How Much Worse the Coronavirus Could Get, in Charts

Play with a model of coronavirus in the U.S.

Comments: 221

  1. This is why I subscribe to The Times. Thank you for your detailed coverage and analysis. Having accurate information, that isn't filtered through a partisan lens (and focused on the stock market), is invaluable to us at this point in time.

  2. @Dan I fully agree. This information is much better than currently available on line in the UK (I am a British citizen living in France.) I also commend the latest measures by the French government to close all schools and educational places, from infant creches to universities "until further notice".

  3. @Dan This is not accurate it just confirms your believe.

  4. @Dan You also need governments that will seek our and act on information provided by scientists and doctors. This means Democrats, and Republicans who are not drinking Trump's Kool-aid. Considering Christie's handling of Hudson River rail tunnel replacement, how would he have balanced the welfare of New Jersey citizens against his political ambitions in the Republican Party?

  5. Awesome! In addition, delaying infections and flattening the curve give us time to develop, manufacture and distribute drugs which reduce the severity of the illness (not just vaccines which experts think will take 6-18+ months). Also, flattening the curve helps non-hospital logistics. For example, instead of huge sudden losses in truck drivers, food service workers etc we can maintain a steady flow of food and other essentials. Finally, flattening the curve reduces panic, which needlessly stresses everything from toilet paper to traffic to staffing to guns. You can't shoot a virus.

  6. Great article and the model helps to explain what is going on. We have all been drafted to fight this war, and we should have that mentality. Stay home as much as reasonable possible. Work from home if you can, so that there are less people around to infect each other. Clean your hands-a lot. Go out just once a day, if needed, to get supplies. It’s gonna be a long wait and we have to be patient and buy enough time to slow the epidemic.

  7. @Louis I have limited stamina and no car. Sometimes I need to go out come back to my apartment, take a nap, and run the other errands later. I also have a galley kitchen with only a mini fridge. I may have to stock up 2x a week for 3or 4 days rather than purchase enough for 2 weeks with limited fridge space.

  8. @Louis People who work in grocery stores where they come into contact w/ hundreds of people a day are likely to become infected. Does anyone have any idea how we can protect those people? They can't wash their hands between customers--the skin of their hands eventually will be injured by the process. Wear and change gloves between customers? Disinfect check out lanes between customers? Does any have any ideas?

  9. @JJ Soap & water or purell at entrance to every store. You must wash your hands before going in (enforced). Space all customers 3feet apart at checkout lines (put tape on the floor). Purell or soap & water at check-out. Gloves &/or masks for folks on tills. We can beat this--we just need to be clever & quick about adapting. Our enemy is simple & can't think for itself.

  10. I hope the Times will give this excellent piece more prominent play. It should be read by as many people as possible.

  11. This is scientifically correct. Also, the models do not require panic buying of staple items. Supermarkets and essential stores can stay open. The key is distancing from other people as a means to avoid aerosol-borne transmission. One should also note that we have more ICU beds in the US than seem the case by a simple tally. Remember that most hospitals have operating rooms, which will not be used as such, aside from emergency surgery, during a pandemic. OR's have the equipment, configuration, ventilators, and monitoring equipment equivalent to ICU's. Pre- and post-acute care (PACU) areas can also be converted to the equivalent of ICU space. The limitation is ventilators and staff.

  12. So why do we not have mass testing already? It seems like mass testing, combined with banning large gatherings and cruises (one of the largest gatherings of them all!), and strictly limiting air travel could dramatically flatten the curve. With the time we buy, we could be expanding our supply of ventilators and other medical equipment by a WWII style shift in domestic factories, which could operate 24/7, supplying jobs to those who need them.

  13. @Rose What? Didn't you hear President Trump announce several times that "anyone who wants a test gets a test." ? It would help immensely if our President, who has actual facts at his disposal, would speak accurately about this health crisis. Of course, it also would have helped if Trump hadn't fired our ENTIRE pandemic response team, including the White House management infrastructure in 2018.

  14. @Rose You can’t treat what you don’t know, and we can’t know the extent of the problem without testing. It is criminal that we are so far behind other countries both in timing and in availability of test kits. That is the real scandal of our generation

  15. @downeast60 The current admin DOES give us the chance to study our bizarre American version of autocratic politics--the transparently fearful egotism, the incomprehensible circles of complicity, the incompetence! This then is an opportunity to build alternative systems for level headed strategic planning and operations. Thank you NYTimes for your excellent coverage. Thank you commenters for reminding me of our collective sanity. And thanks could be due, I'd say, if we deployed them, to a critical mass of underutilized professionals --how can we hire back members of Trump's fired pandemic response team?; I myself, as adjunct in the higher ed labor force, am ready to be a soldier for strategic thinking. And then there's the potential army of autodidacts--I googled "aerosol transmission" and the first link up: https://bmcinfectdis.biomedcentral.com/articles/10.1186/s12879-019-3707-y That to me is reason to stay calm and hopeful.

  16. Fascinating. Here in NYC, the closures have begun, and my sense of confidence that our elected officials are doing the right thing has grown stronger. There is, though, the subject of the public school system. Mr. Mayor, is it time to close them?

  17. @dc In the Boston area, schools are closing. My kids will be home for two weeks. We have not a single known infection in our community of 115,000, but it is important that we reduce our public activity to flatten the curve.

  18. It is increasingly clear that we have a president who only looks one step ahead, and only acts out of selfishness regarding his re-election concerns. I’m not sure how you model for that, but it’s certainly not very reassuring for a major crises. Frankly he can’t lose soon enough.

  19. @Thomas Wright - there is, however, 8 months to go before the election. A lot can change, and by that time events may spin the narrative in such a way that many will vote for his victory.

  20. @Thomas Wright who would want to be president of a country where the market is wrecked and 250 million people are grieving?

  21. @Thomas Wright play with the model and set the bar all the way to the right.

  22. One of the aspects cooked in to the model is the rate or serious infection and resulting mortality rates. However, sciencenews.org's summary of the "Diamond Princess" cruise line experience, probably the closest thing to a controlled model we have, suggests a much lower rate than what's apparently in the NYT model. I understand that's arguable, and we need to understand how to compare that with, say, Italy, but the presentation of the NYT model reinforces certain conceptions in readers about what ICU and mortality rates are going to be and that affecting transmission rates is the only thing variable. Transmission rates may be the only thing in our *control*, through cancellations, hygiene practices etc, but I think the article tends to foster a bleaker and perhaps more frightening picture of what's coming down the pike than we really know at this point. I wish that'd be more thoroughly highlighted. I'm not saying things are rosy and that this is not a serious health care challenge, but we really do want people to keep things in perspective, to understand that there's aspects of this illness and its transmission and outcomes that we understand well and others that we don't and are still learning. The modeling exercise makes the article more fun, but I do believe it also lends a certainty to outcomes that are still not fully understood.

  23. I read an more than one article that stated the 2017-2018 flu season broke all records and that the vaccinations developed for that year were highly ineffective. 60,000 died in the US with the majority being in the 65+ range. I don't remember this being front page news back then. It's been implied that there is some control over the flu, that is until there isn't. It seems to me the only thing that's certain here is uncertainty. I'm not implying that the situation is not serious but more facts are needed. While these models can be helpful, there is a degree of conjecture that I find worrisome. Blasting the public with worst case scenarios and inducing panic is the last thing we need. Calmly stated, good orderly direction is a must.

  24. @Rob Actually the virus is mutating as we write these words. The Hunan virus exploded in Italy because it became a stronger version (Source: a friend who is a Dean of an Italian Medical School & working in hospitals in Northern Italy). Management and containment are crucial because things could go from worse to catastrophic by July.

  25. @Rob -- "I think the article tends to foster a bleaker and perhaps more frightening picture of what's coming down the pike than we really know at this point." So here's the ethical dilemma people in positions of leadership are currently facing: Human beings have a natural propensity to disregard and minimize depictions of upcoming realities which they are not experiencing personally. I see people all over my city not observing sanitary protocols. In 16 weeks, a bunch of us may all be dead. How worthwhile is it to overestimate the danger now and scare people a little into practicing social isolation and sanitation properly in order to save people from excess deaths 16 weeks later. You be the judge. But that's the dilemma. It's already too late to prevent the pandemic, because we didn't act 8 weeks ago (like Taiwan). How much longer should we wait to act?

  26. Any theories out there to explain Italy's nearly equal death and recovery numbers? It must take longer to declare someone recovered than it does for a victim of the disease to die, so the ratio will eventually go down some, but Italy's numbers have been running consistently from day to day, and we're beginning to see similar numbers in other western countries.

  27. I wish I shared your optimism expressed in the last line. The right is blaming the left for spreading a hoax. If the tech companies could could preempt Fox with the doctor in San Mateo's communiques, maybe there would be chance of taking rational action.

  28. As an educator who teaches systems dynamics modeling in my courses I would *love* some technical details or a reference for the model employed here. I have used similar models in my course and I would love to use (steal) this one.

  29. @Paul Fisher me too! This was already going to be Monday’s first online lesson.

  30. @Paul Fisher Robert Schiller's book, Narrative Economics, includes an appendix that describes the mathematics of epidemiology.

