It’s a nightmare.
One spring that saw more than 7,000,000 cases and ૨ 2,70 deaths per day in the U.S. coronavirus toll, followed by a summer in which daily cases rose to over 1,000,000 and daily deaths recurred. Clears 1,400, appearing for the moment that the entire route to the U.S. The epidemic is moving in the right direction. On average, new daily cases return to 36,000. Daily deaths have dropped to about 700.
Yet experts have long feared that cold weather, increased indoor activity, school resumes and holiday gatherings could produce a fall wave with the possibility of dwarfing previous peaks – and now America’s leading COVID-19 modelers are presenting the same. .
So is it a vague time about the fall?
Probably not yet.
On Thursday, a team from the University of Washington’s Institute for Health Metrics and Evaluation (IHME) released its latest epidemic forecast. It was not encouraging. Earlier, IHME researchers now predicted that an additional 220,000 people would die in COVID-19 by the end of 2020, bringing the total number of deaths in the U.S. to 410,000 and more than doubling the current 190,000, which would hurt the U.S. : In less than half the time to reach the month of Khad.
Estimated daily infections, which IHME currently kills around 140,000, will nearly triple to 373,000 by mid-November. In early December, daily deaths will surpass the April surge.
And things could get even worse, the organization warns. If, for example, the United States continues to “gradually relax the order of social distance” and refuses to “re-impose” them even if the epidemic worsens – essentially following a natural “mob immunity” strategy in favor of President Trump’s new coronavirus adviser Does – then the IHME predicts that by the end of the year the U.S. Covid-1 deaths will bring the total death toll to 2,050,000, daily deaths to over 15,000 and daily infections to over 1.8 million.
According to the IHME, the only way to keep the death toll below 300,000 this year – 95 percent of Americans – is to start wearing a mask as soon as they leave home. Given that only 45 percent of Americans currently wear masks in public, that’s probably not going to happen.
So yes, IHME’s forecast is horrible. The question is whether it is reliable.
The truth is that no one – not even the top coronavirus experts – has a clue how the epidemic will develop over the next four months. It’s of course and that’s just awful. But that’s why presenting a certain number of deaths and infections four months in advance is sure to be misleading.
Just consider the number of variables in the game, especially summer gives way to fall.
Scientists are still unclear what role children play in transmission, and how that role changes with age. So expecting the effect of reopening the school is basically impossible.
We know that the virus spreads more easily in the air than previously thought, but what happens when the temperature drops and people inside the house retreat? Will they be less likely to accumulate as a result? Or they will be More Possibility of gathering instead – and in dangerous, uninitiated settings?
The same goes for the effect of temperature and humidity; Studies suggest that weather “affects the Covid-19 transmission,” like other coronaviruses. But researchers do not know how much.
And what about relieving fatigue? What if the inevitable temptation of a hearty Thanksgiving dinner or a warm Christmas morning persuades more Americans, once and for all, to stay 6 feet away from friends and family?
All of this and we haven’t even mentioned the potential effects of the vaccine, which is expected to reach at least a limited amount by the end of 2020 – or the ripple effects of the current cage outbreak, which could cause hot spots elsewhere. As students return home for the holidays.
The whole point of prediction, like IHME, is to share as many variables as possible in mathematical form. She is a model. And to be fair, IHME’s main projection makes some reasonable assumptions: once the cases start to climb, the slightly lower test will go up again; The use of that mask will continue; The distance will increase as the epidemic grows.
But IHME also makes another assumption that may not be so reasonable. According to the analysis According to Stanford University infectious disease and modeling expert Joshua Salman, “220K more death predictions by New Years” are “no behavior” and “no use of masks” – “nor is it even a test.”
Instead, according to Salomon, it is “seasonal” – IHME expects quid deaths to increase in autumn and winter, as well as pneumonia deaths to increase in autumn and winter.
“Se Tu is captured using weekly, state-specific critical statistics data on pneumonia mortality between 2013 and 2019,” said Sal Loman. Explains. “As far as I can tell, the estimated increase in daily deaths from 900 to 2900 is due to this seasonal effect.”
It is certainly possible that whatever seasonal factors stimulate our annual spikes in pneumonia mortality have the same effect on COVID-19 and as a result, we will double our national mortality rate by charging in the last month of the year. Dangerous ly rate of 36,000 new cases every day.
But again, no one has yet enough information to make a mathematical calculation of how the pneumonia and COVID-19 seasonality are connected – let alone whether they are perfectly connected. Putting that assumption on other assumptions to predict a very specific number of deaths (410,451) by a very specific date (Jan. 1) suggests there is a lot of certainty that there is indeed very little.
To understand how inaccurate the IHME prediction may be, it is worth comparing it with a different model. At Covid 19-Projection.com, Yang Gu, an independent data scientist trained in technology at the Massachusetts Institute of Technology, is using a data-driven approach with levels of artificial intelligence (read: less subjective assumptions) to predict the course of an epidemic. .
According to crime model Dell, daily deaths will drop to 400 by November 1 – less than a quarter of IHME’s November 1 launch of 1,825. Until then, according to Gu, U.S. The total number of Covid-1 deaths will be about 39,000 – about 35,000 less than IHME’s forecast.
Beyond November 1, the crime model stops guessing. Most other models do not appear for more than a week or two in the future. Gu used to project out three and a half months, but he has extended his deadline by two months because “Mobility can change in the fall and how unpredictable it is [they] Will change”
“Back in June,” he said Explained on Twitter, “The main driver [was] Reopening, and that’s something we can reasonably model. “But” no one knows what’s going to happen next November, so any model predicted from November 7 onwards needs to be taken with a grain of salt. ”
Incidentally, the crime model has been the most accurate to date – in fact, more accurate than the IHME, which fluctuates wildly and rarely matches reality.
Guy wrote on Twitter, “Predictability of certainty when very low,” can weaken people’s confidence in the scientific community. “
It can also scare people. To be sure, a little fear during a deadly global epidemic can cut a long line. The more Americans mask, maintain their distance, avoid indoor gatherings, isolate and test students at the first sign of an outbreak and do different holiday trips, the less likely they will die this fall from COVID-19.
But it is as early as possible to predict the impact of the measures, or the weather, or how those measures and the weather will interact.
For now, that should be enough to say Gu recently did it on TwitterKnowing that, as we know, thousands of additional Americans, and perhaps more than 100,000, lost their lives to COVID-19 this year. The hope here is that the unknown is no more deadly than that – and that fear of the unknown allows America to be more terrible than that.
_____
Read more from Yahoo News:
