Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Model built by researchers suggests coronavirus outbreak began in November (utoronto.ca)
108 points by hardmaru on Feb 10, 2020 | hide | past | favorite | 73 comments


I am writing this from a cruise ship in the south china sea. We have been at sea for ten days because one port after another has been closed off to us. Some have closed to all cruise ships, some have closed to use because we have passengers who transited through Hong Kong. Every passenger now has to have their temperature taken twice a day. We have been told that we will be able to disembark in Singapore two days from now, but God only knows.

Corona virus has infected fewer than 100,000 people and killed less than 1000. To put these numbers in perspective, seasonal flu results in 3 to 5 million cases of severe illness every year, and kills 290,000 to 650,000. [1]

Personally, I am much more worried about public officials freaking out about the corona virus than I am about the virus itself.

[1] https://www.who.int/news-room/fact-sheets/detail/influenza-(...


So what do you think would happen if no quarantine was in place (because “the flu exists”), the virus spreads and suddenly that 1% mortality rate starts to stack up.

Is your country ready to handle 1% of its population dying, and many more falling ill? Flu is seasonal - predictable, known and accounted for. When flu season hits the system copes. This virus? Not so much.

Sorry you’re stuck on a boat, but it’s for the best.


Further, it's very clearly higher than a 1% mortality rate.

Those deaths are from the lower outbreak figures from weeks ago, as people typically don't die in the first two weeks of infection. According to doctors dealing with this in China, it's the third week that is where patients are dying.

The questionable official numbers have us at 40,171 infections with 908 dead. That's a 2.2% mortality rate off the max figures. More realistically let's go back in time, conservatively, two weeks ago: on 1/27 there were only around ~4,500 infections.

That's a mortality rate a lot closer to 20% (even if you inflate the infection numbers, it's high). Those 4,500 people are the ones that have overwhelmingly contributed to the 908 dead figure that exists today. Unless one assumes there were a lot more than 4,500 infections two weeks ago; or people are dying a lot faster than reported.

With tens of thousands of infections in the last two weeks (again, at the supposed numbers), how many will be dead at the given rate, in a few more weeks? Something obviously doesn't add up with these figures, one direction or another.


I'm not saying that nothing should be done. I do think, however, that focusing so much attention on cruise ships is a little ridiculous, and not just because I happen to be on one. The virus us much more likely to spread through airports, but I don't see international air travel being shut down.


List of airlines that suspended flights to China. https://www.reuters.com/article/us-china-health-airlines-fac...


But we have not been to China and we are not going to China. We were originally going to Hong Kong but that plan went out the window over a week ago.

We have been denied entry to the Philippines, Taiwan, Viet Nam and Guam (despite the fact that our last landfall was in Guam). How many airlines have suspended service to those destinations?


Two thoughts the first is that quarantines work and the second is that restricting travel across borders do not.

You are under quarantine firstly. As a consequence of this you are also being restricted with travel. A boat makes it very easy to do both.

Generally closing borders are bad for the sentiments you give: it's unjust, I'm not infected. People will cross a closed border illegally and officials will not be able to know. If you had the opportunity you would leave, right?

Open borders allow officials to track those coming in and if necessary quarantine them. Leaky borders have people actively avoiding quarantine and surveillance which could lead to much worse effects


> You are under quarantine.

No, we're not. But even if we were, it would be a moot point. Taiwan, Philippines and Guam are closed to all cruise ships. But the airports are still open.


Wait...you're NOT under quarantine? That's messed up.

What were the start / destination ports?


We started in Guam. We have not made landfall since.


Just in case anyone is still following this thread, this is the sort of insanity I'm referring to:

https://www.businessinsider.com/westerdam-cruise-ship-denied...

The Westerdam has been in effective quarantine for two weeks, no one on board has the virus, and they are STILL being refused permission to dock just because they came from Hong Kong.

We are still en route to Singapore, scheduled to arrive this afternoon. No hint of anything more going awry for us so far, but I'm not going to count any chickens until I've cleared customs back home in the U.S.


Glad they let you off in the end.


I was actually not on the Westerdam, I was on the Crystal Symphony. We ended up disembarking in Singapore on Feb 12.


The Westerdam appeared to have someone that was infected when they were allowed to disembark (https://www.nytimes.com/2020/02/16/world/asia/coronavirus-cr...).

I'm hoping this doesn't happen on the Crystal Symphony.


Has anyone on board been to China in the last few weeks?

Also with the H2H transmissions from Singapore (UK & Korea), it might become difficult to get back from Singapore if they go to Code Red.


heading Singapore tomorrow and hopefully the UK on the 17th...


