5 ms·

Here is some chart showing vaccination effect on hospitalization for severe forms. Very telling. https://ourworldindata.org/grapher/israel-severe-hospitalizati

by lucd 6y ago

Here is some chart showing vaccination effect on hospitalization for severe forms. Very telling. https://ourworldindata.org/grapher/israel-severe-hospitalizations-age?tab=chart&stackMode=relative&time=2021-01-08..latest&country=Early%20vaccinated-0~Late%20vaccinated-0~Late%20vaccinated-1&region=World https://ourworldindata.org/grapher/israel-severe-hospitaliza...

EvgeniyZh 6y ago
I'd expect that case fatality rate would significantly reduce after 60+ population (which is 90%+ of deaths) is vaccinated. However, even though more than 80% of 60+ got second dose already, case fatality rate is approximately the same. I don't have any good explanation for that.
- corty 6y ago
  Death usually comes 2 to 4 weeks after hospitalization which comes 1 to 2 weeks after infection. So the drop in death rates should be visible 2 to 4 weeks after the drop in hospitalizations, and should be less steep initially.
  - EvgeniyZh 6y ago
    There was large amount of 60+ vaccinated with second dose 2 to 4 weeks ago. Iirc it was at least 50%. If half of population responsible for 90% of deaths is vaccinated it should affect death-per-case ratio.
    - corty 6y ago
      Vaccinated with a second dose means full immunity 1 week after, possible death 3 to 7 weeks after (1 + (1 to 2) + (2 to 4)). So a decline should start to be visible since last week maybe. But I wouldn't get nervous for another 2 to 3 weeks in case it doesn't fall sharply, data is always noisy and everything is spread out the more dependencies on previous events and times there are.
      - EvgeniyZh 6y ago
        We don't need full immunity to see the effect tho.
        corty 6y ago
        Right. But still, don't worry yet, worry when it hasn't shown improvement in a few weeks. Edit: Also, when you look at the JHU data (e.g. via google), it shows a peak at Jan 15th in the number of cases followed by a steady decline. The number of deaths peaks on Jan 28th, followed by a similar, steady decline. Which is as expected, so I would say we are already seeing the effects on the number of deaths. Sorry that I didn't look at the data earlier before replying.
        EvgeniyZh 6y ago
        I'm looking on case fatality rate: cases and death count are too volatile and depend on many factors. My hypothesis (it's not mine but I agree with it): if population with high death risks is disproportionately vaccinated, at the level it can make visible effect on case count for this population, it should have effect on case fatality rate. Of course, multiple factors can reduce/slow down the decline, but it should be there. Reducing death probability by half for the 60+ people without changing anything else should significantly reduce case fatality rate. Yet the data does not support it: https://ourworldindata.org/coronavirus-data-explorer?zoomToSelection=true&minPopulationFilter=1000000&time=2020-03-02..2021-11-22&country=~ISR&region=World&cfrMetric=true&interval=smoothed&aligned=true&perCapita=true&smoothing=7&pickerMetric=location&pickerSort=asc https://ourworldindata.org/coronavirus-data-explorer?zoomToS... The case fatality rate peaked at the end of November, slightly declined until mid-January, then grown a bit and is almost constant in last three weeks. So, the options I see are: either no effect at all can be seen yet behind the noise (it's hard to believe for me), or there is some factor compensating for the case fatality rate reduction (I can think of what could increase fatality that much and exactly compensate the effect), or the hypothesis is wrong (I can't see why either)
        corty 6y ago
  - Plasmoid2000ad 6y ago
    Good points, and Vaccination itself has a lag effect as well, If i understand it correctly. I've heard that someone vaccinated with 2nd dose today will not have the protection percentage for 10days+ or so - Which needs to taken into account with these figures. The vaccination also does no good if you are already infected, and with the spike in infections in Israel coinciding with vaccination ramp up, those will be tricky to separate out.
- dennis_jeeves 6y ago
  >I don't have any good explanation for that. A good explanation is that the vaccine is not working. ( people are in denial) I looked at the Pfizer numbers and my conclusion is the it's doing nothing.
  - pault 6y ago
    It will take 6-8 weeks after large scale vaccination to start seeing the results. Deaths usually occur six weeks after infection, and it takes a week to reach full immunity after vaccination.
- runako 6y ago
  > case fatality rate is approximately the same If you vaccinate more people, you would expect that the cases that present at a physician will be selected to be the most severe. If the vaccine reduces most infections to very mild or asymptomatic cases, they will not be counted in the statistics at all. So it would make sense that introducing a broadly effective vaccine would increase the case fatality rate. (You would also expect fewer severe cases, which is the whole point.) A functioning vaccine should affect the CFR by changing the denominator of measured cases. I am reminded of an analogy from improvements in battlefield medicine. As battlefield protocols (on-site treatment, rapid evacuation, etc.) have become more effective, battle fatality rates have fallen. But they have been replaced by a rise in severe chronic injuries like amputations.
  - EvgeniyZh 6y ago
    > If you vaccinate more people, you would expect that the cases that present at a physician will be selected to be the most severe. Most of cases both before and after vaccinations are not severe. Moreover, the share of severe cases seems to stay the same or even increase ( https://datadashboard.health.gov.il/COVID-19/general https://datadashboard.health.gov.il/COVID-19/general , I hope it can be google-translated or something). As I explained in other reply ( https://news.ycombinator.com/item?id=26142482 https://news.ycombinator.com/item?id=26142482 ), I expected change in fatality rate due to disproportional vaccination of the group with most of the fatal cases. For 60+ fatality rate is very high, and many of them are vaccinated. For everyone else, the opposite. I agree that if only severe cases were registered, we should've been looking at case number instead (which would be more stable since almost all of those would be registered).
    - runako 6y ago
      > the share of severe cases seems to stay the same or even increase I can't read the dashboard, but as I indicated this is what you would expect if you have a vaccine that broadly works at controlling the most severe forms of the disease. Vaccinated people whose infections manifest as nothing or a day or two of lethargy are not going to get counted in the statistics. > I expected change in fatality rate your expected change in the fatality rate needs to take into account that the real-world denominator has changed and that will not be apparent if you just divide number of fatalities by number of cases. (Because vaccinated people may be more likely to be asymptomatic, and we expect them to not get sick enough to ever present as a case.) It's important to look at the number of people hospitalized or dying.
- whiddershins 6y ago
  Most vaccinated people will not test positive. This reduces the number of cases.
ImaCake 6y ago
This is a really uplifting chart. These trends should continue as vaccination rates increase even more.
- call_me_dana 6y ago
  Not for anyone aged 0-60. It looks like there’s a giant upward explosion on hospitalizations for this age group. Or am I reading this wrong?
throwaway_Ie8th 6y ago
The chart shows that patients below age 60 have been hospitalized 80% more for severe Covid after vaccination. Hard to not comment this cynically. That is bad. Early vaccinated patients over 60 have 40% improvement. Keep in mind that the group older 60 is smaller in size compared to under 60, and the effect is smaller that adverse response in younger. Net negative.
- yokaze 6y ago
  That is hospitalisation in _regions_ where vaccination started early or late. It does not explicitly discriminate between vaccinated people or non-vaccinated, only possibly by proxy of age. People under the age of 35 were previously not included in the vaccination scheme. At the same time, the two more infectious variants are spreading.
it_robot 6y ago
ehhhh im not sure how you are reading this but the chart looks quite bad
- graeme 6y ago
  Remember that hospitalizations take at least 2-3 weeks post infection to occur, and that you don’t see protection from the first dose for at least five days or so. So you’d expect hospitalizations to lag 18-26 days or so. Also the chart is not saying all those groups had vaccines: those under 60 largely haven’t been. The labels could have been clearer. It’s just a chart of change in hospitalization by age from a starting date.
  - throwaway_Ie8th 6y ago
    You are saying "Late vaccinated 0-59 years old" should be read as "Unvaccinated and late vaccinated 0-59 years old"? Why didn't they use that label then?
    - nicoburns 6y ago
      Because they have been vaccinated. The vaccine just (likely) hasn't taken affect and started affecting the statistics yet.
      - RobinL 6y ago
        I'm not sure that is true. As I understand it, it's a cohort of people of that age who live in a geographical area of Israel where vaccination was started early. So it includes people of that age who declined the vaccine/didn't get it fr other reasons. Of course, you're also correct that another important factor is that some of the cohort are not yet fully protected. The definitions are here: https://www.medrxiv.org/content/10.1101/2021.02.08.21251325v1 https://www.medrxiv.org/content/10.1101/2021.02.08.21251325v... Quote: In order to distill the possible effect of the vaccinations from other factors, including a third lockdown imposed in Israel on January 2021, we compared the time-dependent changes in number of COVID-19 cases and hospitalizations between (1) individuals aged 60 years and older, eligible to receive the vaccine earlier and younger individuals (0-59 years old); (2) early-vaccinated cities compared to late-vaccinated cities; (3) early-vaccinated geographical statistical areas (GSAs) compared to late-vaccinated GSAs;
ad404b8a372f2b9 6y ago
I don't get it, the chart goes down after the lockdown starts, like they always have in the past year. How does it say anything about the effects of vaccination? edit: nvm a different chart is shown when javascript is not activated

by lucd 6y ago

EvgeniyZh 6y ago

I'd expect that case fatality rate would significantly reduce after 60+ population (which is 90%+ of deaths) is vaccinated. However, even though more than 80% of 60+ got second dose already, case fatality rate is approximately the same. I don't have any good explanation for that.