  31. Our public health criteria for testing is fundamentally flawed and needs to change. Currently if you have not been exposed to someone with the virus you will not be tested even if you have symptoms. Given the fact that less than 10,000 Americans have been tested to date few people will know if they have been exposed therefore few will be tested. To fix this we need to test everyone who has symptoms period. We need test kits that work period. Until this intervention is implemented we will fail. Not sure why South Korea can test 10,000 people everyday and we have had nearly 3 months and can't seem to make it happen -- maybe we need people in charge of this who can do the job right.

  32. @Julie We still aren't testing because Trump doesn't want "the numbers".

  33. @Julie It seems like with aggressive widespread testing starting today, we could keep the death rate down to about what H1N1 was, but without that, (just moderate intervention) it could be hundreds of thousands of deaths. We really need to test.

  34. What I haven't seen in any of the epidemiological models are estimates of how long the interventions and social distancing need to go on. Right now there are sporadic approaches with some schools closing for one to several weeks, and others for the rest of the year. With models of U.S. infection peaking in July, does a couple of weeks' closure accomplish anything? Also the prediction that 1/3 of the population will eventually contract the virus is not very encouraging. It would seem from that that distancing should be a long term thing over the next year.

  35. @Robbi The length of time for social distancing measures is inherently temporally correlated with data about the rate of infection, and outcome, and geographical data. In other words, a clearer picture will probably emerge over time. However, what is critical for any type accuracy in both model prediction, and public health recommendations is that we have testing procedures and widespread availability in order to predict, and behave according to data and information, and not from hunches, conjecture, or faith.

  36. Thankfully, leaders in sports, education & entertainment are heeding these warnings and executing temporary shutdowns. Disney has closed its parks, NBA & NHL have shut down for a while, and the governor of my state has ordered all K-12 schools closed until the end of March. The problem is, the leaders of these entities cannot do what can only be done by the US president: primarily, offer the leadership, psychological assurance and genuine calls for unity that this crisis demands. Even if Trump were to suddenly metamorphose into the sort of leader who guided us past the attacks of September 11 or the financial crisis of 2008, which is extraordinarily unlikely, too many opportunities have already been missed. What we're seeing now is too little, too late, and of course constant, major miscalculations in regard to focus. While thinking, caring Americans are concerned with the safety of the American public, Trump is concerned with how this makes him look and how it will affect his reelection chances. A major portion of the electorate has been riding this Trump carnival train for a while now, and they bear responsibility for the major mishandling of this crisis. Unfortunately, we will all end up paying the price.

  37. @TLMischler TL, actually all Mich schools are closed until April 6.

  38. Thank you Nicholas Kristof and the New York Times. The information you are gathering and presenting is clear and accurate, with uncertainty acknowledged. It is a life raft thrown into dark times. I have used information read in these pages to successfully argue for the cancellation of events within the control of groups I work with and to inform decisions about social distancing by my family. We are all on the front lines. Flatten the curve!

  39. Very educational piece. It’s important for the public to be able to visually capture the potential impact of the pandemic and the consequences of different levels of containment. Please note, however: In aggregate the age of the US population is little different from than that of China. The median age of the US population is 38.1 yrs and that of China is 37.4 yrs. European populations are older. The reason for the relatively young US population (compared with other developed countries) is - you guessed it - immigration and that’s good.

  40. Your model assumes a 1% mortality rate. This 3/12/20 article in The Lancet, Infectious Diseases, suggests the # of confirmed Covid19 cases, or the denominator used in mortality calculations, is artificially low, and thus the current estimates of mortality are artificially low. This study suggests that a more accurate rendering of the mortality rate outside China is 15.2% https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30195-X/fulltext

  41. @Alex “the denominator used in mortality calculations, is artificially low, and thus the current estimates of mortality are artificially low” That’s not what the article says, and that’s not how math works. If the denominator is artificially low, then the real number is higher. A higher denominator gives a lower mortality rate. 1/10 vs 1/100. See how that works. In the article they actually lower the denominator by comparing only cases that were confirmed 14 days before the death. Which makes sense for the death rate to be higher in the beginning, as testing would not be common. Source: math

  42. @Mark Yes, but the model estimates mortality rates above 5%! That's the point.

  43. @Mark Another way to put it: The point of the lancet article is that dividing “people who died today” by “people who are sick today” is mixing apples and oranges. The thing we really want to know is more like “people who died today” divided by “people who got sick at the same time as the people who died”. Otherwise you are in fact dividing by a denominator that is too high, because it includes recently infected people who haven’t had a “chance” to die yet. But probably poor Alex just mistyped. He did link an important article.

  44. There are dogs that have been trained to detect the odor of some kinds of cancer. I wonder if bloodhounds or other especially odor-sensitive breeds could be trained to detect the virus, given that it spreads via aerosol-borne transmission. While the virus may not have an odor, it is at least possible that the aerosol from an infected person would typically have saliva that does have a unique odor. Separately, it would be of value to mass test several sharply defined geographical areas simultaneously. Then compare subsequent doctor visits, hospitalizations and deaths 30 and 60 days later to the original percentages of positive test results. That data would lead to development of the model parameters unique to this virus. DPD

  45. The WHO estimates 60% of the world population will become infected. In the US with 380 million, that means 220 million will become infected. With a 3.5% mortality rate over 7 million will die. The WHO also estimates that 20% of the infected will be severe cases requiring hospitalization or other health care treatments; that means 44 million people will need help. With 7 million deceased, 37 million are left to the current health care system to help. We have 1 million beds in hospitals and at any point 700 thousand are occupied. Do the math, the health care system will not be able to keep up. That is why China built hospitals in 10 days. Health care workers will get sick too, so who will staff the hospitals just when the 37 million need help? The answer is to flatten the curve and impose draconian policies and force people to stay home. For how long? Who knows, but if we don’t do something quickly, the curve will be too great and even more people will die.

  46. @Ski bum Mortality rate for those those under 40 is approximately 0%, so you have to adjust to the population age distribution when calculating the number. the mortality rate is very high now because we are only counting the people that show up at the emergency room with severe symptoms. If you calculated the mortality rate for the flu based on that same criteria, you'd have something like 10-15%

  47. @Ski bum The US population is not 380 million, it's ~330 million. Rounding out, 60% of that is 200 million.

  48. @Ski bum "In the US with 380 million, that means 220 million will become infected." The current US population is about 330 million.

  49. What strikes me about the graph(s) is that it shows the peak happening around July. Most of the more aggressive actions, e.g., banning gatherings of 1000+ here in Illinois, are currently slated for 30 days. While it could be extended, it seems that timing will be important not only in initiating actions, but also in maintaining them, perhaps for longer than most of us currently imagine. That is also difficult as folks who are not ill will get restive as the time with "nothing to do" begins to drag on. It will take strong, forceful leadership to keep whole communities cooperating over a longer term. Then, too, if the virus takes a hiatus until fall it will be challenging (emotionally as well as financially) to gear up the public health measures for round two.

  50. There is a chance a treatment can be found utilizing new biologic-ACE modulating drugs to reduce the viability of the angiotensin conversion enzyme that the COVID19 virus invades. While not preventative, it could save those in the most danger.

  51. Excellent, if sobering charts. But I have one big question after reading this: the various models show peak infections in the July-November range, depending on the aggressiveness of the interventions. Should we take this at face value; i.e., things will almost certainly get worse through at least July before they get better under *any* scenario? Second question: Has the current administration (and members of congress, and our governors, and mayors) seen these charts? If in doubt, it would be a great public service if the Times could email this article to *every* elected official it has knowledge of, and *follow up* with their staff to confirm it has been read--and understood. This is actionable knowledge. The goal must be to flatten the curve. There is still time, but time is of the essence.

  52. Ok, I get worst case...but can we also project a best case? Perhaps a bit of optimism?

  53. @sadjoepafan I think the assumptions underlying the model presented in this article could actually be called a "best case" scenario because the case fatality rate is estimated at 1 percent. On the basis of actual current data, estimates of the case fatality rate are well over 3 percent. The lower case fatality rate estimate in this model is plausible because the current data are probably skewed by a failure to detect many mild cases, but we cannot be sure that it is as low as 1 percent.

  54. @R. Rodgers They assume a very high transmission rate and do not test their hypothesis if the transmission rate is more like the flu. in fact, it could be much lower than the flu (which is highly contagious). If the transmission rate is half what they assume, the spread will be much slower and much easier to contain.

  55. @R. Rodgers You are a victim of the poor reporting on this subject. If you dig deep into credible literature you will see scientists acknowledge that the death rate is likely much lower than current reports suggest. It is quite possible that the majority of cases in China were never identified and resolved on their own. How would a person with a cold go and get a test for a virus that no one knew existed? Old people with health conditions are certainly at risk, but look to South Korean data for a better estimate of the impact this has on healthy people. We have flu seasons where close to 50,000 people die. Do you get a flu shot? Do you avoid the public from November to March?

  56. Excellent scientific analysis. Now if this could be coupled with a model of the Stock Market, Trump might pay attention! A million deaths are not good for the economy.

  57. @AKJersey Or, his chances of being reelected.