But there are people on your ship who may have ncov2019 in it's incubation period. I don't see why it's necessary to risk the lives of our nation's elderly, who are very much loved, who are most at risk from these viruses. In addition while this viruses transmission to cats is unconfirmed unlike the h1n1 where it did happen, I also don't see why it's necessary to risk it. It just takes 1 person in the incubation period infecting someone else who frequents a cat cafe to devastate a community (and the cats would get a bunch of people sick as well).

The US has a travel ban for foreigners who have been in China, but they also ask people to self quarantee if they've been in contact with those from areas where the virus is more widespread, which is everyone on your ship.

It's unfortunate to be stuck on a ship and it's unfortunate that your captain didn't anticipate these measures.


Our last landfall was in Guam. The passengers who transited through Hong Kong embarked in Guam. They were admitted to the U.S. Technically, we are still in the U.S. So any U.S. travel ban is (or at least should be) irrelevant to our situation.


I don't think the Corona Virus is as benign as a flu virus. Sure the mortality rate is 1% but that is what we know today. No one knows what this Virus is capable of, in terms of mutation etc, and it is because of these unknowns organisations and public officials are on alert and taking quarantining steps - rightly so.

It is foolhardy to undermine WHO's cause for concern.

I am sorry you are stuck on a ship and I do hope you get to disembark in Singapore.


Yeah from the Singapore numbers, 45 cases and 7 in critical care. I don't recall seeing those sort of numbers with flu.

edit - googling a bit the % going into critical care seems to be about 0.2% for flu vs the above about 15% for coronavirus.


Corona viruses are a well known class of virus.

Sure, this is a new strain, but the flu is a new strain each year too


N. Taleb wrote a very good note [1] on why conventional risk-benefit considerations don't apply in the case of an unknown contagion like NCov19 - the comparison to the seasonal flu is not helpful in these cases (thin-tailed risks vs fat-tailed risks which are an existential threat on a long-enough timeline).

[1] https://www.academia.edu/41743064/Systemic_Risk_of_Pandemic_...


Really good read, thanks. Tl;dr, Taleb et al think extraordinary measures to reduce transportation flow can be very effective and that policy makers shouldn’t shy away from employing them when facing a viral outbreak of unknown tenacity.


> Corona virus has infected fewer than 100,000 people and killed less than 1000

I heard this bogus reasoning so many times in the last weeks. You're comparing a disease that's global and reached pretty much everywhere to the one that's at its initial stages of spreading.

Coronavirus is highly contagious, at least 50-200 times more deadly than flu. It's exactly the virus you should be worried about.

When you walk with a flu into a doctor's office, he doesn't bat an eye. Coronavirus is treated in hazmat suits.


I'm so sick of people comparing this to the flu.

It's nothing like the flu -- coronavirus death rates are 2%, flu 0.01 -- it's 200 times as deadly and it can spread before any symptoms.


We don't know the death rate of nCov-2019, because we don't know how many people are infected, and we don't know how many people are dying.


Do we really know how many people are infected with the flu? I'm sure only the more severe cases get tested for both.


I’m totally out of my element here but I’d wager that because the flu is much more pervasive we’re able to extract more reliable statistics, even if it’s from a relatively small sample set.


I'm no statistics expert, but given they can predict election outcomes from samples of a couple thousand people for tens of milions of voters, 40k+ infected with coronavirus is enough to have a very good idea about death rates.


The table on Wikipedia Influenza page gives <0.1% for fatality rate of seasonal influenza.

So perhaps just 20x as deadly as the seasonal flu.

https://en.wikipedia.org/wiki/Influenza


you are comparing 2019-nCoV Case Fatality Rate with Influenza mortality rate? why?

Wikipedia states:

Asian (1956–58) flu ~ 0.1% [44] Hong Kong (1968–69) flu ~ 0.1% [44] Influenza A, typical pandemics < 0.1% [44]

So 2% would be roughly 20x Asian or HK flu, or in terms of of a typical Influenza A pandemic at the very least 20x harder...


I'm not sure he is comparing Case Fatality Rate with Influenza mortality rate. I thought 1-2% was a reasonable estimate of the % of people who catch coronavirus who'll die. The CFR seems higher.


because it’s a severe illness we are all familiar with.


read my comment more closely, why are we comparing a voltage with a current?

CFR Case Fatality Rate: fraction of people diagnosed with an affliction that eventually died.

For example if the Sniffles has a CFR of 1%, then 99% of people diagnosed with Sniffles will survive.

Mortality rate: fraction of all deaths in a group, that are attributed to a specific affliction.

For example if mortality rate of Sniffles in Belgium is 4%, then 96% of deaths in Belgium are attributed to other afflictions.

For a comparison check out the following lists:

https://en.wikipedia.org/wiki/List_of_human_disease_case_fat...

https://en.wikipedia.org/wiki/List_of_causes_of_death_by_rat...


Interesting distinction, but I don't think anyone was talking about mortality rate over all people. All the discussion I've seen is the mortality rate of people with the disease.


so people are discussing CFR's but calling it mortality rates

consider how mortality rates are extensive quantities i.e. proportional to the number of people infected while CFR's are intensive quantities i.e. independent of the number of people infected


I agree, 2% is terrifying, try an online rng and see if you survive. Then try it for close friends/family. Its russian roullete with a few more chambers.


The 2% figure is across all groups. For older people, diabetics it would be much higher. And that 2% assumes a scenario where there are infinite medical resources. The


> Corona virus has infected fewer than 100,000 people and killed less than 1000.

Of the 42,638 who have caught coronavirus, 1,016 have died, and 3,996 people have recovered.

Because it's not simply ongoing but accelerating, we can't say the fatality rate is simply 1k/42k = 2.4%. The overwhelming majority of the people who have caught the disease still have the disease, and we can't say for sure that it is not going to kill them. We can say for sure that the mortality rate is significantly higher than 2.4%.

https://bnonews.com/index.php/2020/02/the-latest-coronavirus...


I'm seeing a lot of posts on why the seasonal flu is worse.

This to me seems ridiculous by any reading of even the reported numbers.


We don't know the final statistics, especially since the numbers we're getting are coming from the Chinese government.

It's simply too early to say whether we're overreacting or not compared to how we handle the common flu.


Really sorry to hear your stuck.

This virus is a stress test for many aspects of our globalized society. There’s a risk in both overreacting, and under-reacting. Both carry costs.

It’s also a test of how we can all align interests in an age where trust in institutions is being eroded. We’re awash in information, and disinformation — and we’re experiencing in real-time the effects of that.


The virus has impacted hospitals so badly they can’t keep up and they’ve had to turn away patients. China isn’t building new emergency hospitals for flu but is for the Coronavirus, so something is different.


>Personally, I am much more worried about public officials freaking out about the corona virus than I am about the virus itself.

That's understandable given the circumstances you are personally encountering. I hope you are allowed docking somewhere soon, followed by a suitable period of quarantine in comfort, and a confirmation of no infection.


[flagged]


That's unlikely, this isn't the first time nor will it be the last cruise ships will be implicated in the spreading of infectious diseases.

Everybody knows cruise ships are inherently problematic in this regard due to the large numbers of international travelers living in close quarters for an extended period. It doesn't stop millions from using them, for some reason.

https://www.cdc.gov/h1n1flu/guidance/cruiseships.htm

https://www.cdc.gov/nceh/vsp/surv/gilist.htm

https://wwwnc.cdc.gov/travel/yellowbook/2020/travel-by-air-l...


Not surprising considering for a few days there, after control was implemented, the National Health Commission numbers adhered to an exponential trend with an R2 value of 0.995-9.

https://twitter.com/whistle_posse/status/1221964216040509440

An exponential curve is a great way to catch up to the numbers of infected that already exist.

The good news is this means the outbreak is less contagious and uncontrolled than was reported.

Its also probably harder to hide deaths than recoveries, so the mortality rate is likely lower.

The bad news is, people may be infectious for longer before they show symptoms.


"The bad news is, people may be infectious for longer before they show symptoms."

There is evidence that the study claiming the virus can be transmitted by people without symptoms was flawed.[1]

Also, according to that article, even if asymptomatic people transmit the virus "asymptomatic transmission likely plays a minor role in the epidemic overall, WHO says. People who cough or sneeze are more likely to spread the virus, the agency wrote"

[1] - https://www.sciencemag.org/news/2020/02/paper-non-symptomati...


>There is evidence that the study claiming the virus can be transmitted by people without symptoms was flawed.

There's also a recent news story that a UK citizen is a "super spreader", and that he had contact with all the cases in France while being asymptomatic. He even went home and felt fine, carrying on his normal activities for four days before becoming Ill and being confirmed with the virus.


Of course people who cough and sneeze are more likely to transmit it! I can’t trust the WHO when they make what to me seem to me to be intentionally misleading statements.


>Its also probably harder to hide deaths than recoveries, so the mortality rate is likely lower.

1) you obviously mean Case Fatality Rate, instead of mortality rate, since the latter is independent of number of recoveries when keeping deaths fixed (unhideable).

2) What would China's motive be to hide recovered infection counts with the result of reporting a higher than truthful case fatality rate? Do you actually believe China wants local panic and fear, or that it actually wants to loose 2% annual productivity per week in quarantine?


Cool username!

