Is big data a "panacea"?
Written by | Lost little bees
Source | "Medical Community" public account
On March 1, 2020, the first case of COVID-19 infection was reported in the U.S. state of New York.
Nineteen days later, New York Governor Andrew Cuomo issued a statewide "stay-at-home order."
As the "heart" of New York State, New York City began to take measures such as working from home, block lockdowns and mass testing. People were asked to practice 2 meters of social distancing. Pharmacies make their own "protective nets" to protect staff and isolate potential patients.
Photo caption: In March 2020, workers sell medicines through protective plastic film at a pharmacy in Queens, New York City, USA. /News UN
But these initiatives have had little effect. By April of that year, there was a major outbreak in New York City, with an average of nearly 1,000 deaths per day and a six-fold increase in funerals.
The City of New York mobilized more than 300 refrigerated container trucks. The car was loaded with the bodies of the deceased. Soon, the freezer truck was not enough. The Forensic Service stipulates that the period of refrigeration of corpses is 14 days. When it expires unclaimed, it will be sent to Hart Island Cemetery for burial – described by the media as a "mass grave".
Image from AP
Epidemiologist Sharon Greene's office is in a high-rise building in Queens, New York City.
In her office, as she watched COVID-19 sweep across New York City, she worked with her team to develop an open source program called SaTScan.
Through this process, the research team produced a "real-time map of the spread of covid-19 in New York." "With detailed data from hospitals and laboratories, we found that the virus affects different regions differently."
Combined with maps, Sharon Greene and others can mediate resources and put them where they are most needed. For example, let the new crown pneumonia test truck park in the neighborhood with the most potential outbreak in a short period of time; guide the government to distribute masks, gloves and other epidemic prevention materials to the institutions most affected by the epidemic.
As the outbreak spread, Sharon Greene began to focus on "the rate of transmission growth," predicting "where rapid rates of cases are likely to occur."
"It's hard for us to move resources in real time." "But if people can be alerted earlier and avoid high-risk areas, they have more opportunities to protect themselves and avoid infection," Sharon Greene said. ”
Modern epidemiologists, represented by Sharon Greene, believe that this highly targeted approach, with information technology at its core, is an "emerging weapon" for public health and contributes to the realization of "precision public health."
Photo caption: A man is being tested for the coronavirus on a street in Manhattan, New York, USA. /The New York Times
Origins: Complementary to precision medicine
In 2013, Academician Tarun Weeramanthri, director of public health at the Western Australian Ministry of Health, took the lead in proposing "precision public health", saying that it was a supplement to precision medicine.
"Precision medicine" is to tailor the appropriate medical plan according to the different genes, metabolic conditions, lifestyle and environment of individuals. The cost is high and most people cannot benefit at this stage.
The definition of "precision public health" is broad and varied. In general, it refers to the population-centered, through genomics, bioinformatics, and cross-application of tools such as big data and artificial intelligence (AI), to define the risk of disease occurrence in the population, and determine feasible interventions, so as to accurately predict and prevent the occurrence and development of diseases and promote population health.
In theory, the two complement and promote each other. One focuses on individual therapy, and the other focuses on extensive prevention.
Development: It has been maturing for a long time
In the view of John Quackenbush, a biostatistician at Harvard University's Chen Zengxi School of Public Health, the birth and development of epidemiology is for "precision public health".
In the 1850s, cholera broke out in London, England. Dr John Snow went door-to-door and marked all confirmed cases of cholera on a map. He found that people who fetched water from the Broad Street pump had higher infection rates.
Since then, municipal public health departments have been set up in major cities such as London and New York, and the first thing they do is to purify drinking water, establish a sewage treatment system, remove garbage, and collect birth and death information. The implications are broad and most people can benefit from these interventions.
Dr. John Snow's attempt is seen as the prototype of "big data analytics."
"Dr. Snow's statistical method is simple and slightly clumsy. But in 1854, it was the most advanced idea. If Dr. Snow had been able to use SaTScan by then, he might have been able to stop cholera faster. John Quackenbush said.
Photo caption: Cholera outbreak in London in 1854. Dr. John Snow maps and accurately tracks cholera cases. /On the mode of Communication of Cholera
"Nature" pointed out that with the passage of time, the medical science community has learned more about diseases, epidemiology has become complex, involving a lot of knowledge and data, and the requirements for "precision" have increased accordingly.