  58. Great article. Remember we are still hoping that warmer weather will also flatten the curve and buy us some time to prepare for the fall rebound of the virus.

  59. @Tom German Virologists have found that the virus is not temperature sensitive. In fact it loves temperatures of ca. 27 Celsius the most

  60. @Tom Hmmm....this didn't seem to work for Australia, New Zealand, Brazil or South Africa. However, maybe the warmer weather has kept their Confirmed, Infected and Death Rates low(ish).

  61. China, with a much larger population and less time to prepare, appears to have peaked with under 100,000 infections. And your best case scenario for the US is 3 million? No. Just, no. The idea that the United States of America is prepared to tolerate an outcome thirty times worse than in China is unacceptable. How do we defend our system against China's if it leads to such a disproportionate outcome?

  62. @Leonard Flier, Ascertainment of the number of total cases is tough, I would think. I don’t know if there has been a random sample in non sick individuals from China to estimate the incidence of asymptomatic infection.

  63. @Leonard Flier after the smoke clears, we will collectively need to re-assess our reliance on Chinese-made products and supply chains. The risks they pose to the health of our society and industry have been glaringly revealed by this situation.

  64. @Leonard Flier China had effective testing and quarrenting measures in place rather quickly. We are not mobilizing at the same speed, so it is predicted that the disease will be worse here.

  65. What's missing here to me is a note to say that to keep a sustained peak under or near the capacity for a while, you have to ensure your measures are not too restrictive. In other words, you still need people to keep passing the virus on. If not, we'll experience a resurgence in cases and have periodic-like bumps in the curve (and the resumption of damaging measures).

  66. Excellent analysis that helps bring the sometimes confusing but vital concept of curve flattening into perspective. One factor that it would have been nice to explicitly consider is the impact of aggressive testing as a tool for contact tracking and containment. In much of the country, it is not too late to contain local outbreaks. The key to doing so, however, is to test widely and early. Rather than waiting until a patient has fallen so ill as to be hospitalized, every patient showing even the earliest sights of CoVid-19 should be tested, monitored, and quarantined while contact tracking begins. By the time the patient is very ill, he/she may have passed the illness to countless others in the community. Aggressive testing and contact tracking might not lower the total number of cases, but it can flatten the curve significantly, particularly in America's rural heartland where relatively few cases have been reported thus far and where medical facilities to deal with a spike in cases are in short supply.

  67. dear gentlemen, as i am Italian and i do business with China, i am under pressure for virus and I have gathered many information on this virus problem and , therefore, i can underline that usually spring season ends the impact of flu, generally speaking, and that will help the temporary solution. BUT, WHAT ABOUT NEXT WINTER FLU SEASON WHEN THE VIRUS, AS ALL VIRUS, COULD COME AGAIN WITH DIFFERENT STRUCTURE AND STARTING FROM OCTOBER, WHEN THE IMPACT WILL HAVE SO MANY MONTHS FOR HIS SPREADING?

  68. @ROBERTO Dear "concittadino", spring season is here (and we never had winter in Northern Italy during 2020) and the virus is spreading like Australian fires, so dont bet on this temporary solution

  69. Very insightful. This helps the American people (read Republicans - lawmakers and voters alike) understand that even in the best of scenarios envisioned, 10k's of Americans may die; and in worst case scenarios, an unthinkable 1M dead. This event is unlike any we have seen in a lifetime, and must be treated that way -- with a robust, public health response that extensively tests, quarantines, and treats -- to keep the American people as safe as possible. We are NOT on that path now.

  70. As I understand it most, if not all, countries are claiming to be following the "best medical/scientific advice". However, that advice seems to vary from country to country. Some countries are pursuing a much more aggressive response than others. Some countries are trying to be more proactive than others. The UK government's approach is to slow the spread of the virus with the peak infection rate being as much as 3 months away. However, no closing of universities or schools, museums & libraries, no closing of cinemas & theatres, no closing of restaurants, pubs or clubs & no cancelations of sporting events & care homes remaining open to visitors If your approach is to slow the spread of the virus, why provide it with opportunities ideal for infecting a large number of the population? I'm in a high risk group (age, asthma, scarring on my lungs, compromised immune system). The advice to people like me is to avoid crowds & wash my hands repeatedly. I'm fortunate that I have family who make it possible for me to remain comparatively safe. That doesn't sit easily for me as I dread my wife, children & sister becoming infected. Finally, what of those who don't enjoy the support I have?

  71. @VambomadeSAHB Thanks for your info from Scotland. I have to decide to pay the last installment on a much anticipated trip to Orkney in July. Having the curve move to July is not good news for me at 73. I will hope I can postpone a year.

  72. Thanks for this excellent column. It is especially welcome in this atmosphere of fear and uncertainty. If we all comply with social distancing, more of us will stay healthy and those of us who get the virus willl have a greater chance of survival if we need medical care. Alas, iIf only more tests had been available, earlier!

  73. I appreciate much of what Mr. Kristoff and Mr. Thompson are trying to do. They understand the problems with attempting too much precision. Some of their discussion might help to calm nerves, but no amount of science, at this stage, will stem some of the coronavirus panic that is occurring. There is little doubt that this administration's initial tendency to minimize problems and withhold information made matters worse: https://www.foreignaffairs.com/articles/2020-03-10/science-cant-save-us-coronavirus-panic?utm_medium=newsletters&utm_source=twofa&utm_campaign=Science%20Can%E2%80%99t%20Save%20Us%20From%20Coronavirus%20Panic%20&utm_content=20200313&utm_term=FA%20This%20Week%20-%20112017 .

  74. We the people still are in the dark. Though there is only early data, the information transmitted seems to be woefully incomplete. Your graphics, if a true representation of the spread, is more compelling than any Administration depiction. First we know that cases are increasing in the US, but in China, they are falling, and in other places they are not increasing geometrically. Flu also moves quickly and cases increase rapidly, but we are not given comparisons. The ID and Epidemiologists are in offices at NIH and CDC, and for this brief moment they get to be the stars, but the science at this point seems to be modeling worst case scenarios that may not come about. For instance, in China, perhaps they should have a million cases, and the cases would be going up according to your graph, but they are reporting cases going down, so the data is contradictory. I don't disagree with the current strategies, but I would like to hear CDC epidemiologists give modelling updates frequently and not just tell us what to do, but keep explaining why.

  75. @William Trainor The difficulty in both the US and China is a large population without access to affordable medical care; anyone working for low wages will tend to go to work even if sick. I suspect the number of US cases is in the tens of thousands but since there is no capability to determine if a sick person has the new virus no one can be sure. I also assume that many recent deaths have been attributed to "natural causes" rather than the new virus. The current government seems very happy to keep it all secret and pretend anyone reacting (like the NBA, cancelling games) is panicking. Whereas the real centre of panic seems to be in the White House.

  76. Our country too often displays a bizarre reluctance to use its great wealth for public goods. Mr. Kristof and Thompson show us how rapid deployment of testing facilities, ventilators, and support for quarantines will save many thousands of lives and allow economic recovery much sooner. Dithering and refusal to spend for the public good is the path to disaster. Thank you Mr. Kristof and Thompson.

  77. @Charles Knight Using our power for the public good and to help people is socialism and the surest way to hurt our country. The only good Socialism is Corporate Socialism, where the rich benefit, and those benefits Tickle Down.

  78. @gratis Watching CNBC, the host actually said it would (and I'm paraphrasing here, but accurately) be better if everybody got it and we were then done with it. Then markets would recover faster. In other words, better to kill off a million or two people rather than have the markets suffer a long drawn out period of uncertainty. He later was forced to apologize but it was striking the thinking of these types. Neither of his two co-host pushed back or disagreed, but that typical of corporate media, especially the business shows.

  79. @gratis... Dead man walking.

  80. I am at a loss to understand why the Mayor of NYC has not closed the public schools. He must know that this will inevitably happen. He must understand the basic point of this article. Because of his inaction, more people will die in NYC. More people than those victimized by police misconduct. More people those killed in motor vehicle accidents. What is he waiting for?

  81. @David A. The issue is the mayor has not figured out a plan for the homeless kids in particular, and those who count on school for meals and even clean clothes. Many working parents who cannot take time off to take care of kids, or may lose their etc.

  82. @Oh My Time is of the essence here. Every little helps bend the curve. Telling people who *can* keep their kids home from school to do it *now* reduces the risk of cross-infection (which is the name of the game for the next few months). It's a no-brainer. It reduces population density & risk of cross-infection. It also keeps teachers safer. Those that have to come to school for food/clothing can still do so until we figure out how to supply stuff for individual families, wherever they are. We need to move fast on this.

  83. @David A. Not single child has died from the virus. Not a single adult under the age of 50 (parents of school aged children). How many degrees of separation do you go? When does the cure become worse than the disease?

  84. For those who recover we have no testing for antibodies or insight into how long they might last. This is a second level of testing that would be useful as those who exhibit some immunity would curb the infection but also predict a second wave if the antibodies wear off. I would like to see the impact of antibodies on the flattening of the curve.

  85. @poslug The question is understandable but the knowledge as to persistent community immunity is a long way off: https://onlinelibrary.wiley.com/doi/abs/10.1002/jmv.25685 ("Coronavirus infections and immune responses," 2020) As for use of antibodies as a therapeutic tool to treat the virus, that is likely even further away, both in terms of discovery/testing and affordability: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5627241/ ("Antibody therapies for the prevention and treatment of viral infections," 2017).

  86. Under-reporting is the issue. Many people have no access to health care or even if they do, they can't get an appointment. Think Naples Florida in season when the health care system is normally maxed out without the Corona impact. A one month wait to see a person's primary care physician is normal. How about the spare capacity of hospital beds? Doesn't exist. As for urgent care walk-in clinics, they will be hot spots for catching the virus. Cleanliness is not a hallmark of the U.S. Look at many restaurant toilets and sinks and those along interstate highways. This illness will run it's course and it's doubtful that politicians will learn form it. Hopefully a vaccine will be developed that will minimize the next recurrence of it.

  87. The mortality rate in this model is far above those known for every country to date and those are overestimates for a variety of reasons.

  88. @Jim G. As of this writing, Italy has 15,113 cases and 1,016 deaths, or 6.7% mortality rate. Source: (https://www.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6). The practice of medicine in northern Italy is more like Switzerland than China where the case-fatality rate is ~3% according to the same data. So much of the early experience around this epidemic suggests that mortality rate is very much about the ability to prepare for the surge.

  89. @Jim G. Look at Diamond Princes: 696 cases, 7 deaths, 364 active cases with 32 in serious or critical condition, 325 recovered so far. The death rate is already 1% (hope it stays that way). Of course, demographics differ from the general population.

  90. @AL Another fancy graphic, but isn't it odd that your reference leaves out South Korea, the country that has one of the most reliable figures for number of infections. The more accurate this number the more meaningful the mortality rate calculation. For South Korea this is well under 1% with the elderly making up the vast majority of deaths.

  91. Given that many universities and high schools are being closed and instructors are scrambling to provide courses on line, a short module available from CDC should be developed on the material presented in this article. People would understand how the decisions that affect our lives should be made via an all too real case study. A few university students with a background in epidemiology could collaborate to develop the interactive module (with input from Kristoff and Thompson and other faculty with relevant expertise) in the next few days. It is a teachable moment for all of the students who are out of their classrooms and we would have a more informed public. When the next outbreak occurs, CDC would be adequately supported so fewer people would be infected with less serious disease and less economic impact.

  92. Very well done. So illustrative, I admire the analysis and graphics. What a great team effort. I also appreciate the frank assessment. Testing was indeed completely bungled (as it was in many European countries... the US is not alone). Hopefully aggressive measures will be taken, including by the public, and we will flatten the curve. Now, thanks to you, we can really visualise that.

  93. My second post on this-we need really good leadership at this time-local, county, statewide, national. First aim is to flatten the curve-but then what? How do we not permanently ruin people’s livelihoods and hollow out towns and cities of businesses? We have hopefully 12-18 months to get a vaccine or better treatments in place. That should be the expectation right now for everyone. Though China has flattened their curve, how well are they sustaining it and how can we do it for America? We need ubiquitous testing so that we know where the hot spots are and quarantine quickly. So where the disease is more controlled or not at epidemic levels, we can go about our daily lives (with smaller gatherings) and not ruin the local economies. On a separate note, do we have to do year-round schooling for a couple of years to make up for lost school time? Can we find ways to reduce school density so that staff, students and parents are not so stressed to be in class?

  94. I grew up in the forties and fifties when hospital floors were always washed up by older women who were down on their hands and knees behind tin buckets utilizing scrub brushes and bars of hard yellow Octagon soap made of lye. My own mother was devoted to this soap, and I personally knew a few of these women who were, to put it mildly, fanatically good hospital-floor-washer-uppers. I believe this soap is still commercially available, but I greatly regret to say that these good women are not. I made a personal vow to myself a long time ago to stay out of hospitals unless and until some adequate substitute is found for these women, and it again becomes safe to eat off of hospital floors, which currently and for the last 30 years it has not been. I leave it to Vice President Pence, who recently was placed in charge of this matter, to give urgent attention to it before our hospitals death rates skyrocket.

  95. @A. Stanton While I generally agree that we as a society are indebted to “older women” for many things, and I applaud you for noticing these good women while they scrubbed, we might be ready for a coed scrub team among some other more radical approaches to dealing with an unprecedented public health emergency. Informed leadership acting quickly comes to mind.

  96. @A. Stanton A small correction made "with" lye as an ingredient, not made from lye, which could cause serious burns.

  97. @Warren Courtney Thank you!

  98. This article again makes me fantasize that our President was named Nicholas Kristof. What I have not heard about are any plans to back up the clinicians who might be called upon to devote themselves to the care of seriously sick hospitalized patients. As a retired generalist who volunteers in a free clinic, I would be a logical doctor to take over the care of stable patients with chronic conditions if their usual specialist was unable to see outpatients. But, I am unaware of concrete plans to “preposition” and “activate” backup clinicians if the size of the coronavirus case load approaches a point of stressing our disease care (non)system. I looked yesterday at the websites of my state and local health department, my local medical society and the nearest Medical Reserve Corps group. Hopefully this is being worked on and potential backup clinicians will be asked to sign on. Stephen Rinsler, MD

  99. I generally am not a fan of Ohio’s Governor Mike DeWine, but he has done the right thing by proactively closing our schools and prohibiting gatherings and workplaces of more than 100, except for groceries and the like.

  100. @J. Stay well, Ohio.

  101. Question? If, as they likely are, the infection cases are MORE than we have currently reported because we are currently behind the curve on testing, how does that affect the model? Wouldn't it mean that the predicted peak is a lot closer to us than where your model has it in July?

  102. @Sharon Sheppard I’m also wondering about this: and then I’d like to know how fast cases can increase in a locality that doesn’t have an early head start from overseas travelers or (alas) a nursing home.

  103. Irresponsible. Even the most credible scientists are sharing "death rate" numbers that are based on such limited and flawed data that they are worthless. Most people now believe that 2 or 3 of every 100 people who get infected will die. Look at the country with perhaps the most comprehensive testing program, South Korea. Out of several thousand cases, how many people under 30 have died? Zero. How many people under 40? One. How many under 50? One. Even these numbers probably overstate the mortality rate because there will be many people who only experience mild cold symptoms and are never get tested. People die in America every day from preventable causes and we carry on without any concern. The greatest public health emergency we face is obesity and the corresponding complications, but this is nothing to people. The same goes for distracted driving. Many thousands of innocent, healthy people are maimed and killed every year but try to pass red light camera or automated traffic enforcement laws and watch people howl about the unfairness of it. This new virus is a serious issue but there is a lack of perspective in our response. And BTW, scientists who quote a 0.1% mortality rate for seasonal flu are using an educated guess that estimates the total number of infections (this brings the mortality rate down) and they are ignoring the fact that in less developed countries the flu mortality rate can be closer to 1%.

  104. @Teal I agree this is really bad data science: 1. The current case number in the US is likely much higher than the base line assumed in the model since we are testing very sparsely. In that regard the mortality rate is probably much lower. 2. The testing is also not statistical representative because it is not random but along people who show symptoms and their contacts. Severe cases are much over-represented in the data we have. 3. At least for the next weeks we will not be able to distinguish between an actual increase in cases and the results of improved testing (both will happen in parallel). 4. Current death numbers in China (3,500) don't seem support a best case scenario of 51,000 death in the US. 5. Sad that people are uncritical if well-meant bad science confirms their believes.

  105. @Falk Schuetzenmeister I would trust the numbers coming out of Italy as they evolve, although even there the quick action of the government to lock down northern Italy is likely a very different response than would be tolerated here by many

  106. @Teal the greatest threat to our future with the Coronavirus is people who think like you who pretend there is some conspiracy behind these projections or regular people like you know better. Give me a break!

  107. Visuals and sliders graphically depict the storyline. Can someone pass this on to our “very stable genius”?

  108. @Tabula Rasa Done. Tweet sent to POTUS with link to Times article.

  109. @Tabula Rasa You can't convince a person who doesn't want to listen, and frankly doesn't care about anyone else (save, perhaps, his oldest daughter).

  110. Trump's shoddy leadership of the feds coronavirus pandemic response has resulted in actions and messaging that fall far short, probably dangerously so, of what is needed to protect Americans and aggressively treat those who become seriously ill. America's response to the public health crisis is not wholly dependent on the response out of Washington, D.C. The states possess considerable power and messaging capability, too, especially if they work together. Politics must be minimized. The manufacture and distribution of coronavirus test kits, masks, and other protective gear; having experienced public health leadership, including virologists and epidemiologists, with the authority to manage resources, and equipping and supporting nurses, physicians, and other medical staff adequately, are priorities. America has a great many people qualified and dedicated to protecting our citizens and caring for those who need it. Stop focusing on those who should be leading the fight, but are not doing so. Let the professionals lead the fight. Let them do their jobs.

