New York City hospital ‘looks completely out of sight’ of U.S. math as life-saving straw

Over the past few days, New York City’s hospitals, large and small, have become “completely out of sight” as thousands of new coronary pneumonia patients have flocked to emergency rooms and ICU wards. In Seattle, 3,000 miles away, Lisa Brandenburg saw a similar scene: the hospital lobby was transformed into an isolation ward, nurses made of garbage bags to protect themselves, and cadaver trucks kept moving through the streets.

New York City hospital 'looks completely out of sight' of U.S. math as life-saving straw

A week ago, Seattle was at the center of a new outbreak of pneumonia in the United States. The first confirmed case of new coronary pneumonia in the United States occurred in Seattle in January, and in February, the first new case of pneumonia in the United States occurred in Seattle.

As head of the Washington University School of Medicine Hospital, Brandenburg leads the region’s largest medical network, with more than half a million outpatients a year. In early March, she and many public health officials were shocked by a report by computational biologists at the Fred Hutchinson Cancer Research Center. Their analysis of genetic information shows that the new coronavirus has been “quietly” spreading in the Seattle area over the past few weeks, infecting at least 500-600 people. The outbreak has become a time bomb that could explode in Seattle at any time.

Seattle’s mayor declared a state of emergency, schools were closed, King County and Snohomish Counties called a halt to rallies of more than 250 people, and the tourist attraction, the Space Needle, stopped receiving visitors. Seattleers think the government should take further anti-epidemic measures and petition the governor to implement a statewide ban.

But Brandenburg is concerned about other questions: How many patients will need to be hospitalized? How many patients need first aid? When will patients rush into the hospital? Does the hospital have enough ventilators?

No one can answer these questions with certainty. Hospital managers such as Brandenburg can only estimate the future based on past experience, and use it to procure as many ventilators as possible, recruit ICU nurses, empty beds, and be ready to arrive in the new coronary pneumonia epidemic.

Mathematical model into anti-epidemic pioneer

Chris Murray and his computer simulations have worked.

Murray is director of the Institute of Health Indicators and Evaluation at the University of Washington, a group of about 500 statisticians, computer scientists and epidemiologists with strong data analysis capabilities. Each year, it publishes the Global Burden of Disease Study, which quantifies the incidence of each disease in 195 countries and territories around the world and their impact.

In February, Murray and dozens of his subordinates put all their energy into predicting the spread of the new corona epidemic in the United States. In particular, they are trying to help hospitals prepare for the impending crisis, the first of which is the University of Washington’s health system. Mr Brandenburg said the collaboration between the two agencies had saved many lives and “their research will enable hospitals to know when they will flood into the large number of patients”.

However, it is debatable whether Murray’s research will work. During the outbreak, it is impossible to obtain exact data. Chinese researchers have published several papers on the spread of the new coronavirus in Hubei. Inadequate detection capacity means that researchers do not have enough samples to study the spread of the new coronavirus in the United States. Since the 2009 H1N1 outbreak, researchers around the world have increasingly relied on mathematical models and computer simulations to predict the spread of infectious diseases. Like many universities, federal agencies such as the CDC and the National Institutes of Health have their own modeling teams.

As with Earth’s climate change or nuclear bomb explosion simulations, computer simulations of the spread of infectious diseases are used to make well-founded predictions in the face of a great seamount of uncertainty.

When less data is available – usually when a virus begins to pass from person to person – different models differ in assumptions, conclusions, and so on. But government departments still advertise their models, and well-known modeling labs regularly publish large numbers of reports, and policymakers still rely on models when making decisions. There are still significant differences in the prediction of new crown pneumonia deaths globally by different models. No model is perfect, but with proper use, it’s better to have a model than nothing.

The principle of virus propagation calculation is not complicated

The basic mathematical principles of the infectious disease calculation model are not complicated, and with a little explanation, the general public can understand them. The most basic model of transmission of infectious diseases is the SIR model, which divides people into three groups: susceptible, infected, and re-recovered or removed (i.e., those who live, gain immunity, or die).

