3 ms·

The numbers of hospitalizations and deaths unfortunately are quite proportional to the number of positive tests. Depending on the testing regime the exact facto

by _ph_ 6y ago

The numbers of hospitalizations and deaths unfortunately are quite proportional to the number of positive tests. Depending on the testing regime the exact factor varies a bit, but unfortunately, not very much. By now we really have enough international data to show that. And the positive tests are leading hospitalisation and deaths by a couple of weeks. So they are a good indicator about what to expect. And obviously, during an epedemic, raising infection numbers mean even higher numbers following until something reduces further infections.

belorn 6y ago
Yes and no. The numbers of hospitalizations and deaths are proportional to the number of tests, but only if the amount of testing is constant. As linked elsewhere, https://experience.arcgis.com/experience/09f821667ce64bf7be6f9f87457ed9aa https://experience.arcgis.com/experience/09f821667ce64bf7be6..., the numbered of confirmed new positive tests are around 5k compared to April, and the number of deaths are 10-20 compared to April 100-120 per day. What the graphs would need is normalization to the amount of testing being done. Such graphs would then have hospitalizations and deaths proportional to the number of positive tests.
- _ph_ 6y ago
  Perhaps I was a bit short, but the numbers of the badly affected are proportional to the number of positive tests. The proportionality factor does indeed change over time, but very slowly. Unfortunately, we have a plentiful of data internationally, as many countries handled the epedimic differently. But the fundamental data looks very similar between countries. In Germany, there is about a factor of 4 between positive/death in April and in November. The number of people in intense care has reached April levels. But on shorter time scales the variation is much smaller. And if the number of positive tests doubles within a week and nothing fundmental had happened in testing, then this is extremely relevant. Especially if the trend follows an exponential curve closely.
  - belorn 6y ago
    Going by the above graphs, looking at the day of 4/15/2020, we had 604 people tested positive and 115 who died, with a ratio of 5 positive per death. If we look at peak for positive tested, we got 11/5/202 with 4744 people tested positive and 12 people who died, with a ratio of 395 positive per death. If we are going to get a factor of 4 positive per death we would need to see a spike of 1186 deaths within the next month or so for a single day. While anything is possible in this pandemic, I am not that certain of it. > The number of people in intense care has reached April levels. Per the graph, we are not there. April had as it peak 49 new person in a single day. November has its current peak at 20. It could get above 49, but again, time will show if that is the case. It is more alarming than the death graph, but less so than the positive graph. For the number of tests, it has been increased from around 160k to 250k per week in the span of 5 weeks. (https://www.folkhalsomyndigheten.se/smittskydd-beredskap/utbrott/aktuella-utbrott/covid-19/statistik-och-analyser/antalet-testade-for-covid-19/ https://www.folkhalsomyndigheten.se/smittskydd-beredskap/utb...). In comparison, April had around 20k tests per week. Interpretation of data is naturally up for debate but personally I am bit of a middle line person. I don't think we will see over 1k deaths per day any time soon, but we are going to see an quick increase in the near time. A lot will depend if the government decided again to lock up the elderly, something which worked but was quite controversial last time.

by _ph_ 6y ago

belorn 6y ago

Yes and no. The numbers of hospitalizations and deaths are proportional to the number of tests, but only if the amount of testing is constant. As linked elsewhere, https://experience.arcgis.com/experience/09f821667ce64bf7be6f9f87457ed9aa https://experience.arcgis.com/experience/09f821667ce64bf7be6..., the numbered of confirmed new positive tests are around 5k compared to April, and the number of deaths are 10-20 compared to April 100-120 per day. What the graphs would need is normalization to the amount of testing being done. Such graphs would then have hospitalizations and deaths proportional to the number of positive tests.

_ph_ 6y ago
Perhaps I was a bit short, but the numbers of the badly affected are proportional to the number of positive tests. The proportionality factor does indeed change over time, but very slowly. Unfortunately, we have a plentiful of data internationally, as many countries handled the epedimic differently. But the fundamental data looks very similar between countries. In Germany, there is about a factor of 4 between positive/death in April and in November. The number of people in intense care has reached April levels. But on shorter time scales the variation is much smaller. And if the number of positive tests doubles within a week and nothing fundmental had happened in testing, then this is extremely relevant. Especially if the trend follows an exponential curve closely.
- belorn 6y ago
  Going by the above graphs, looking at the day of 4/15/2020, we had 604 people tested positive and 115 who died, with a ratio of 5 positive per death. If we look at peak for positive tested, we got 11/5/202 with 4744 people tested positive and 12 people who died, with a ratio of 395 positive per death. If we are going to get a factor of 4 positive per death we would need to see a spike of 1186 deaths within the next month or so for a single day. While anything is possible in this pandemic, I am not that certain of it. > The number of people in intense care has reached April levels. Per the graph, we are not there. April had as it peak 49 new person in a single day. November has its current peak at 20. It could get above 49, but again, time will show if that is the case. It is more alarming than the death graph, but less so than the positive graph. For the number of tests, it has been increased from around 160k to 250k per week in the span of 5 weeks. (https://www.folkhalsomyndigheten.se/smittskydd-beredskap/utbrott/aktuella-utbrott/covid-19/statistik-och-analyser/antalet-testade-for-covid-19/ https://www.folkhalsomyndigheten.se/smittskydd-beredskap/utb...). In comparison, April had around 20k tests per week. Interpretation of data is naturally up for debate but personally I am bit of a middle line person. I don't think we will see over 1k deaths per day any time soon, but we are going to see an quick increase in the near time. A lot will depend if the government decided again to lock up the elderly, something which worked but was quite controversial last time.