MulletApps: COVID19 Stats
TLDRUpFront: Over 10 redesigned applications for COVID19 Stats including Cases, Death Rates, Rate of Change in Cases, Hospitalization Rates, Remaining Hospital Beds, and Remaining ICU Beds by State. For Countries we track Cases, Death Rates, and Rate of Changes in Cases. Users can create custom comparison charts between States or Countries and download as images or data files.
Hello, its me.
FullContextInTheBack:
Bookmark this page! The Mulleteer Data Science Team has revised and updated our popular interactive applications for tracking, understanding, and advocating about COVID19. All charts are interactive allowing users to create their own custom Comparisons and save that as an image or download the data for additional analysis.
US COVID19 Stats
COVID19 Cases by State
Tracks Confirmed Cases (Tested Positive), Estimated Actual Cases (Derived), and Confirmed Deaths for an individual State. Note Y-axis is logarithmic. See notes below for methods and caveats. Source: covidtracking.com Updated: Daily
COVID19 Hospitalization Rates, Remaining Hospital Beds & Remaining ICU Beds by State
Tracks estimated Hospitalization Rate for COVID19, Remaining Hospital Beds, and Remaining ICU Beds for an individual State. Does not include 95% CL. Source: https://www.cdc.gov/nhsn Updated: 12pm EST Tuesday, Wednesday, and Friday.
Compare COVID19 Cases by State
Compares Confirmed Cases (Tested Positive) across up to five States. Note Y-axis is logarithmic. See notes below for methods and caveats. Source: covidtracking.com Updated: Daily .
Compare COVID19 Case Rate of Change by State
Compares the Daily Rate of Change in Confirmed Cases across up to five States. Source: covidtracking.com Updated: Daily
Compare COVID19 Death Rate by State
Compares the Crude Fatality Rate across up to five States. Source: covidtracking.com Updated: Daily
Compare COVID19 Hospitalization by State
Compares estimated Hospitalization Rate for COVID19 Patients across up to five States. Does not include 95% CL. Source: https://www.cdc.gov/nhsn Updated: 12pm EST Tuesday, Wednesday, and Friday.
Compare COVID19 Hospital Beds Remaining by State
Compares estimated Hospital Beds remaining across up to five States. Does not include 95% CL. Source: https://www.cdc.gov/nhsn Updated: 12pm EST Tuesday, Wednesday, and Friday.
Compare COVID19 ICU Beds Remaining by State
Compares estimated ICU Beds remaining across up to five States. Does not include 95% CL. Source: https://www.cdc.gov/nhsn Updated: 12pm EST Tuesday, Wednesday, and Friday.
Global COVID19 Stats
COVID19 Cases by Country
Tracks Confirmed Cases (Tested Positive), Estimated Actual Cases (Derived), and Confirmed Deaths for an individual Country. Note Y-axis is logarithmic. See notes below for methods and caveats. Source: WHO Updated: Daily
Compare COVID19 Cases by Country
Compares Confirmed Cases (Tested Positive) across up to five Countries. Note Y-axis is logarithmic. See notes below for methods and caveats. Source: WHO Updated: Daily .
Compare COVID19 Case Rate of Change by Country
Compares the Daily Rate of Change in Confirmed Cases across up to five Countries. Source: WHO Updated: Daily
Compare COVID19 Death Rate by Country
Methods & Notes
General Use
- At this time the charts are not sized well for mobile – we suggest using a computer or pad device.
- If the chart is flowing into the side bar making it hard to read, adjust the zoom on your web browser. (CTRL and the “-” Key)
- Below the chart is a bar where you can select the entity level (State or Country) you wish to review.
- You can hover over any line to see the value of a point, or download the chart as a *.PNG file to share.
- You can download the chart as a *CSV, these downloads will hold all data for all entities (States of Countries) on that report type so you don’t need to download per entity.
- For our Estimated Actual charts, because the chart auto-calculates new each day based on yesterday’s reported deaths, charts can change appearance from day-to-day.
Primary State/Country Charts
Primary charts for individual States and Country Confirmed Cases, those that have tested positive, Confirmed Deaths, and a calculated estimate of Actual Infected. We’ve known for some time that Confirmed cases is the lower-end of potential actual because of limitations of who gets tested, asymptomatic cases, policies don’t allow for testing, or simply tests aren’t available.
Our method of calculating Actual Infected does not rely on any testing data and may be more ‘accurate’ as to the total infected population in an area than Confirmed, even if it is just an estimate. That method, described below, gives us a good understanding of total infected up to a point 20days ago. At that point we forecast three scenarios which represent the range in which the contagion may have changed between our last good-estimate and today.