corty 6y ago
Death usually comes 2 to 4 weeks after hospitalization which comes 1 to 2 weeks after infection. So the drop in death rates should be visible 2 to 4 weeks after the drop in hospitalizations, and should be less steep initially.
- EvgeniyZh 6y ago
  There was large amount of 60+ vaccinated with second dose 2 to 4 weeks ago. Iirc it was at least 50%. If half of population responsible for 90% of deaths is vaccinated it should affect death-per-case ratio.
  - corty 6y ago
    Vaccinated with a second dose means full immunity 1 week after, possible death 3 to 7 weeks after (1 + (1 to 2) + (2 to 4)). So a decline should start to be visible since last week maybe. But I wouldn't get nervous for another 2 to 3 weeks in case it doesn't fall sharply, data is always noisy and everything is spread out the more dependencies on previous events and times there are.
    - EvgeniyZh 6y ago
      We don't need full immunity to see the effect tho.
      - corty 6y ago
        Right. But still, don't worry yet, worry when it hasn't shown improvement in a few weeks. Edit: Also, when you look at the JHU data (e.g. via google), it shows a peak at Jan 15th in the number of cases followed by a steady decline. The number of deaths peaks on Jan 28th, followed by a similar, steady decline. Which is as expected, so I would say we are already seeing the effects on the number of deaths. Sorry that I didn't look at the data earlier before replying.
        EvgeniyZh 6y ago
        I'm looking on case fatality rate: cases and death count are too volatile and depend on many factors. My hypothesis (it's not mine but I agree with it): if population with high death risks is disproportionately vaccinated, at the level it can make visible effect on case count for this population, it should have effect on case fatality rate. Of course, multiple factors can reduce/slow down the decline, but it should be there. Reducing death probability by half for the 60+ people without changing anything else should significantly reduce case fatality rate. Yet the data does not support it: https://ourworldindata.org/coronavirus-data-explorer?zoomToSelection=true&minPopulationFilter=1000000&time=2020-03-02..2021-11-22&country=~ISR&region=World&cfrMetric=true&interval=smoothed&aligned=true&perCapita=true&smoothing=7&pickerMetric=location&pickerSort=asc https://ourworldindata.org/coronavirus-data-explorer?zoomToS... The case fatality rate peaked at the end of November, slightly declined until mid-January, then grown a bit and is almost constant in last three weeks. So, the options I see are: either no effect at all can be seen yet behind the noise (it's hard to believe for me), or there is some factor compensating for the case fatality rate reduction (I can think of what could increase fatality that much and exactly compensate the effect), or the hypothesis is wrong (I can't see why either)
        corty 6y ago
- Plasmoid2000ad 6y ago
  Good points, and Vaccination itself has a lag effect as well, If i understand it correctly. I've heard that someone vaccinated with 2nd dose today will not have the protection percentage for 10days+ or so - Which needs to taken into account with these figures. The vaccination also does no good if you are already infected, and with the spike in infections in Israel coinciding with vaccination ramp up, those will be tricky to separate out.
dennis_jeeves 6y ago
>I don't have any good explanation for that. A good explanation is that the vaccine is not working. ( people are in denial) I looked at the Pfizer numbers and my conclusion is the it's doing nothing.
- pault 6y ago
  It will take 6-8 weeks after large scale vaccination to start seeing the results. Deaths usually occur six weeks after infection, and it takes a week to reach full immunity after vaccination.
runako 6y ago
> case fatality rate is approximately the same If you vaccinate more people, you would expect that the cases that present at a physician will be selected to be the most severe. If the vaccine reduces most infections to very mild or asymptomatic cases, they will not be counted in the statistics at all. So it would make sense that introducing a broadly effective vaccine would increase the case fatality rate. (You would also expect fewer severe cases, which is the whole point.) A functioning vaccine should affect the CFR by changing the denominator of measured cases. I am reminded of an analogy from improvements in battlefield medicine. As battlefield protocols (on-site treatment, rapid evacuation, etc.) have become more effective, battle fatality rates have fallen. But they have been replaced by a rise in severe chronic injuries like amputations.
- EvgeniyZh 6y ago
  > If you vaccinate more people, you would expect that the cases that present at a physician will be selected to be the most severe. Most of cases both before and after vaccinations are not severe. Moreover, the share of severe cases seems to stay the same or even increase ( https://datadashboard.health.gov.il/COVID-19/general https://datadashboard.health.gov.il/COVID-19/general , I hope it can be google-translated or something). As I explained in other reply ( https://news.ycombinator.com/item?id=26142482 https://news.ycombinator.com/item?id=26142482 ), I expected change in fatality rate due to disproportional vaccination of the group with most of the fatal cases. For 60+ fatality rate is very high, and many of them are vaccinated. For everyone else, the opposite. I agree that if only severe cases were registered, we should've been looking at case number instead (which would be more stable since almost all of those would be registered).
  - runako 6y ago
    > the share of severe cases seems to stay the same or even increase I can't read the dashboard, but as I indicated this is what you would expect if you have a vaccine that broadly works at controlling the most severe forms of the disease. Vaccinated people whose infections manifest as nothing or a day or two of lethargy are not going to get counted in the statistics. > I expected change in fatality rate your expected change in the fatality rate needs to take into account that the real-world denominator has changed and that will not be apparent if you just divide number of fatalities by number of cases. (Because vaccinated people may be more likely to be asymptomatic, and we expect them to not get sick enough to ever present as a case.) It's important to look at the number of people hospitalized or dying.
whiddershins 6y ago
Most vaccinated people will not test positive. This reduces the number of cases.