Some models add an “E” population (SEIR model) – a group of people who are not infected but have been exposed to the virus. The researchers made decisions based on their knowledge of the spread of the virus. Variables that affect the spread of the virus, including the number of people infected before they are treated or die, and the amount of time it takes for an infected person to infect a person.

“When the epidemic began to spread, everyone was susceptible, there were very few infections, they were infected, they were infected, and the number of infections would grow rapidly,” said Helen Jenkins, an infectious disease scientist at Boston University’s School of Public Health. “

The assumption of the proportion of each population and the rate of conversion between the population is very important. “If only 5 percent of the population were cured and immunized, that would mean that 95 percent of the population was vulnerable, and the risk of an infectious disease outbreak would be much higher; if 50 percent of the population was infected – many of them asymptomatic and we didn’t know they were infected, the epidemic pressure would be much less,” Jenkins said. “

The next question to answer is the “ability” of human transmission of the virus, known as R0 (basic infectionnumbers), which depends on the difficulty of the virus spreading from person to person – regardless of the patient’s symptoms. Equally important data also includes how many close contacts the infected person has and the time it takes to truly become infected (this is why maintaining a safe social distance helps to curb the spread of the virus, which reduces close contacts). The interval between human-to-human transmission, the time when an infected person infects another person, or when an susceptible person becomes infected or cures (or dies) of an infected person.

R0 is important only at the beginning of an outbreak, when the virus has not mutated and most people are susceptible. As the proportion of different populations changes, epidemiologists are looking at another data: the number of effective spreads.

At this point, it is assumed that readers have understood that different data can generate different models.

Some data is deliberately magnified to simulate the worst-case scenario. It makes sense to simulate the worst-case scenario, because it can prompt people to take actions that help curb the spread of the virus. Unfortunately, once these measures take effect, models will be “unjustly wronged”: they will be considered wrong. The real significance of these mathematical models is to prompt people to take steps to ensure that the worst of predictions does not become a reality.

At a White House press conference on Thursday, Deborah Birx, co-ordinator of the New Coronary Pneumonia Task Force, called on the media not to pay too much attention to the models, “which are not true.” “

New York City hospital 'looks completely out of sight' of U.S. math as life-saving straw

Keeping a safe social distance can change the speed at the time of virus transmission

In response to Mr. Burks on Twitter, Marc Lipsitch, an epidemiologist at Harvard Medical School, said Mr. Burks had mentioned his lab’s research, which the federal government had asked the federal government to do two days earlier. In a paper published on the preprinted website, his team used the SEIR model to simulate the spread of the virus under strict and loose isolation conditions. He simulated the spread of outbreaks at multiple R0 values.

In the model, stopping strict quarantine measures without a vaccine or special drugs would dramatically increase the number of people infected, with two serious cases per 1,000 people – meaning 660,000 serious illnesses or deaths in the United States. Even if the strictest quarantines are extended from April to August, the new coronavirus will make a comeback in the fall.

The real purpose of maintaining a safe social distance is not to limit people’s social activities, but to slow the spread of the virus, to ensure that the number of patients requiring admission to the hospital exceeds the capacity of the hospital at any time, and to buy time for scientists to develop vaccines and special effects.

Model Prediction Leads To Change Epidemic Prevention Strategy in Britain and The United States

If the Lipschurch team’s conclusions are correct, the new corona outbreak could be repeated by the end of 2022. Without any leaks, large-scale nucleic acid testing, isolation of patients and aggressive bans would reduce the number of people infected, he said, reducing the duration of the outbreak. But Lipsey said on Twitter that he had not seen the U.S. government taking such steps.

So here’s the question: Is The basis of Burks’s decision that the model’s most optimistic expectations will prove correct? “I was so impressed by her comments,” Yonatan Grad, an epidemiologist at the Harvard School of Public Health, said at a news conference Friday. “

Mr Burks also referred to an influential report published earlier this month by Imperial College London, which predicted that the number of new deaths from pneumonia in the UK would reach 500,000 this year, prompting the British government to abandon its policy of inaction and waiting for Britons to gain group immunity.

The report also predicts that 2.2 million Americans will die from new coronary pneumonia if the government does nothing, drawing the attention of U.S. President Donald Trump. Shortly thereafter, the White House issued a 15-day ban, encouraging Americans not to go out and stay at home if they were not necessary. Neil Ferguson, a researcher at Imperial College London, attracted the attention of the Trump team last week when he submitted a new report to the British Parliament predicting that the number of deaths from new crown pneumonia in the UK would be less than 20,000. “The number of deaths is expected to drop from 500,000 to 20,000, and we’ll learn more about why,” Burks said at a news conference Thursday. “

Ferguson didn’t really give up the original estimate or model. As he explained in a series of subsequent tweets, the new forecast is the result of two factors: the UK government has adopted a ban on footing, a slightly higher R0 (which suggests that the outbreak is spreading faster than previously thought, so the number of people infected is higher than anyone expected, and most of them are mildly ill). Ferguson says this is more evidence of the importance of maintaining safe social distance.

It’s important to note that Ferguson updated the model when the new data was released. But Mr Ferguson’s changes come at a sensitive time in politics. Just a few days ago, the British media began to preach that there was no need to panic over the new corona virus. Half of all British people are infected with the new coronavirus and get immunity, a new study suggests. In fact, the study’s conclusions are not, but the simultaneous appearance of both in the media will give the public the impression that there is no need to worry about the new coronavirus.

Researchers at the University of Oxford studied the number of deaths in Italy and the UK before the ban was issued to see what contributed to the high mortality rate. They found that one possible explanation was that the virus had only just begun to spread in the UK, and that a high proportion of confirmed patients were seriously ill – as the imperial institute’s model shows.

According to the model, another possible reasonable explanation is that the new coronavirus has been circulating in the UK since January and could infect up to half the UK population. In this case, most patients with new coronary pneumonia are mild, and only a small number of patients need admission. In other words, in the first case, the new coronavirus has only just begun to spread;

Sunetra Gupta, a theoretical epidemiologist and head of the Oxford University study, said the second scenario was “great news” because it meant that a significant proportion of Britons were already immune. While this is only one scenario that Mr. Gupte’s model predicts, it is a far cry from Ferguson’s prediction, enough to make some of the media hype.

It should be noted that some asymptomatic infections , which may also be quite large , have unknowingly helped spread the virus , and they are infected in the form of susceptible populations. This is illustrated by the strict city closure measures taken in Wuhan in January, which slowed the spread of the virus. Wuhan city for the rest of the world to fight the epidemic time, it is regrettable that many countries and regions, including the United States, have wasted China at a great cost to fight this precious time.

When will the outbreak end?

Given the large amount of data generated by more than 160,000 confirmed cases in the United States, the next question the model needs to predict is when the ban will be stopped. “No one can finalize this time, and all that can be done is a new coronavirus,” Anthony Fauci, director of the American Institute of Infectious Diseases, told CNN. “It is not clear whether children can transmit the new coronavirus, nor is it clear that the infectiousness of light patients in adults is not clear.

“One of the things we have to do is to test for serological tests to see if there are antibodies in a large population,” Jenkins said. “Testing antibodies can tell if a person is infected — whether it’s symptoms or no symptoms. Large-scale antibody testing can understand the number of susceptibility and recovery population, usually, human infection with the new coronavirus about a week, antibody testing is effective.

“There is no doubt about the importance of the model in predicting how we will survive the current crisis and the long-term impact of this outbreak,” Jenkins said. The evidence is not complicated for outbreaks in Europe and the United States. If we take swift action, we can reduce the number of early deaths and get through the early stages of the outbreak more quickly. “

Jenkins says previous studies have been clearer than models, especially if humans have not yet fully recognized the new coronavirus, “and one drawback of the model is that it quickly becomes complex, and the quality of the model depends on the data entered”.

