Number of new COVID-19 cases worldwide is declining now?

The last four World Health Organization Coronavirus disease situation reports, 63-66, show an almost flat number of new tested-and-reported COVID-19 cases worldwide each day:

  • 40,788
  • 39,825
  • 40,712
  • 49,219
  • 46,484 (report 67; March 27; bold face because added as an update)
  • 62,514 (report 68; March 28)
  • 63,159 (report 69; March 29)
  • 58,411 (report 70; March 30)
  • 57,610 (report 71; March 31)
  • 72,736 (report 72; April 1)
  • 72,839 (report 73; April 2)
  • 75,853 (report 74; April 3)
  • 79,332 (report 75; April 4)
  • 82,061 (report 76; April 5)
  • 77,200 (report 77; April 6)
  • 68,766 (report 78; April 7)
  • 73,639 (report 79; April 8)
  • 82,837 (report 80; April 9)

Original text, before the data points in bold were added: Almost, but not quite flat, right? After four days it grew from around 40,000 to around 50,000. But consider these data against the background of a rapidly ramping up testing infrastructure. More people are being tested, not just celebrities and political elites with mild symptoms and peasant hospital inpatients. If dramatically more people are being tested every few days (and therefore fewer mild cases never get noticed/recorded as COVID-19), don’t the above data suggest that the actual number of new cases (humans actually infected as opposed to the proper subset of humans tested and reported as infected) is going down?

Massachusetts Update:

3/27 analysis: Not an obvious exponential growth process and, even if it were, patients aren’t generally tested until admitted to a hospital (i.e., current growth in cases reflects a growth in infections that happened 1-2 weeks ago)

[Update, evening 3/27, text message from friend: “I asked [physician wife] how her nurse friend is doing at the hospital. She said she has not worked in a week because the ER is so dead she can’t get hours. The hospitals are empty.”]

3/28 analysis: The Massachusetts data are worrisome. The number of reported tests stayed constant from March 27-28, but the percentage of positives grew. We have no information about when samples were taken, however, so this growth could have occurred several days earlier. (Test results show up in this report on the day that the tests are completed.) The WHO data suggest that the answer to the question posted in the headline is “no” (though it is tough to say given that the WHO situation report does not say how many tests are being performed worldwide).

3/29 analysis: Massachusetts testing actually fell. Only 17 percent of tests were positive, about the same as the average of the two previous days. Linear growth is a better fit than exponential growth. The WHO data suggest weak exponential growth.

3/30 analysis: Massachusetts testing fell again, but 21 percent tested positive. Suggests that doctors are now better at figuring out who needs to be hospitalized (since it is hospitalization that leads to being tested). WHO data show a slightly reduced number of both cases and deaths.

3/31 comment: Where’s the exponential growth?

4/1 comment: Okay, maybe we do have (slow) exponential growth in both worldwide and Massachusetts cases.

4/2 comment: The world is doing better than Massachusetts, but if Farr’s law is in operation, it is tough to guess the top of the bell curve.

4/3 comment: Sweden (open) has 2X the population of Denmark (locked down), so the 519:279 ratio of new cases works out to roughly the same as the population. The ratio of deaths (43:19) is also roughly the same as population ratio. So it looks as though the Swedish epidemiologist’s prediction that government policy would have a minimal effect is proving correct. For the world overall, it has been 11 days and still the number of new cases has not doubled. WHO is still not publishing data regarding how many tests are being performed, so there is no way to know if this slow doubling corresponds to a 4X increase in testing and a falling number of actual new infections or a steady rate of testing and a doubling-every-two-weeks rate of actual new infections. Massachusetts data suggest doubling every 8-9 days. They also confirm the wisdom of the New Yorkers who fled via Gulfstream to their Nantucket mansions: only 9 cases on the island.

4/4 comment: Sweden continues to have roughly the same number of new cases, per capita, as Norway and Denmark, and Swedish hospital/ICU capacity remains sufficient (and public). It seems to be essentially impossible to die from COVID-19 in Norway. Do they have a better treatment that they’re not sharing? Or do they record the death of someone with underlying conditions differently? Or are they keeping people on ventilators beyond the point at which the Swedes would have given up hope?

4/5 comment: Sweden has 365 new cases; Denmark has 320. Keeping in mind that Sweden has 2X the population, this suggests that the Danish “lockdown” is not “making a difference” as rich Americans like to say. Massachusetts is back to the same number of new cases as 9 days ago. Farr’s law, but with an extremely poor fit to a Gaussian?

Monday, 4/6 comment: Sweden has 387 new cases; Denmark has 292. I.e., Sweden continues to have a lower per-capita new-case rate despite the lack of a lockdown. Sweden even has a lower per-capita death rate, 28:18. Spain and the UK are contributing hugely to the world total of new cases, each with roughly 6,000. Massachusetts numbers are back where we’d expect them to be, perhaps due to reduced lab activity on Sunday.

4/7 comment: Sweden has 376 new cases; Denmark has 312. The per capita new death ratio is no longer favorable to Sweden, however, at 4.75X (small sample, however; Denmark had only 8 deaths). Globally, COVID-19 has killed more than 72,000 people, i.e., about the same as the number of Americans killed each year by taxpayer-funded (via Medicaid) opioid addiction and overdose. But the new case count seems to be declining in accordance with Farr’s law. The Massachusetts numbers are worrisome. New cases are stubbornly high and 356 people have died so far (roughly the same, per capita as in Sweden; i.e., a poor argument for our state’s school closure and other shutdown efforts).

4/8 comment: Sweden has 487 new cases; Denmark has 390 (i.e., more per capita in the locked-down country). Worldwide new cases are still on a plateau. Massachusetts cases continue to grow, consistent with a doom-and-gloom forecast from University of Washington that Massachusetts will end up worse than New York, adjusted for population. Fully 2.6 percent of people in Massachusetts with confirmed cases are already dead, suggesting that we are not better at caring for patients than were the doctors working in Wuhan (death rate roughly 1.4 percent). As in China, though, our data are skewed by limited testing.

4/9 comment: Sweden is at 726 new cases; Denmark 331 (comparable when adjusted for population size). Worldwide new cases continue on a plateau. Massachusetts in shutdown seems to delivering the exponential growth that the media was hoping to see from wide open Sweden.

18 thoughts on “Number of new COVID-19 cases worldwide is declining now?

  1. Well, your president has already said that the number of cases “within a couple of days is going to be down to close to zero.””

    Making Baghdad Bob seem like a well informed source of truth, your president does. The collection of presidential statements below is from, and it is not mine.
    January 22: “We have it totally under control. It’s one person coming in from China. It’s going to be just fine.”

    February 2: “We pretty much shut it down coming in from China.”

    February 24: “The Coronavirus is very much under control in the USA… Stock Market starting to look very good to me!”

    February 25: “CDC and my Administration are doing a GREAT job of handling Coronavirus.”

    February 25: “I think that’s a problem that’s going to go away… They have studied it. They know very much. In fact, we’re very close to a vaccine.”

    February 26: “The 15 (cases in the US) within a couple of days is going to be down to close to zero.”

    February 26: “We’re going very substantially down, not up.”

    February 27: “One day it’s like a miracle, it will disappear.”

    February 28: “We’re ordering a lot of supplies. We’re ordering a lot of, uh, elements that frankly we wouldn’t be ordering unless it was something like this. But we’re ordering a lot of different elements of medical.”

    February 28: “Now the Democrats are politicizing the coronavirus, you know that, right? Coronavirus, they’re politicizing it. We did one of the great jobs. You say, ‘How’s President Trump doing?’ They go, ‘Oh, not good, not good.’ They have no clue. They don’t have any clue. They can’t even count their votes in Iowa.” “They tried the impeachment hoax. That was on a perfect conversation. They tried anything. They tried it over and over. They’d been doing it since you got in. It’s all turning. They lost. It’s all turning. Think of it. Think of it. And this is their new hoax.”

    March 2: “You take a solid flu vaccine, you don’t think that could have an impact, or much of an impact, on corona?”

    March 2: “A lot of things are happening, a lot of very exciting things are happening and they’re happening very rapidly.”

    March 4: “If we have thousands or hundreds of thousands of people that get better just by, you know, sitting around and even going to work — some of them go to work, but they get better.”

    March 5: “I NEVER said people that are feeling sick should go to work.”

    March 5: “The United States… has, as of now, only 129 cases… and 11 deaths. We are working very hard to keep these numbers as low as possible!”

    March 6: “I think we’re doing a really good job in this country at keeping it down… a tremendous job at keeping it down.”

    March 6: “Anybody right now, and yesterday, anybody that needs a test gets a test. They’re there. And the tests are beautiful…. the tests are all perfect like the letter was perfect. The transcription was perfect. Right? This was not as perfect as that but pretty good.”

    March 6: “I like this stuff. I really get it. People are surprised that I understand it… Every one of these doctors said, ‘How do you know so much about this?’ Maybe I have a natural ability. Maybe I should have done that instead of running for president.”

    March 6: “I don’t need to have the numbers double because of one ship that wasn’t our fault.”

    March 8: “We have a perfectly coordinated and fine tuned plan at the White House for our attack on CoronaVirus.”

    March 9: “This blindsided the world.”

    Edit: March 9: “The Fake News Media and their partner, the Democrat Party, is doing everything within its semi-considerable power (it used to be greater!) to inflame the CoronaVirus situation, far beyond what the facts would warrant.”

    March 10: “It will go away. Just stay calm. It will go away.”

    March 13: National Emergency Declaration

    March 13: “I don’t take responsibility at all”


    March 16: “I give myself a 10 out of 10”

    March 17: “I’ve always known this is a real—this is a pandemic. I felt it was a pandemic long before it was called a pandemic.”

    • Thanks, Jonas, but I’ve been getting my coronavirus info from WHO and similar sources, not from any politicians (though, bizarrely, Professor/Dr. Trump and the Imperial College epidemiologists now seem to agree that the worst will be over by Easter!).

    • Philip: If you think the growth in confirmed cases is being driven by wider testing, look at the number of deaths instead.

      “I’ve been getting my coronavirus info from WHO and similar sources”

      WHO is saying that the rate of new COVID-19 cases worldwide is declining?

      I’d describe it differently: You’re getting your raw data from WHO. Your interpretation of the data, on the other hand, seems to be driven by confirmation bias (which is of course a universal failing), and your belief in what might be called anti-anti-Trumpism – that criticism of Trump simply springs from personal prejudice against Trump. This is the curse of high intelligence: you don’t trust what you read in the New York Times, or what the WHO says about the pandemic, because you think you’re smarter and more rational than they are.

      If you have no reliable news source you can trust, you end up in this kind of epistemological vacuum, where the unpublished Oxford study and the Aaron Ginn essay have just as much credibility with you as what the WHO is saying. I assume they came to you through Facebook or through email?

      As Jonas points out, Trump’s statements so far have been wishful thinking, completely detached from reality. If COVID-19 turns out to be far less dangerous than the conventional wisdom, this will of course make Trump’s critics look stupid, and make you look smart. I would respectfully suggest that this gives you a strong bias towards thinking that COVID-19 isn’t that dangerous, despite the deaths in Italy, Spain, and now the US.

      COVID-19 is a new and highly contagious virus; we have no immunity to it. Trump may not understand exponential growth, but you do.

    • Why does the steady rise in number of deaths cast doubt on a falling number of new infections? Wouldn’t deaths lag infection by 2-8 weeks? (see ) So a change in the number of deaths tells us how the rate of new infections was changing perhaps 3 weeks ago? Maybe 4 or more weeks ago in a country with a lot of intensive care capability?

    • Just updated the original post with today’s WHO number of new cases. It is lower than the previous day’s (i.e., declining!).

    • “So a change in the number of deaths tells us how the rate of new infections was changing perhaps 3 weeks ago?”

      Both the number of deaths and the number of confirmed cases are lagging indicators. Median time from infection to death is around 17 days. Median time from infection to confirmation is about 14 days (five days to first symptoms, seven more days to hospitalization, two more days to get test results). Looking at changes in either number tells you what was happening to the rate of infection about two weeks ago. (In the US, from the number of deaths, we can see that the rate of infection was doubling every three days.)

    • Deaths are not an accurate indicator either, and could be simply and artifact of more testing simply because it’s not deaths from covid-19, it’s deaths where the virus was detected. The “normal” rate of pneumonia deaths is still much greater than covid-19’s.

  2. Honestly, the corona virus is not too big of a concern. Most cases are mild, and some are even unknown. Governments are making a big deal out of it so they can obsess over something.

  3. “Nothing to fear but fear itself” [1]

    When this is all over, I think history will record today’s event as one of the biggest man-made blunder of all times. We are letting fear over take us vs. some common sense.

    This is first hand account: Both of my kids are nurses, one is NP the other is RN, each work at different hospital here in the Boston area. At both hospitals, they converted at least an entire floor for COVID-19 patients some 2 weeks ago. How many beds have patients as of yesterday both hospitals combined? Less then 10. And those who are in a hospital bed are all older patients with preexisting conditions such as heart issues due to smoking or drinking to name some. All other beds are simply left empty and nurses are getting bored having nothing to do.

    Yes, there are COVID-19 can be deadly, painful and is contagious. But so is a flu. What do Drs. tell a flu patients? Use common sense, stay home so you don’t pass it around, wash, rest and watch you health. In fact, that’s exactly what Dr. are tell their COVID-19 patients to do, but yet somehow COVID-19 has become this virus that makes it sounds as if you get it, you will die from it no matter what. Thus, we injected “fear” into the public.


    • And to make it super clear. Those COVID-19 patients at my kids hospital, none are on ventilators or life support. Their symptoms is no different from a severe flu symptom but is now aggravated due to preexisting conditions that they have.

  4. A co-worker died at senior management’s company, in New York City. All in the company have been working from home the last two weeks. There were about 100 working in the office. That is the first coronavirus fatality I can personally confirm.

    There are two trends to track: coronavirus and fear. The personal fear index in my household just skyrocketed as the personal coronavirus death toll reached to about what we should expect.

    Thanks for linking to the WHO reports. How the report organization and presentation changes over time is noteworthy.

  5. Is testing capability growing at at least the same rate as infections in the regions experience the most absolute growth?

    What does “If dramatically more people are being tested every few days” main in terms of actual growth rate? Linear testing growth in the face of exponential infection growth would still mean that the ability of testing to capture the true infection landscape is getting worse day-by-day not better.

    • For the thousands time: epidemics are NOT exponential (they only appear to be so at initial stage) and follow the bell curve, aka Farr’s law.

      Everyone talking about exponential growth or making projections based on these basically has no clue. That includes most of the fake “experts” in govt and media.

    • averros: We are still in the initial stage of the epidemic. Even if the number of infections in the US is one million (a bit less than 100X what it was in Wuhan shortly before China imposed a near-total lockdown on the city), that would be only 0.3% of the population.

      Philip thoughtfully posted links to the Massachusetts numbers. On March 27, the total number of confirmed cases was 3240 (823 new cases). On March 24, it was 1159 (382 new cases). On March 21, it was 525 (112 new cases). Philip comments that this is not an obvious exponential growth process. I’m tempted to add that the doubling time is obviously not three days.

    • Hooded A: If there is exponential growth in infection rate AND linear growth in testing, shouldn’t we expect to see (a) dramatic growth in “new cases”, and (b) a higher percentage of positive tests?

      The WHO summary reports don’t tell us how many tests were performed, but Massachusetts does. Massachusetts has neither a dramatic rise in new cases (a) nor a significant change in the percentage of positive results when a test is actually done (b). As noted above by various people (including me), because “new cases” are limited to hospital inpatients, this is a window into how things were growing at least a week earlier (Russil says 14 days).

    • Philip: glad to see that you’re updating the post with the daily Massachusetts count, and that today’s count was lower. Logarithmic graph of cases in the most-affected states. Massachusetts now has more cases than Washington state. Hopefully the slope of the curve will continue declining, reflecting the effectiveness of social distancing (whether voluntary or mandated) from two weeks ago. (It doesn’t have to be perfect to have an effect!)

  6. averros: No natural exponential process is exponential forever, there is always a limiter. You claim it is not exponential then claim “it only appears be at the initial stage”. What does it mean for growth to “appear” to be exponential? It literally is exponential (for a period) is the explanation. You can plot the growth on a log graph and take the slope.

    philg: The behaviour depends on the cohorts being tested versus the cohorts getting sick.

Leave a Reply

Your email address will not be published.