  111. I unfortunately work for an outpatient healthcare practice that has balked at rolling out updated measures to protect staff and patients, despite multiple pleas from myself and other doctors to update front desk screening questions and provide more protective equipment. Like Trump, my practice's administration can't seem to wrap its head around the concept of good prevention paying off dividends later, even if there is some short-term loss. If there are other healthcare and general workplace settings like mine that are dragging their feet, I worry that we will see the worst-case curves depicted in this article. As a physician, it is upsetting when my inbox is flooded with emails from various non-medical businesses and services explaining how they plan to limit exposure, yet I am a physician and am hearing crickets from the (non-physician) "leaders" of my practice.

  112. @anonymous you are not alone. it's happening in my city, too. it doesn't help that the official guideline are weak. look at the cdc guidelines for outpatient clinics. they say "explore telemedicine options" or "consider" rescheduling non-urgent appointments. it is all very wishy-washy and optional sounding. The only thing I can say is keep speaking up, or threaten to leave if they don't change their tune and get others to speak up with you (doctors have power only in numbers when going against administrators). I imagine there will be lots of locums jobs for telemedicine in the coming weeks and months.

  113. Really? China just made it through the coronavirus epidemic that it started with less than 4,000 deaths, and Nicholas Kristoff is leading us to believe that the best case scenario for the US is 324,000 deaths, although a million dead is also reasonable in his simulation. It's so absurd one has to wonder what the angle is here.

  114. @Chuck French China was able to lock down an entire city. We have not done that so it will continue to spread unabated. The best we can do as citizens is to lock ourselves down.

  115. @Chuck French I more or less agree with you about China. The thing is, can the US do anything on the order of magnitude of intervention that China did and in the same quick time frame. I don't see people realizing the need for extreme draconian measures like massive quarantines and so on for at least 2 months. Maybe the stock market crash will ring the bell, I don't know. In China, the leader says it and it happened. How much pain will the US need to experience before it will be acceptable to go that far. Also, I have a student in China and I spoke with him 3 days ago. Every movement is monitored now to make sure that sick people are not in the public. Still, all restaurants are closed as well as supermarkets according to him. I think it is necessary to go that far but I doubt it will happen. ON THE OTHER HAND, there is this article today in the NYT: https://www.nytimes.com/2020/03/13/opinion/coronavirus-best-response.html

  116. @Bill think of this comparison: China is able to build 1,000 bed hospital in 10 days to confine and treat patients. Does anyone believe we could do the same? My response is not a chance as permits, public hearings and federal approval would have to come first. Even though we have bungled testing thus far, all projections predict cases in the millions so perhaps we should get started!

  117. Sadly, these worst case scenarios are too optimistic. When the health-care system is overwhelmed, more people die. Take Hubei as comparison: 4.6% case fatality rate. Thus, I'm counting with 5 million deaths if uncontrolled. The US health care system is spotty… how worse can it get? dunno.

  118. @Volutes But in the province, the fatality rate is about 0.4%, so maybe...

  119. @Volutes IMHO the focus of the article is not so much about numbers here but on the dynamics of time with regard to interventions and there possible impact

  120. @OneView Maybe not. The epidemy was stopped in Wuhan. Uncontrolled in the US it would...

  121. Science and technology can not solve almost any problem if the people in charge don't understand or refuse to accept the issue in the first place and whom to solve it for. It's also the responsibility of the scientists, the real ones, not those phony guys with a PhD in STEM or a job with scientist designation, but without those 'core values' that make people aware of the society and the world they live in. Two basic qualities of a scientist must be 1) honesty to talk straight, and 2) the courage to stand up for truth and honesty, particularly when those "core values" are challenged. Yes, democracy is not about elections or majority opinion to decide or evaluate truth, but about what to do with that. But that would fail miserably if we fail to understand what is *Truth* in the first place. And it's the scientists that actually define what is truth. Philosophers and other original thinkers later can add to that. Forget about communism and other forms of autocracy, even democracy would not work if we fail to define *Truth*.

  122. Great analysis and graphics, but the peak occurs in July, and there's quite a bit of evidence this is a seasonal virus. Time will tell.

  123. @Robert Could you elaborate on this? From what I've heard virologists say, this is not the case.

  124. @Robert There is no evidence, yet, that this is the case. It's a possibility - not something to count on.

  125. Thiis is a good article. I'm an old geezer, but when I was much younger I taught courses on differential equations in a small university. The students found differential equations boring, so I tried to liven the course up by applying toy models to diseases and epidemics. As an example, I would point to "the Plague" by Albert Camus. The students weren't impressed. I was an unpopular lecturer. However, one implication of the models sometimes generated a modicum of interest. The model showed that if a ship of sailors arrived in the port of a fictional city, let's call it Oran, infected with bubonic plague, two scenarios would unfold. If the city was too highly populated the number of interactions between sick and susceptible would be high and there would be an explosion of cases. On the other hand, if the city were more sparsely populated and interactions occurred less frequently the plague would die out. I wonder if the same principle might hold for COVID-19. Perhaps those who live in farms outside Bison, Kansas, and go into town only once a week for animal feed are better off than those in New York City, whose look out their windows at apartments which stretch to infinity. In other words, the lethal nature of COVID-19 is just one more example of the downside of too much population growth. It started in the highly congested city of Wuhan, and will continue wiping out people around the world, but somewhat selectively, sparing the left-behind areas of the Midwest.

  126. Why should we see infection numbers different from Germany whose Chancellor predicted two thirds of the population would become infected? That would equal 200 million people contracting the corona virus here. With a 1% death rate, that would mean 2 million deaths. I would agree with your worst case scenario if we were taking aggressive action now. But given that we don't see Trump, Ivanka, AG Barr and other self-perceived GOP toughs who were in close contact with infected people self-isolating to protect others like the Canadian Prime Minister is doing because his wife tested positive and we also have the leading cable news and talk radio programs telling its automatons that it's a hoax perpetrated by the Dems or no more deadly than the flu, I think my numbers will turn out to be more accurate. They might even be half as many deaths as might occur in the U.S.

  127. As shown on the graphs very early intervention is absolutely essential. Everything possible to do must be done if that means closing schools, public gatherings, etc.

  128. We need more testing, many thousands every day. This should also include rapid point of care testing. Rapid testing can be antibody or antigen testing, not RT-PCR. Years ago, FDA approved rapid testing for HIV which you can by at pharmacies and test yourself. We need a much better handle on who has the disease and we have the technologies to do this. Hopefully, it will begin very soon. That a country like ours is the worst advanced nation at tracking this pandemic is the culmination of years of gross misinformation.

  129. The world has changed for all who are alive today. The freedoms that we knew to congregate and travel freely will be over for the remainder of the lives of all that are alive now. The virus and other viruses are here to stay. There is no way to eradicate them. They will likely gain strength and virulence. Other viruses will surface. The emotional scarring is developing now and will remain forever. Everybody will be afraid to engage in sports, travel and even work. Socialization is forever changed. Look at he TV shows without audiences. We don't realize how important these things are until we do not have them. Flesh and blood needs flesh and blood. Now they will be taboo. Sexual interactions will be heavily curtailed, even with spouses. The day has come for the apocalyptical naysayers in their bunkers. Life as we know has been dramatically diminished forever.

  130. @rich williams Yesterday a viral expert explained that this particular virus is not evolving, so he doesn't believe it'll become more deadly. And a vaccine is in development. People said things "changed forever" with HIV (a far deadlier virus), Pearl Harbor and 9/11. We adapted and got back to a recognizable "normal" after those catastrophes. We'll adapt and become complacent after Covid-19, too.

  131. @rich williams Seriously? - You are not, by any chance, running an online-shop for "Doomsday-paraphernalia" or trying to start a new church, are you? Of course, viruses will be around. They always have been. New ones will surface. So? We´ll deal with it. Either by vaccination, or by self-developed immunity in our bodies. Please yourself and dig your own hole to hide in, if you like. But stop scaring others. (and f your comment was meant as satire, I apologise I did not get it.)

  132. @Randy sure right about 9/11

  133. What we dont know 1. was this floating around 6-12 months before it was identified. most people had mild cases. now a sever strain has surfaced. which is more deadly 2. If exposed do you become immune or can you get it again? If there is immunity is it only temporary like if you just got it you can fight it off but if it dies off in the summer it can resurface and you dont have immunity. Or is it like Chicken Pox where you get it , survive, you have immunity. 3. My concern everyone goes in isolation , disease goes down. then people congregate again then it goes up again. 4. In a test will there be a difference between currently have it vs you had it and are now done. For people who are wondering. You dont know what the next outbreak is in so there is no way you can plan and store cures for unknowns.

  134. Thanks! A great tool for visual learners, and, perhaps, for the otherwise hard headed. Well done.

  135. I am a professor of medical genetics. It is my opinion that given the number of people infected with C-Virus grows exponentially across the Globe, it is likely that millions are already C-virus carriers. Professor Lipsitch a Harvard epidemiologist says that within a year, 40-70 % of people world wide will be infected, and emerging consensus among epidemiologists is that most likely outcome will be a new seasonal disease. So probably there is no escape everybody eventually will be infected. Efforts should be concentrated on protecting the most vulnerable-ages 70 & up & practice social distancing so not to overwhelm the health care system. Children <9 are asymptomatic to C-Virus and incidence of severe Covid-19 in young people < 50 is very low ~0.1% Thus, analyzing why children & young adults are protected should dramatically accelerate availability of antidots.