ImaCake 6y ago

This is a really uplifting chart. These trends should continue as vaccination rates increase even more.

call_me_dana 6y ago
Not for anyone aged 0-60. It looks like there’s a giant upward explosion on hospitalizations for this age group. Or am I reading this wrong?

throwaway_Ie8th 6y ago

The chart shows that patients below age 60 have been hospitalized 80% more for severe Covid after vaccination. Hard to not comment this cynically. That is bad. Early vaccinated patients over 60 have 40% improvement. Keep in mind that the group older 60 is smaller in size compared to under 60, and the effect is smaller that adverse response in younger. Net negative.

yokaze 6y ago
That is hospitalisation in _regions_ where vaccination started early or late. It does not explicitly discriminate between vaccinated people or non-vaccinated, only possibly by proxy of age. People under the age of 35 were previously not included in the vaccination scheme. At the same time, the two more infectious variants are spreading.

it_robot 6y ago

ehhhh im not sure how you are reading this but the chart looks quite bad

graeme 6y ago
Remember that hospitalizations take at least 2-3 weeks post infection to occur, and that you don’t see protection from the first dose for at least five days or so. So you’d expect hospitalizations to lag 18-26 days or so. Also the chart is not saying all those groups had vaccines: those under 60 largely haven’t been. The labels could have been clearer. It’s just a chart of change in hospitalization by age from a starting date.
- throwaway_Ie8th 6y ago
  You are saying "Late vaccinated 0-59 years old" should be read as "Unvaccinated and late vaccinated 0-59 years old"? Why didn't they use that label then?
  - nicoburns 6y ago
    Because they have been vaccinated. The vaccine just (likely) hasn't taken affect and started affecting the statistics yet.
    - RobinL 6y ago
      I'm not sure that is true. As I understand it, it's a cohort of people of that age who live in a geographical area of Israel where vaccination was started early. So it includes people of that age who declined the vaccine/didn't get it fr other reasons. Of course, you're also correct that another important factor is that some of the cohort are not yet fully protected. The definitions are here: https://www.medrxiv.org/content/10.1101/2021.02.08.21251325v1 https://www.medrxiv.org/content/10.1101/2021.02.08.21251325v... Quote: In order to distill the possible effect of the vaccinations from other factors, including a third lockdown imposed in Israel on January 2021, we compared the time-dependent changes in number of COVID-19 cases and hospitalizations between (1) individuals aged 60 years and older, eligible to receive the vaccine earlier and younger individuals (0-59 years old); (2) early-vaccinated cities compared to late-vaccinated cities; (3) early-vaccinated geographical statistical areas (GSAs) compared to late-vaccinated GSAs;

ad404b8a372f2b9 6y ago

I don't get it, the chart goes down after the lockdown starts, like they always have in the past year. How does it say anything about the effects of vaccination? edit: nvm a different chart is shown when javascript is not activated