Muin Khoury is director of the Office of Genomics and Precision Public Health at the Centers for Disease Control and Prevention. As a pediatrician and geneticist, he has been studying familial hypercholesterolemia (FH) for years.
It is a genetic disorder that occurs in a population of about 1%. Patients experience spiked cholesterol levels, or induce heart disease and stroke. If diagnosis can be diagnosed early and treated correctly, the risk of cardiovascular disease can be reduced by about 80%.
However, public health resources are limited, screening everyone is impractical, and it is easy to waste resources.
Currently, a U.S. team is working on AI to try to identify potential FH patients by scanning big data such as health records, including blood cholesterol values. This is more targeted, saves money and saves more lives.
Caitlin Allen, an epidemiologist at the University of South Carolina, is also testing AI. She recently completed a study that found that analyzing chatbot information on medical websites can predict whether a person needs to be tested for multiple cancer genes.
"This is one of the manifestations of precision public health. It is also presented to the public in many other forms. Nature said.
In 2016, zika virus swept through the western hemisphere. On September 1 of that year, more than a month after the outbreak of natively transmitted Zika in the United States, native mosquitoes were detected in Miami, Florida, infected with Zika virus.
Subsequently, local health officials used geographic information systems to draw "mosquito maps.". Ultimately, they chose to spray pesticides on only two blocks — rather than the entire city — to block the incursions of venomous mosquitoes.
Photo caption: In Miami, Florida, 2016, a health official sprayed insecticides in targeted communities to kill mosquitoes and stop zika virus transmission. /Getty
The "precision public health" based on big data is also reflected in many public welfare projects.
To help governments, charities and others develop informed public health policies, the Bill & Melinda Gates Foundation has allocated a total of $271 million to map and map areas of maternal health and child malnutrition in Asia and Africa.
During the COVID-19 pandemic, the Rockefeller Foundation gathered strategies to prevent the pandemic, created "artificial intelligence," and helped the education industry plan its schedule and make decisions about how to get students back to school safely.
In terms of tracking the new crown virus in real time, "precision public health" is also playing an impact.
"We sequenced the genomes of infectious pathogens to form several databases. Virologists can not only observe the variants first-time, but also unravel the subtle changes in their human-to-human transmission to determine who infects whom. Emma Thomson, a virologist at the University of Glasgow in the United Kingdom, said.
Paradox: should it be "precise" or "public"?
With the spread and application of "precision public health", advocates have begun to worry about its shortcomings.
Sandro Galea, an epidemiologist and dean of the Boston University School of Public Health in the United States, believes that the primary problem of "precision public health" is that its definition is unclear and it relies too much on high-tech technology.
"We understand that epidemiologists and other professionals need to use funky terminology and get financial support. But over-precision or deviation from its essence has resulted in many people not being able to have the most basic public health services. Sandro Galea expressed concern.
David Taylor Robinson of the University of Liverpool in the United Kingdom said that the excessive pursuit of "precision" may lead people to overlook less technological but more challenging health intervention strategies, such as universal health care and reducing income inequality. "We may be blinded by the data and drill the horns."
Angeline Ferdinand, a public health researcher at the University of Melbourne in Australia, pointed out that some areas do not even have basic birth and death information, let alone expect it to be able to sequence the whole genome. "Too precise, it will exacerbate inequalities within and between countries."
"In the wake of the COVID-19 pandemic, information technology companies promised to share private data with public health researchers to help them mine and analyze the data and find ways to overcome the spread of COVID-19." Two years on, it's clear that big data isn't a panacea. The U.S. medical news website STAT noted.
Nishant Kishore, Center for Infectious Disease Dynamics at the Harvard School of Public Health, has been working with health authorities to investigate the spread of the virus and revise epidemic prevention strategies. He said it's useful to get the data, but "unfortunately, the data is generated from parsing to parsing for commercial purposes, not public health." I looked at Google and it told me that more people were staying home during the outbreak, and Facebook said that its open rate dropped at a specific time, so it meant fewer people staying home. Who should I trust as a researcher? ”
bibliography:
Yang Yingying,Fu Chuanxi. Research Progress and Challenges of Precision Public Health. Chinese Journal of Public Health,2021,37(2):371-374.]
[3] Why big data didn’t deliver on its big promises to combat Covid-19.STAT
Source: Medical community
Editor-in-charge: Wan Shunshun
Proofreader: Zang Hengjia
Plate making: Xue Jiao