The obstacles to the model are not just data. It is also unclear how policymakers understand the model. While government departments across the United States are using models to justify the imposition of bans and the addition of beds, political leaders like Mr. Burks can easily deny them.

Mr. Trump also doesn’t seem entirely agreed with the model’s predictions. “I don’t believe New York needs three or four thousand ventilators,” Trump said on a television show after New York Gov. Andrew Cuomo, who predicted that the federal government would support tens of thousands of ventilators, said in a television program. Two days later, In a tweet, Mr. Trump said Ford and General Motors should mass produce ventilators, even hinting that they would use the law.

Simulation to prepare for epidemic prevention

Brandenburg values models more than their own feelings. On March 7, washington University School of Medicine Hospital received the first confirmed case of new coronary pneumonia. Three days later, she asked Murray and his team for help. The next Tuesday (March 17), the Murray team released an expectation of three scenarios, and in the worst case, washington University School of Medicine Hospital will need to see 950 new coronapneumonia patients a day, with a peak coming april 7. With only about 1,500 beds in the four hospital districts, the hospital will be overrun.

Brandenburg began preparations immediately, ordering more masks, gloves, masks and ventilators, according to the model’s predictions, and even opened a stop-and-run test ingress. The Brandenburg team began to respond to the possible peak of the outbreak, eliminating all non-essential operations and freeing up as many beds as possible. Hospitals have set up tents outside emergency rooms to prevent cross-infection, and they have recalled ICU nurses who retired for five years and trained health care workers in relation to new coronary pneumonia.

As new data emerges, Murray updates his forecasts every day. The latest model, updated over the weekend, shows the first decline in new cases and a more flattened growth curve. According to the updated model, the number of patients in the worst case was reduced by 20%. The peak of the outbreak was delayed by 10 days until 17 April.

The number of new coronary pneumonia patients admitted to the Washington University School of Medicine hospital began to decline. Over the weekend, the number of hospital admissions in four hospital districts fell from 75 a few days earlier to just over 60. The policy of keeping a safe social distance seems to be working, Brandenburg said.

She knew in her heart that things could get worse at any time, ready for a bigger test. For the first time, Brandenburg says, she came up with the idea that Seattle might no longer be the next Spain, the next Italy and the next New York.

Other hospital administrators and local public health officials should be aware of the collaboration between the Washington University School of Medicine Hospital and the Murray team. Other U.S. medical facilities began writing to Murray’s team to ask for their help. As more medical institutions seek help, the Murray team decided last week to release their research, which predicts the inventory of beds, ICUs and ventilators in states in the coming months.

The new coronavirus spreads at different rates in different regions and is influenced by a number of factors, including population density, transmission patterns, and whether people strictly adhere to the rules of maintaining safe social distances. Mr Murray hopes that policymakers everywhere can use models to better understand the spread of the virus, “and we want to help them understand when the worst week will occur and how to respond”. The Murray team will update the model every Monday to absorb the latest data on deaths and consider the impact of a national policy of maintaining safe social distance. It’s too early to tell whether Washington state’s fight against the epidemic will be successful, but for now at least it’s doing well.

According to the University of Washington’s Institute for Health Indicators and Assessment, 41 states require more beds than they currently need, and 12 states need to increase the number of ICU beds by at least 50 percent. The model also predicts that the shortage of beds will be one cause of 81,000 new deaths from pneumonia in the United States, with daily deaths peaking in mid-April.

Even such predictions may be loose. Several epidemiologists have pointed out on Twitter that one of the assumptions of the Murray model is that states that have not yet imposed a ban will be enforced next week, and that it will be as strict as Wuhan, China, but many public health experts suspect they can do so. The reality is that enough states, mainly conservative-dominated and still few, are resisting strict bans. Even before the new corona outbreak, scientists failed to get policymakers to heed their warnings.