Feared is the worst case of continued outbreak, Mid indicates a transition between outbreak and decline, and Hoped represents a decline. These three forecasts give us a range of uncertainty at the current Actual Infected but allows us to compare all three to the current Confirmed via testing. Although every chart varies – three notional patterns emerge that aid in understanding what’s going on.
Forecast calculated actual lines are above confirmed line and the gap is increasing. This indicates inability to conduct sufficient testing and means Confirmed cases may not be a reliable proxy. These conditions occur prior to, during, and just after an uncontained outbreak.
However, the above all implies that a government is accurately reporting COVID-19 deaths. Under-counting, missclassification, and/or trying to “hide” deaths can create behavior modes that do not represent real conditions.
Note that the vertical axis is a logarithmic scale of base ten. This allows us to chart entity totals as they rise from under 10 to over 2M. But don’t misread a line “half-way” up the chart as being halfway to 2M. Halfway up the chart is 1,000 meaning it’s still 2,000x smaller than the value at the top. Charts generally refelct most recent 60-90 days of available data.
How we calculate actual infected in the past on primary charts
We determine Actual Infected by using actual deaths due to COVID-19, which are better tracked, and then making a few assumptions to calculate an estimate of total infected. These assumptions are:
- The time from infection to death on average is ~20 days. (1)
- On average 4% of those who become infected will die based on a crude CFR.
- COVID-19 related deaths are reported reliably by the WHO, CDC and other health organizations.
Note: The second assumption is an -assumption- used for mathematical modeling purposes only!! Actual death rates to the virus vary across countries, age demographics, extent of underlying conditions, access to healthcare, the ability of hospital systems to handle surges of patients, and have ranged from under 1% to over 10%. See our interactive chart to compare death rates by country. Also as the size of the contagion increases, broadening the underlying demographic at risk we are reviewing this rate for adjustment.
With these three assumptions we can calculate an estimate of actual infected. We call the “current” day H-Date and 20 days before that I-Date. If 4 died on H-Date, based on our assumptions we can estimate that 100 people were infected on I-Date, because 4% of 100 is 4. If 40 people died on H-Date, we can estimate that 1,000 people were infected on I-Date. Likewise the running total of deaths at any H-Date implies a similar running-total of actual infected at I-Date. When we update our data each day, we advance the H-Date and the I-Date forward, calculating new totals.
How we calculate Actual Infected up to Today
This gives us a reasonable estimate of Actual Infected, but it’s 20 days in the past. To “grow” that 20day old estimate into an estimate for today, we take a seven-day average of the Rate of Change in Actual Infected up to I-Date, and then apply the three forecasting scenarios to that Rate of Change: Feared, Mid, and Hoped. Each scenario is a 20-day segment of a lifecycle of the contagion, and the behaviors of Rate of Change are illustrated below.
Feared represents a continuing outbreak scenario where daily growth continues to climb until it stabilizes. Mid represents this transition phase, where growth rates have stabilized, and after several days, begin to decline. And the Hoped forecast represents a continuing decline in growth rate.
Sources:
(1) We use 20days, rounding up based on a researched finding of 17.8 days see https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext
Death Rate Charts
Death Rates over time tell us the changing impact of COVID-19 as it moves through Wave 1. A rising or declining Death Rate indicates dynamically changing conditions in the State or Territory. It differs from a Per Capita Measure in focusing just on the deaths within the population who has been Confirmed to have the infection. Given that COVID19 outbreaks are highly localized in clusters this is an important measure to use in context with Per Capita as a State overall may be spared but a few hotspot clusters that overwhelm local services can lead to a sharp rise in deaths. Causes for dynamics observed could be from from strain on the healthcare system, different criteria on who gets tested (and thus adds to Confirmed), and a State’s pandemic response policy. Regardless a rising Death Rate should be taken seriously. Use it with the Confirmed Cases Chart (see below) to understand the size of the population impacted by this Death Rate.
How we calculate Death Rates
The rate is calculated by taking COVID19 Tracker death counts and dividing them by the confirmed cases on that day. The chart is set to only display 0-10% for greater clarity. These charts are adjusted to display only the last 90 days of date. Also to avoid noise the death rate calculation was only applied once a country had passed it’s 20th death related to the infection. Keep in mind the ‘death rate’ is of confirmed cases only and should not be taken as applying to the total population of a State!
As daily testing protocols struggle to keep pace with the pandemic, Confirmed cases are lagging far behind Actual Infected. This may also artificially inflate death rates at a point in time. To understand the potential of this impact look up the Primary Chart for a State or Country in question to compare Confirmed to Actual Infected.
About Author
