5 ms·

This is taken as a yearly number but looking at week over week numbers [1] Sweden did have excess deaths in March and April. 1. https://www.nytimes.com/interac

This is taken as a yearly number but looking at week over week numbers [1] Sweden did have excess deaths in March and April. 1. https://www.nytimes.com/interactive/2020/04/21/world/coronavirus-missing-deaths.html https://www.nytimes.com/interactive/2020/04/21/world/coronav...

ggrrhh_ta 6y ago
Other countries already have excess deaths (e.g. US: https://public.tableau.com/views/COVID_excess_mort_withcauses_11112020/WeeklyNumberByAge?:embed=y&:jsdebug=y&:toolbar=n&:tabs=n&:display_count=n&:origin=viz_share_link https://public.tableau.com/views/COVID_excess_mort_withcause...) indicating that not only people have died sooner than they would have (which is irreparable), but more people will have died within this year than in a year without Covid. It could be argued that US & Sweden's policies are not that distant. I guess that US has cities with much higher population density & much more unavoidable commute (as in the higher affected areas in Europe).
arcticbull 6y ago
Sure, that makes sense, although I think the annualized number is quite interesting too as it appears a risk-off approach has led to a commensurate decrease in other fatalities.
- benlivengood 6y ago
  The tail at the end of the reporting period graph looks like lag, not a sudden drop in death rate.
  - arcticbull 6y ago
    Correct me if I'm wrong, is that not just that this year's total is only 10.25 months of data as compared to prior year's full 12 months?
    - DanBC 6y ago
      Don't these death stats have a built in lag because of coroners, medical examiners, and stats authorities?
      - jeofken 6y ago
        I’m not sure, but I’d expect Swedish statistics to update the same day, as we have very little if any record keeping on paper, and databases are centralised here.
        Pinus 6y ago
        I'd expect the same. However, reality does not agree. Apparently, the law/rules/whatever state that a death must be reported within 14 days (I have no source for this). I do not know about deaths in general, but for Covid-19 deaths in that Excel sheet on the Arcgis site, the death count takes about 10 days to stabilize. (Why? My guess would be: because public health care in Sweden is managed by the 21(?) regions, each with their on systems for everything.)
        jeofken 6y ago
        I’d be interested in learning this from the source! Thanks for your reply
        kqr 6y ago
        Digital does not mean instant. There are plenty of batch systems around that process all changes at a regular schedule. A chain of such subsystems (especially when they're not synchronised) can easily delay feedback by days or weeks.
        jeofken 6y ago
        True indeed. Since King Gustav Vasa who indebted the nation to fund his wars and in order to finance this, forever doomed the swedes to suffer centralisation, in order to establish his taxation regime. What I mean with this historical tidbit is I’d expect the Swedish state to keep that data in one database, not different databases around the country. But I have no source, and would love to learn, if anyone working in Swedish healthcare IT is on HN.

ggrrhh_ta 6y ago

Other countries already have excess deaths (e.g. US: https://public.tableau.com/views/COVID_excess_mort_withcauses_11112020/WeeklyNumberByAge?:embed=y&:jsdebug=y&:toolbar=n&:tabs=n&:display_count=n&:origin=viz_share_link https://public.tableau.com/views/COVID_excess_mort_withcause...) indicating that not only people have died sooner than they would have (which is irreparable), but more people will have died within this year than in a year without Covid. It could be argued that US & Sweden's policies are not that distant. I guess that US has cities with much higher population density & much more unavoidable commute (as in the higher affected areas in Europe).

arcticbull 6y ago

Sure, that makes sense, although I think the annualized number is quite interesting too as it appears a risk-off approach has led to a commensurate decrease in other fatalities.

benlivengood 6y ago
The tail at the end of the reporting period graph looks like lag, not a sudden drop in death rate.
- arcticbull 6y ago
  Correct me if I'm wrong, is that not just that this year's total is only 10.25 months of data as compared to prior year's full 12 months?
  - DanBC 6y ago
    Don't these death stats have a built in lag because of coroners, medical examiners, and stats authorities?
    - jeofken 6y ago
      I’m not sure, but I’d expect Swedish statistics to update the same day, as we have very little if any record keeping on paper, and databases are centralised here.
      - Pinus 6y ago
        I'd expect the same. However, reality does not agree. Apparently, the law/rules/whatever state that a death must be reported within 14 days (I have no source for this). I do not know about deaths in general, but for Covid-19 deaths in that Excel sheet on the Arcgis site, the death count takes about 10 days to stabilize. (Why? My guess would be: because public health care in Sweden is managed by the 21(?) regions, each with their on systems for everything.)
        jeofken 6y ago
        I’d be interested in learning this from the source! Thanks for your reply
      - kqr 6y ago
        Digital does not mean instant. There are plenty of batch systems around that process all changes at a regular schedule. A chain of such subsystems (especially when they're not synchronised) can easily delay feedback by days or weeks.
        jeofken 6y ago
        True indeed. Since King Gustav Vasa who indebted the nation to fund his wars and in order to finance this, forever doomed the swedes to suffer centralisation, in order to establish his taxation regime. What I mean with this historical tidbit is I’d expect the Swedish state to keep that data in one database, not different databases around the country. But I have no source, and would love to learn, if anyone working in Swedish healthcare IT is on HN.