagnosis and treatment. There are also many apps that have turned screening into a traffic business, using free screening on the front end to attract traffic. As long as you fill out the questionnaire with common habits such as staying up late and eating takeout, they will directly give you a 90-point risk score for gastric cancer/liver cancer, and then recommend you thousands of high-end physical examination packages and anti-cancer health products, purely by creating anxiety to make money.

It’s interesting to say that my friends who work in public health are all praising this kind of App, and my friends in clinical medicine are criticizing it. Both sides have solid opinions. The core difference lies in the word “compliance.” I now help my family members choose this type of app. The first thing I turn to is always the qualifications page. Is there an "Internet Healthcare Information Service License" and is it registered with the National Health Commission? If not, just leave it out. Another thing to look at is its screening logic. If you fill in your age and say you have frequent stomachaches and then directly label you as a "high risk gastric cancer", it is probably to trick you into buying subsequent products. Really reliable products will either require you to upload inspection reports and imaging data from a formal institution, or they will send you a standardized sampling kit. After collecting the samples, you will be sent back to the cooperating medical examination center to get the results. The algorithm is only an auxiliary risk stratification.

In fact, to put it bluntly, the disease screening app is similar to the automatic blood pressure monitor in the drugstore at your doorstep. If your blood pressure is high, you must go to the hospital for a review. You will not just buy medicine with the results of the machine, right? I have installed an early screening app for bowel cancer in cooperation with the Provincial Cancer Hospital for my mother. Once a year, I receive a sampling box to test for occult blood in the stool. If the app indicates a low risk, go for a normal physical examination. If the app indicates a high risk, go directly to the green channel of the cooperative hospital to register, which saves a lot of time in queuing. As for those who dare to judge you as terminally ill without any information, I have long since deleted them all.

After all, whether a tool is useful or not depends on whether the person using it understands its boundaries. If you really feel unwell, don't make blind guesses with the app. It's safest to go to the hospital and see a doctor.