  136. @lieberma This is a more rational response relative to the "we're all going to die" hysteria that this article is promoting.

  137. The FDA has announced a new test that is 10 time faster than the old one.In the mean time the USA has ordered 300 million of the old test and gotten an unbelievably good price with delivery sometime this month or early next month. The new tests will be available in time for the next pandemic and we'll likely only need half as many.

  138. @kirk ?satire

  139. @kirk What is your source for this information?

  140. @Elizabeth /Leonie My bad. didn't mean to scare you. The first part about the ten times faster was news this morning. When they would be available was not clear.

  141. Insightful presentation of potential scenarios. I strongly encourage you to develop more features of this type. The plots presented are quantitative, however, and it would be helpful (and strengthen the presentation) to have more information on the nature of the modeling--an expanded Methodology section--with links to sites and papers describing more mathematical details of the modeling.

  142. As to whether it will abate with summer, cannot we look now to the Southern Hemisphere, where it’s late summer, as a gauge ?

  143. @Boyd Trump doesn't understand Northern and Southern Hemisphere. A stable genious like him can't be bothered.

  144. @Boyd Yes, and it's apparently not clear from the data. I think Kristof wrote that in response to another commenter. However, whoever it was said that comparing northern to southern China, it looks like there may be a decline in summer. As for why that happens with the regular flu, and colds, I suspect it's because people get more vitamin D in summer-- as without supplements, the main source of vitamin D is sunlight, and when the sun gets too low in the sky, the relevant rays are filtered out. Without sufficient vitamin D in the blood, the immune system is not functioning on all four cylinders. Sufficient D is 40-60ng/ml. In the darker months, I need 2500-3000 IU/day to get to that level. Rule of thumb: if your shadow is longer than you are, the D-catalyzing rays are being filtered out.

  145. This is scary, listening to the media one gets the impression that the worst will be in the next 2-3 weeks. If these charts are correct we are facing massive increases in infections over the next 3+ MONTHS!

  146. This is responsible reporting. This public health professional thanks you.

  147. We have never developed immunity to Coronaviruses (colds) before this one, because they tend to mutate. I would love to know if we have seen a lot of that happening with this virus in the genomic testing. If it is mutating, that does not portend well for reinfection and a second virus season and more beyond.

  148. @Office manager There was a news item from Politico this morning that a tour bus driver in Japan has been re-infected...it appears that reinfection and mutation is occurring already.

  149. This is very interesting and very helpful. I was quite surprised to see that regardless of the choice of when interventions start, this model is projecting the peak of cases in July. That is far later than I think many of us would have expected. It would be helpful for the authors to expound further on that. For many of us - managing home and work lives - that time line is actually very important.

  150. Excellent. A real public service.

  151. China has many millions more people than the US and the number of infected at the same time never exceeded 60000. Total deaths are around 3000. So the numbers in this article seem pretty high by comparison. I would like to see this model applied to China to see if it can predict those results.

  152. @danarlington China took drastic measures to prevent infection. Cities were locked down, transportation was halted. People were not allowed to leave their apartments - except perhaps for short trips to acquire food. Everybody was being monitored for fevers. They took extreme measures of "social distancing". If we don't want to see many number of infected people in US at the same time, similar social distancing measures have to be adopted. That is the point of this article.

  153. @danarlington China is a dictatorship that can demand people stay inside under threat of being shot or arrested. That will not work in the US.

  154. @danarlington Per The Daily podcast (Thursday March 12) China has recently been very aggressive setting up fever/testing stations all over and if one tests positive they are instantly quarantined. They have buildings and beds set up. 75-80 % of cases come from family clusters) thus China’s number of cases is decreasing. Maybe they realize this type of quarantine won’t happen in the US.

  155. Regarding the impact of warmer weather, an article in this newspaper this morning pointed out that the infection is currently active in summertime Australia, as well as in Singapore which is warm and muggy year round.

  156. @RMS That's true, but many cases in both countries have involved infections picked up elsewhere. It's not clear how much community transmission there is in each place, and there's some reason to hope (based partly on southern China vs. northern China) that warm weather may help.

  157. @Nicholas Kristof You can disregard my earlier response. I just read on and found my answer

  158. @RMS Trump doesn't understand climate change, much less the science of Northern and Southern Hemispheres... Donald Dear....when it's winter here, it's summer there. Got it? Confirm with your man of science, Pastor Pence.

  159. I know this is just one facet (and may not be important to some), but restaurants in New York are New Yorkers' "go to". I was just at a restaurant picking up a "to-go" order. I was stunned at the lack of response by management: Servers re-folding used napkins, picking up used glassware & flatware. Why is the restaurant industry/association not responding??? Servers should be wearing gloves, perhaps switch from glassware to paper cups. The restaurant business without this epidemic is difficult enough. We've already seen an implosion in retail here. Now, we're going to see shuddered restaurants, the "heartbeat" of what makes New York tick. How is the government (local, state, federal) going to respond????

  160. Yikes! From the graphs it looks like we'll all be staying home for the summer - maybe not even to attend local funerals. What's scary now is the continued lack of enough coronavirus test kits (not to mention a lack of smarts, compassion and truth telling in the Oval Office).

  161. @nestmaster The model doesn't count on warm weather, which may help; in that case there would be improvement in the summer and a worsening in the fall. No one knows. Funerals are a really big issue. There may be a lot of them, mostly of old people who have elderly friends and are themselves vulnerable.

  162. @Nicholas Kristof Precisely why Italy has banned large funerals.

  163. @Nicholas Kristof As difficult as it is to suffer a death in the family, the silver lining has always been the way family and friends take the opportunity to come together not only to honor those who have passed but to provide physical manifestations of love and support to their immediate survivors—showing up to the wake, the funeral, the repast; hugging one another; bringing food to the bereaved and promising to stay in touch. In many large and far-flung families this can be the only time they see one another. Then there's the necessity of going through all their belongings and distributing things of financial and/or sentimental value. With social distancing, the dangers of traveling in enclosed spaces, and a reluctance to stop at rest areas on long trips, I fear all of that will be lost. People will stay home and I'm sure that's wise for their biological health, but I worry about our psychological and emotional health and the social toll not being there for one another when we're not only losing our loved ones but losing them in the middle of a global pandemic and throwing away all their mementos will have on us as a society.

  164. It won’t be long before testing won’t matter, as too many will have the Covid 19 virus, but social distancing will matter greatly to avoid the peak happening too quickly and severely for hospitals to manage the patients within that 15-20 percent who gain more than a mild case. Let’s do all we can to flatten the curve, to care for the high risk and our medical professionals. We can slow the progression down and alleviate 2/3 if the death toll. We can ‘treat others the way we wish to be treated’. God have mercy on us, if we don’t.

  165. I'm startled that the predicted best-case scenario is 51,000 deaths. And it appears that if we begin aggressive intervention today vs modest intervention today (it seems to have already begun), the result is 269,000 fewer deaths. Thanks for this revealing tool.

  166. The transmission rate based on the experience of the two cruise ships put in quarantine seems much, much lower than projected here. Given the time, exposure and delay in separation, one might have expected infection rates nearing 100%... Assumptions matter and not adjusting the RO to demonstrate the impact on the projections is malpractice.

  167. This is inaccurate. The case fatality rate of documented infections may be around 1% or higher *of people testing positive*, but no country has 30% of their population testing positive for the virus. The highest percentage is Italy with 25/100000 = .025%. The fatality rate for documented positive cases only applies to people testing positive, not to estimates of the number of people be carrying the disease, especially since the latter estimates are wildly high and not backed up by actual data. If you apply the 1% fatality rate for cases that have tested positive to the highest percent of positive cases any country has found, you get 1% of .025%. With the US population that would be 80,000 deaths, not millions. I hope the fatality rate is not even this high, but claims that millions will die do not seem to be justified by the data from other countries based on what is actually happening as opposed to models or unproven theories. I am not trying to minimize the danger of this pandemic. I'm happy that we're taking precautions. I'm just trying to avoid panic based on bad math. Applying the fatality rate for confirmed positive cases to unproven estimates on the number carrying the disease, is not mathematically valid and gives a false picture of the danger. The truth is bad enough. Can we stick to that?

  168. @Hdb But other countries have taken early, aggressive steps to deal with the virus. We've done nothing in a month. If China still weren't testing anyone, or isolating anyone their numbers would be huge. And no, despite the constant misinformation spouted by our president, we are not testing people coming into US airports. None at all. I just flew in. Not even a sign of testing. Considering the size of our population, we aren't testing anybody.

  169. @Hdb Unfortunately, the math is pretty correct. Similarly, in Europe, there are estimates which raise the percentage of the virus carriers to climb to even 70% of the population if measures are not taken "yesterday". To prove those estimates wrong we all need to be responsible citizens, follow the advice of the experts and hope that our governments will take the appropriate measures neglecting any political cost. I wish you the best of health.