1) Sure! I don't think that really changes my points though. 2) That's only true if the rate correlates with panic, and if the Chinese government knows that would be the effect. They may be sacrificing panic, over a higher case fatality rate, in order to retain an illusion of competence in identifying and addressing the outbreak. I don't think it'd be out of character for an authoritarian government to care about appearances in that sort of way.


A pizza a day keeps the doctor away


The good news is this means the outbreak is less contagious and uncontrolled than was reported.

How do you figure? If you're juicing the numbers presumably you're doing so to make things look better than they are. An optimistic curve is going to catch up to reality only if reality stays the same (or improves). About the only thing we know is that the data from China are suspect.


Maybe they juiced the numbers to make it appear like the epidemic was discovered promptly, rather than late.


Maybe hiding the deaths has been working well: https://www.dailymail.co.uk/news/article-7969861/Wuhan-crema...

If they are requesting 100 body bags per day how does that’s square with the official death count?


Wuhan is a city of 11 million people. New York City has a population of about 8 million people. In 2010 NYC saw an average of 144 deaths per day (Table P3). By those numbers you'd expect to see around 200 deaths daily in Wuhan without any added pressure from a pandemic.

http://www.nyc.gov/html/records/pdf/govpub/6551as_2010_final...


100 body bags per crematorium, 49 crematoriums in Wuhan, supposedly 11 working around the clock.


Which I guess would put deaths at nearer 2000 a day than the official 100 or so.


Daily mail? I'm going to need some more sources for whatever it is they're claiming.


That's why we would all be sleeping easier in our bed if the Chinese would just accept externla help from the USA/WHO/anyone. So far as I understand, there are currently no external observers. There are no reliable sources. Daily Mail is currently no more or less reliable than any other source currently.


I think “bad information from China => daily mail as good as anything else” is a pretty cynical view. I don’t think it’s true in this case


TLDR: A research by Uni Toronto suggests that the coronavirus epidemic started one month earlier than is commonly reported (Nov 2019) and has yet to be brought under control.

David Fisman, one of the model’s creators said: “You can’t get up to that level of cases if the epidemic started in December. even if you pushed the reproduction really high. The plausible start date seems to be mid-November,”

Wikipedia lists[1] start date to be Dec 1, 2019 and with Fever, cough, shortness of breath etc as symptoms[2], I think it's difficult to predict even for doctors exact date of the outbreak.

[1] https://en.wikipedia.org/wiki/2019%E2%80%9320_Wuhan_coronavi...

[2] https://thestrife.co/wuhan-coronavirus-resources


Yeah, running a regression back to the Y intercept to get a 0 time is nice in retrospect, but you'll never find the index patient, the cell or origin, etc.


Playing with their model as linked gives a much better match with reported numbers with a shorter serial interval, 5 or 6 days instead of the 7 assumed by the default model settings. This is also a better match with the more recently reported case studies.

I find it a little irresponsible to make such an extreme claim from such a limited modelling tool... Small changes to the assumptions can account for the discrepancy seen, and it's unrealistic to expect a basic model like this to accurately represent reality in such an unusual situation as in Wuhan.


There was an article in China News Weekly that investigated the origins of the Virus and also pointed out that 1) the first cases were in November and 2) that the first cases were not linked to the Huanan Seafood Market.

https://chinachange.org/2020/02/09/the-regret-of-wuhan-how-c...


If the blue dots represent the data and if the yellow line represents the model -- doesn't the picture show an obviously bad fit?


I'm in China and it's astonishing how different out lives changed with less than 10 days.

Quarantine and lockdown came so quick a lot of people still haven't got used to it.

I've made a video about it and you can see it for yourself: https://youtu.be/uxW7k5462Ms

This is something I've never experienced before.


How many coronavirus articles are you going to post this to?

https://news.ycombinator.com/item?id=22294779


Remember ~15% of common cold cases are caused by a coronavirus, one that has evolved so as not to kill it's host .... there's a good chance it started out a lot like the current coronavirus


Oh good so the German scientist who visited for a few days last week, who was working in Wuhan in December and caught a pathogen (“but it wasn’t the flu”) may have been carrying coronavirus?


If he's been symptom free for 3+ days then no.


dont you mean 24 days? https://www.medrxiv.org/content/10.1101/2020.02.06.20020974v...

"The median incubation period was 3.0 days (range, 0 to 24.0 days)."


I read it that the person got sick after the visit to Wuhan but recovered from that before the visit to OP. Hence 3+ days should be enough to rule out a possible infection of OP with whatever the scientist had.

Since they traveled to Wuhan in December, from then to last week, that would've been more than 24 days, so in case they did not get sick yet and I read that wrong, it should also be pretty much impossible, since that would be the longest incubation period of ncov by quite a bit.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: