This article is from WeChat public account: biokiwi (ID: biokiwi) , author: Tadpole
We often say that we should learn from history, and after epidemics, people will always think of ways to prevent the recurrence of world epidemics.
After the SARS incident in 2003, China has successively revised the Law of the People’s Republic of China on the Prevention and Control of Infectious Diseases, and formulated the “Administrative Measures for Reporting Information on Surveillance of Public Health Incidents and Infectious Disease Epidemics”, which are applicable to infectious diseases The reporting system allows health care institutions to report according to the severity of the disease.
“Administrative Measures for Reporting of Surveillance Information of Public Health Emergencies and Infectious Disease Epidemics” (screenshot from the official website of the National Health and Medical Commission)
These regulations and methods are major advances in epidemic surveillance, and they have also enabled China to perform well in the face of bird flu and swine flu.
But think about it, would it be a little slower if you waited until the infectious disease appeared? If the virus or other pathogens can be predicted in advance, like weather forecasting and earthquake monitoring, before they become epidemic in humans, and stop it in timeMaybe the epidemic would not be so serious?
I have to mention an emerging discipline here—epidemic prevention, which has three main purposes:
1. Identify endemic diseases early;
2. Assess the probability of an endemic epidemic becoming a global epidemic;
3. Stop deadly endemic epidemics before they become global epidemics.
Back to the question in the title: How does epidemiology prevent epidemics like weather forecasts?
Staring at the “Sentinel Crowd”
New viruses or other pathogens often jump from animals to humans and then spread among humans.
For example, recent new coronaviruses, previous outbreaks of SARS virus, Ebola virus, or the rabies virus that occasionally occur, are all derived from animals.
Before viruses attack humans, they are often parasitic on wild animals. For example, the coronavirus that we all know now comes from various wild animals. p>
In the process of jumping, a small number of people may be more likely to be infected than others. Such people are called-sentinel crowds.
What are the sentinel crowds?
For example, those hunters who still hunt for a living, they are in close contact with the wild animals in the forest, and they may even touch the blood of the animals.
If a virus wants to jump from these wildlife to humansIf you go, they are likely to be the first infected group.
Except for hunters, rangers, forest park managers, slaughterhouse staff, etc., all belong to the “sentinel crowd.”
Through regular monitoring of blood samples from hunters in hunting areas, infectious disease scientists can track the migration of pathogens such as viruses from these forests to humans.
If you use the same model to establish a global control system to monitor the physiological status of all “outposts” and block viruses and microorganisms that want to behave in a human group, you can avoid the epidemic.
2015-2017 Guangdong animal health workers survey (Zhu Yanshan et al., 2018)
The sample from the sentinel crowd is just one of the monitoring kits. There are many other tools that can be used to gain information on pathogens and trends in infectious diseases.
Be alert to abnormal deaths of wild animals
In addition to the sentinel crowd, the wildlife itself deserves attention.
Wild zoologists can observe whether the animals in the forest are in a normal state, and whether they have strangely died in large areas. If so, what are the reasons.
A large number of wild animal deaths are likely to be trailers of human epidemics. If we understand the trailer, we can calmly cope with the next story.
The rangers in Uganda Nature Reserve require the timely report of the death of wild animals (source: Science)
For example, the epidemic virus Ebola, which has caused several fevers and severe internal bleeding in Africa for decades.
Because of the lack of specific drugs and vaccines, the vast majority of infected people cannot escape the fate of death.
Ebola virus under the microscope (source: Wikipedia)
To this end, scientists have collaborated with the Frenchvile International Medical Research Center of the Gabonese Republic, recording from 2001 to 2003 in forests spanning more than 20,000 square kilometers across Gabon and the Republic of the Congo, including gorillas, chimpanzees, Necropsy information of various wild animals including monkeys and antelopes.
It was found that the deaths of 98 animals mostly occurred during the two peak periods of the Ebola virus epidemic.
Researchers in protective clothing subsequently found the remains of 21 animals infected with Ebola virus, collected and tested the bones, muscles, and skin tissues of animals that had not been decomposed, and found 10 of them. Three chimpanzees and one antelope still tested positive.
Schematic diagram of Ebola virus transmission, mainly transmission and infection caused by contact with humans and wildlife
Before the two Ebola virus outbreaks in December 2002 and November 2003, researchers monitored wild animals several weeks or even months in advance, alerting them to rapid progress. Initiate precautionary measures.
But due to the lack of efficient and reliable communication channels between scientists and public health agencies and local people, even if the surveillance network has issued warnings, the virus may still infect humans and spread widely.
The Internet is a very convenient communication channel. Not only that, information from the Internet itself can also help predict epidemics.
Use AI and Big Data
People spend a lot of time “surfing the web” every day, posting updates on various social media, and searching for information on search engines.
These informational data from people going online can also be used to predict the risk of epidemics.
Because once an epidemic occurs, everyone will search the Internet for information on how to deal with various symptoms of the body or pathogens such as viruses.
The high-frequency keywords and data appearing in these search engines are very helpful for predicting the process of identification, occurrence and development of some epidemics.
For example, the flu epidemic trend prediction system launched by Google in 2008, Google Flu Trends, built two billion weeks before the H1N1 outbreak in the U.S. in the following year, and constructed 450 million different numbers. Model, the relevant influenza prediction index is obtained, and the result is highly correlated with the official data of the US Centers for Disease Control and Prevention (CDC) (97%) .
If this system can maintain such predictability and accuracy, this will undoubtedly be a powerful weapon against influenza.
But contrary to expectations, an article published in Science in 2014 pointed out that the results given by Google Flu Trends are often higher than the actual situation, and are extremely high: From August 2011 to August 2013, Among 108 weeks, 100 weeks were higher than the flu incidence rate given by CDC, and the highest period was even twice the CDC data.
Comparison of Google Flu Trends and CDC Reports (source: Science)
Baidu has also previously launched epidemiological trend prediction systems including influenza, hepatitis, tuberculosis and sexually transmitted diseases.
But whether it is data collection and screening, or the form of model construction, it may affect the final results.
If you add artificial intelligence (AI) algorithm on the basis of artificial analysis, use more complete technical strategies and prediction models to reduce Unnecessary “noise” may lead to more accurate results. At present, there are several companies around the world, and such programs are being developed or even launched.