  170. This article is enormously helpful as well as hopeful, since it tells us that the best experts see a way to vastly minimize the destruction and death this virus will cause, IF our government acts quickly and responsibly, which means that WE the people need to make our voices heard. We don't want this to turn into a political football - we don't want to hear different sectors of government pointing fingers and assigning blame. That helps no one. We want all sides to work together, quickly,intelligently and effectively. This means that we need to listen very very carefully to the scientists, epidemiologists, doctors and other health experts who are at the forefront of the research, and we need to exert all the pressure we can muster to compel our elected officials to do the same. Moreover, we should be willing and grateful to accept any kind of assistance that foreign governments, especially those who have successfully managed the spread of the virus in their respective countries. China, South Korea and Singapore have shown us what works and what doesn't. This gives us a great advantage since we are not facing the same unknowns that our friends in Asia faced when their battle began. The real question becomes: Is our government prepared and willing to take advantage of that advantage. I would love to see a committee of 100 non-partisan leading scientists, doctor and other experts, issue a joint plan of action, and then challenge our government to enact the plan.

  171. "Any disease model is only as good as the assumptions programmed into it, and there’s so much uncertainty about the coronavirus that we shouldn’t see this model as a precise prediction." Why not use the algorithms used to exactly predict climate change. It is scientific.

  172. The virus cannot be controlled without data. No tests no way.

  173. Unfortunately not every citizen reads the NYT. In a related article, people are divided on the severity of the outbreak and the likelihood of effective mitigation, along party lines. With our “illustrious leader” stepping on his own words, contradicting health professionals, all outcomes are uncertain and trend towards the dire, towards the non flattened curve on your chart. My husband and I are both in our seventies. We’re not panicking, but are definitely afraid.

  174. Excellent points made here regarding Korea. Using the most recent data that I can find, the mortality rate in Korea is 0.9% It is important to note, however, that Korea has taken an entirely different position than many of the other countries have been able to take. They have been aggressive with testing, biosurveillance, and transparency -- above and beyond the “social distancing” efforts that we've been seeing here in the US or some other countries. The Koreans have ready access to World Health Organization testing kits and are deploying 10,000 tests, roughly, per day. They have high-efficiency drive-through testing system set up throughout the country to minimize spread to healthcare workers and other individuals. Patients do not get out of their car for a pharyngeal swab. They have adopted a position of amazing transparency and have used SMS technology to their advantage. As I understand it, when a COVID-19 test is positive, other individuals in the country of Korea receive SMS Messages with the location of the positive test while withholding the name of the patient for privacy purposes. These texts are quite detailed however indicating *possible locations* that the COVID-positive patient visited. There are obvious downsides to this approach, but one is not unrecognized spread of illness. I believe the data supports my assertion that mortality rates are greatly impacted by preparation and preparedness. Are we in the US more like Korea or more like Italy?

  175. Something seems off on the model. The time of peak infection count seems insensitive to the degree of intervention. Surely more aggressive intervention shifts the peak farther into the future (along with decreasing it somewhat). Can you explain the model in detail somewhere?

  176. I see the dash/blue line peaks on about Election Day 2020. Everyone is aware that Trump wants an excuse to call off the election and this will provide him with all the cover he needs.

  177. I don’t find articles like these helpful; it just accelerates anxiety and panic, which is high enough already. Yeah, we know it’s bad and going to get worse. We understand that.

  178. @Brad Disagree. I have friends who are freaking out, assuming 200m infected and 1.5m deaths. This makes it clear that taking action makes a huge difference and will help us avoid that worst-case model's outcome. If we take it to heart, that is.

  179. @Brad That was not the point of the article. It was to show that active measures on all levels: individual, community, state and federal can be effective in responding to a novel virus. It is hard to feel empowered by that amid the worries of economic fallout that can be more front and center for many of us, but if we can flatten out the curve, we will all be better off, not only in terms of our health but it will also be better economically.

  180. Italy did it like Trump: deny that corona is anything to be taken serious. So it is doubtful that the lower estimates proposed here are still realistic.

  181. I am dreading the inevitable economic impact this will have, and the fallout we will all experience with inconvenience and shortages. A lot of businesses and individuals are going to be terribly hurt; this will have consequences for all of us. Some of the damage will not be apparent for awhile, maybe many months.

  182. Here's my best guess: mild interventions beginning now, ramping up to moderate by late April or early May (remember, "president" Trump thinks the pandemic will, like the flu, subside during warmer weather - but it's not the flu), then in mid to late June, aggressive measures will be taken. Alternatively, perhaps this will behave like the flu pandemic of 1918-1919 and we'll see a tapering of the pandemic in early May. Because of the "dynamo", /s, in the White House and his "crackerjack crew", /s, "running" everything, things taper off for the summer and then come roaring back with gusto and greater and more potency. We will be caught unaware, flat-footed (much like now) and things will be dire. Other countries will have developed vaccines and treatments; however, the US will have nothing. Why? Because of what we have for a government led by a narcissist who believes he knows everything.

  183. Dr Anthony Fauci said on tv this morning that the peak will be "at least a matter of several weeks...a few weeks up to eight weeks." So I am confused, or maybe skeptical, of the much longer time scales of many months in these graphs.

  184. I have seen this equal-area curve with the horizontal “care capacity“ line in a number of contexts now, and it is a great concept visualization that helps frame how our individual choices affect us as a society. Missing is a conceptual vertical line far off to the right that represents the development of an effective vaccine that will cut off the flattened curve and actually make its area = number of infections less.

  185. Thank you for printing this. Here's the problem though. This information should be coming, not in an opinion column for the NYT, but as an announcement from the CDC. Fortunately, I live in Oregon. Yesterday in their news conference state officials made clear that, without interventions, more less 75,000 Oregonians would become infected. Farther, although 21 cases had been confirmed (at the time of the press conference), the officials stated that the real number was closer to 150 to 200. In the abject failure of the federal government to respond to this crisis, I am glad to see that the states (some of them at least) are stepping up.

  186. A lot of attention in this article is focused on "flattening the curve". And, of course, I think this is right. However, when I look at the idea “flattening in the curve”, I think of two curves: The first curve, and the one identified here in this article has to do with keeping caseloads beneath our capacity for treatment. But, importantly from an epidemiological perspective, we should think about this curve as it relates to our capacity for testing. If we cannot test, we cannot understand who to quarantine, and therefore cannot control this outbreak. Current epidemiological data supports the assertion that mortality rates are greatly impacted by preparation and preparedness. We are woefully unprepared from a diagnostic testing standpoint. Using a "foreign" WHO testing kit was deemed un-American by someone in the current administration. Sadly, we had at minimum six weeks to prepare for this. Soon, I fear, this will be more than sad, it will be tragic.

  187. Fantastic chart, very informative and succinct, the kind that even almost the White House could understand. Would like to bring one sobering-scary fact to mind: this model is based on a 1% mortality rate. So far, the rate has fluctuated at around 3%... Hope you're right, very much so.

  188. In the face of this crisis the executive branch has launched an all-out effort--to maintain Donald Trump's self-image.

  189. South Korea and Singapore have been very successful with this approach. Kind of hard to believe that America can't come up with cotton swabs at the end of little sticks for testing. Trump deliberately discouraged testing for political reasons. He needs to resign and take Pence with him.

  190. I was just telling someone 12 hours ago that this was the sort of analysis that was needed to explain what and when things needed to be done. Thanks NYT.

  191. Trump's and the Republican's grossly inadequate response only reinforces my view that the formerly grand old party has become genocidal. This wasn't sudden: back during the AIDS epidemic they refused at first to check the blood supply for that virus, and then some "people of faith" claimed that "AIDS is [G]od's judgment on gays." Today's Republicans' mantra seems to be Make America White Again, and the victim groups of choice now include not only the poor, elderly (privatize Medicare! or lure people into "advantage" plans they can't easily switch out of--all for corporate donor profits), people of color, Muslims, and LGBTQ people. When that's the strategy, NO group of Americans is safe. And epidemics are far cheaper than actually creating jobs (and using resources and money) in an traditional genocide scheme, which involves rounding up victims, building and staffing murder facilities, and transporting the victims there. In other words, mass murder by malign neglect. Call them out on it...and vote them out on it.

  192. Wonderful, if scary article. Truth is, at this point, I am taking it all personally, wishing I could just go to some central point and get tested. If no one knows if they are sick, and since we all need to shop and travel, sooner or later we will be around someone who has coronavirus, or we will make someone else sick. This debacle reminds me of some other stupidities my government muddled into, including the Vietnam and Iraq wars. We like to think we elect competent people, but geez... Hugh

  193. You are playing mathematical models with very hypothetical parameters. Are you really being responsible by writing that up to two thirds of the US population "could" become infected? Later you do relativise the statement. But, that simple sentence is what the reader will take away from your article. At a time when panic is taking hold and reason is in short supply, I would have expected less sensationalism.

  194. If nothing else hopefully this will prove to ALL Americans that the current administration is incompetent and the worldview that free markets will take care of things will finally see its death knell before the body count becomes unbearable.

  195. This clear and concise call to action is precisely what the president should have, could have, presented to the country. Instead we get word salad and self-congratulation. The House should vote articles of impeachment today for criminal negligence. Trump's abject incompetence and willful misdirection would be actionable in any sane society He is literally causing death through intentional obfuscation and malfeasance. The Senate could convict in a 15 minute trial. I am not being lighthearted. This action might save hundreds of thousands of lives.

  196. Michael Osterholm, MD, Epidemiologist, explaining why the US is ill-prepared for this pandemic...in part because the majority of the simple everyday medications we use and many of the acute care medications we depend on are manufactured in China and India...these supply chains are broken right now. Video released on Tuesday. https://www.youtube.com/watch?v=E3URhJx0NSw&t=2540s

  197. The ignorance running through the general public is discouraging and the fact that it is being validated by right wing talk radio is just amazing. Turn on your AM radio and listen to people minimizing this disease, complaining people are over reacting and otherwise down playing the potential impact. The folks who get their news from Fox really are way under informed on the facts of Covid-19 also. Ironically their prime demographic is the most vulnerable. We also seem to be handicapped by the group in charge on the Federal level. The president has spent all the early preparation time down playing the threat. Pence is outright lying about the availability of testing. All the other political officials follow their lead because they've seen what happens when they contradict Trump ("you're fired!"). They now seem to be coming around but will need to perform much better in order to be effective. We need more medical supplies, respirators, ICU beds, many more tests, anti virals and a vaccine. We need a war effort as Sanders said. We need someone to rally every medical scientist in the country to shorten the time to develop a vaccine. We have the resources we need someone to marshal them to the cause. So far our Federal officials have let us down but that does not have to continue.

  198. The virus was here before your graph, and real data needs to be modeled from prior conditions. Your graph is bad and will only increase pandemonium.

  199. By refusing to listen to the best available (or even the 10th best....) expert scientific advice FROM THE BEGINNING, Mr. Trump is personally responsible for each and every excess death that occurs because of his cavalier attitude towards what was a "clear and imminant" threat from this new virus. Its a pity he can't be tried and convicted for criminal negligence causing death - or is that criminal narcisssim, since until the fat REALLY started hitting the fire around the world, all he was worried about was his election campaign being damaged by the economic fallout. VERY sad. VERY criminal.

  200. We have currently about 3,500 death from Corona Virus in China (over 1 billion people). How can a model be trusted that assumes 50,000 death for the US as best case scenario? I agree, very irresponsible.

  201. Nope. Your numbers are wildly inflated. Follow the South Korean data in particular closely. Once the ramp started in South Korea, it has declined to very few new cases in about 5 weeks. Four weeks from now the US will have very low lumbers of new cases. My estimate: 10,000 cases with 600 deaths...That is %0.6 death rate.

  202. @TJ Wow. Have you paid no attention to how South Korea has responded vs how the U.S. has responded? It is not a matter of raw numbers, it is a question of the public-health response once the virus has been detected.

  203. @TJ The reason to follow South Korea's numbers and not Italy's numbers?

  204. This is important and far better graphically than what I was seeing on TV, where the infection curve was a little above hospital capacity. We can only hope Dr. Fauci continues to be in the forefront, but he is at high risk himself. Dr. Redfield appears to not be very competent, judging by his performance with Rep. Katie Porter yesterday. The Cleveland Clinic and the 11 other hospitals who have rapid testing with results in 8 hours would do us all a favor by sharing the technology nationwide. It's time for all of us in the rest of the country to pul together, because help may not be coming from DC. I can't help but wonder what a highly functioning CDC would have done by now. Elections matter.

  205. Excellent article, informative and non inflammatory, just well presented facts and facts are what are missing in much of the media coverage at the moment. Your average person just wants to know what is this bug and how do I avoid it, what are the signs to look for and what happens to me if I get it or someone I know gets it. There is an absolutely essential article in the SMH which readers here should take a minute to peruse. It covers even more than this article and when the two are read in conjunction and the information taken on board...your fairly well on board of what can happen, what to look for and how to behave. https://www.smh.com.au/national/what-does-covid-19-do-to-the-body-and-what-s-it-like-to-have-the-illness-20200302-p5465a.html It has a chart titled: The risk of dying if infected with COVID-19 (by age group) Based on more than 72,000 cases in China until February 11. I can't reproduce it here but it is in that link and it will allay some fears for sure for many people worried about this. It is indeed the elderly who are most at risk here. It explains how they tried to save the life of the young doctor in China who first alerted the world to the virus and also shows why medical staff are most at risk. Would love to see the article published in the NYT as well.

  206. BBC this AM reporting India's caseload has begun, and China begins to propagandize that there is no proof the virus began in their wet markets—the world should thank them for their unprecedented response. So. They won't look into these markets, then? A little early for the brand of self-congratulation China is beginning to promote. We don't know if people can be re-infected, and if India flares incredibly, the West of China may be re-infected. And while it is generally true viruses fare better in the cold—half the globe is always in winter. Our virus needs only winter over in the Southern hemisphere, as the Spanish flu did—and return again in the fall. A Journal of the Coronavirus Years—we need a volume to match Samuel Pepys, or Gabriel Garcia Marquez. Simply seems so surreal.

  207. Now if 'Fox News' would ever produce and air a similar tract, would that audience understand, accept and integrate the premises and results? That is the failing of news that does not educate its audience but attempts to herd lemmings.

  208. The fact remains that we do not have an adequate ability to test people, as well as "leadership" that is utterly incompetent. So by that qualification alone we are in for a serious epidemic.

  209. US has not tracked and traced infections - massive case of under reporting. Washington state had a serious problem and no effort to contain. You guys have a serious problem and the government is not taking it seriously. China has shown that lock down works and it a model you need to start thinking about. Watch Italy with interest and see how long it take to contain. Spain/Germany and France may need local or country wide lock down by mid April. UK may follow by mid-may. If you don't prepare for that then you will have people dying at home because hospital system cant cope with 1m to 2m cases.

  210. Trump will NEVER acknowledge the seriousness of the virus because he is incapable of admitting that he was wrong. Furthermore in his mind the flu kills more people annually than does the virus making it less dangerous. Also social distancing is for him an abstract concept that he does not understand.Particularly since it interferes with the adrenaline charging mass rallies of his fervent supporters.

  211. Any ... model is only as good as [its] assumptions and the way it’s designed to work.

  212. Thanks for this article. Well done. One minor correction: some of your red lines are labeled as "infections", which makes sense when you are pointing out that the US could experience 100 million potential infections. Others of the red lines should be labeled "hospitalizations", eg when you say we could have 9.4M or 533,000 infections (should be hospitalization) at peak.

  213. If Trump's current behavior and the distorted blathering of conservative media is any indication of what is to come - we are headed towards the worst case scenario. Where are our true leaders who will make a difference? It's not just one person trying to boost the Dow. It's all of us who take appropriate precautions now.

  214. What's interesting with the models is they will not be applied equally. In real time you will get to see if the aggressive measures in, say, New York work better in states that don't adopt them. We will get to see the 'overreacting' perspective play out. This pandemic is just a worse (Not necessarily worst) episode. More will follow. Some less severe, perhaps some more. How we handle this one will determine how we handle the next. My biggest fear is human arrogance. Trump will have you believe we are better than viruses simply because we are human. And Americans, being the 'best' humans on the planet, are better at defeating this than, say, the Germans. Isn't it so ironic the biggest threat to Americans aren't 'invaders' at the border? It's the ones sitting right in front of your mucous membranes. The ones you can't see. And Trump wants to Pooh, Pooh the experts who suggest building a 'social distance' wall to keep them out.

  215. Trump, FOX, and Republican "leadership" believe the real threat is to Donald Trump's re-election prospects. That is clear from their rhetoric.

  216. "We’ll be honest: We worried that the clean lines in the graphics here risk suggesting a false precision. None of us know what lies ahead. But the wise uncertainty of epidemiologists is preferable to the confident bluster of television blowhards. The one thing we can be confident of is that enormous risks lie ahead — including a huge loss of life — if we don’t take aggressive action." This paragraph should have been the lead. Not everyone reads a long article through to the end, even a good one such as this. Peace.

  217. BBC this AM reporting India's caseload has begun, and China begins to propagandize that there is no proof the virus began in their wet markets—the world should thank them for their unprecedented response. So. They won't look into these markets, then? A little early for the brand of self-congratulation China is beginning to promote. We don't know if people can be re-infected, and if India flares incredibly, the West of China may be re-infected. And while it is generally true viruses fare better in the cold—half the globe is always in winter. Our virus needs only winter over in the Southern hemisphere, as the Spanish flu did—and return again in the fall. A Journal of the Coronavirus Years—we need a volume to match Samuel Pepys, or Gabriel Garcia Marquez. Simply seems so surreal.

  218. Only yesterday, I think, I read comments about chancellor Merkel´s warnings, which she issued in Germany. Comments to the extend of "How can she scare everybody like that? If this was true it could mean 0,5 mio dead." Well,... Already yesterday I felt she was (involuntarily) raising awareness about the seriousness of our problem not only in Germany but also in the rest of the world, while other so-called leaders are gone golfing. I am glad to see this article and simulation by the NYT - and to see that commentators do not consider this alarmistic any more. What a difference a day makes.

  219. What’s really frustrating is that America had a few weeks to see what was happening in Asia (both with the virus spread, & countermeasures being taken) and to prepare, but little was done except lies about how it’s nothing to worry about, just like the seasonal flu.