Real COVID-19 Case Numbers Derived from Death Data, T-cell Misinformation and Nanobodies

T-cells and Immunity

Lots of more mainstream news sources have now latched onto the cross-reactive T-cell results and are misinterpreting them either purposely for an agenda or out of ignorance.

These results have been highlighted and discussed for months in the scientific community, including in updates posted by the author. [1–4]

Nevertheless, here the newest data will be very briefly summarised, as well as what they may mean and what they certainly may not.

New York Magazine published a piece that is one of the best non-technical summaries, though I still think there are important parts of information missing and the speculation goes a bit too far at times. [5]

First, a note on terminology that is particularly important. When talking about T-cells, cross-reactive immunity does not equal protective immunity. To (somewhat inaccurately) simplify, protective immunity describes a case where an individual’s immune system mounts a response such that very few or no symptoms are experienced and the patient is minimally infectious or not at all. Cross-reactive immunity may also be termed cross-reactive immune memory.

Overall, more than five different research groups have identified cross-reactive T-cells in 20–50% of healthy people unexposed to SARS-CoV-2. Mostly these are CD4+ memory T-cells and the epitopes share a high similarity to common human coronaviruses (HCoVs), including those that cause the common cold. [6–10]

This cross-reactivity may help explain the very high variation in disease severity seen for various patients. In other words, patients that are asymptomatic or experience extremely mild symptoms may have cross-reactive T-cells, whereas others require who hospitalisation may not. This is speculative and no data exists for or against this as of yet (18/08/2020). In addition, it is extremely unlikely that cross-reactive T-cells alone would confer protective immunity and render people safe from SARS-CoV-2. Notably, it is also possible that this cross-reactivity is detrimental to people, as it could cause antibody-dependent enhancement or the Hoskins effect; no support for this exists either. [11–13]

Finally, herd immunity, the need for a vaccine and the public health response do not change significantly with the discovery of the T-cell reactivity, because the trajectory of the pandemic has been as it is with this cross-reactivity present in the population. [14]

The reason the scientific community has discussed these results so much is because they can affect the way in which clinical trials are designed, conducted and interpreted, for both vaccines and therapeutics. In addition, they may help to raise alarms for potential detrimental effects (see above).

In the meantime, the research (still) supports that universal adoption of face-masks and social distancing remain the most effective non-pharmaceutical interventions. Protective eye-wear has also been identified as beneficial but has remained a bit more under the radar. It has also been speculated that use of face masks may reduce the acquired viral load in the case of infection and that this may in turn lead to less severe infection; an interesting hypothesis but one not supported by data yet. [15–16]

Deriving Case Numbers from Death Data

The European Center for Disease Prevention and Control (ECDPC) has daily updates on the raw data for worldwide coronavirus cases and deaths. [17]

Using this freely available data, one may attempt a rough estimation of the “real” number of SARS-CoV-2 infections per country, by using death data. The deaths per day divided by 0.007 (best current estimates of IFR/mortality are 0.7% [18]) gives the number of cases per day 14 days prior.

There are several limitations to this approach:
1) it will always be 14 days behind the current date (or whatever time lag one wishes to apply)
2) it assumes accurate death reporting
3) IFR can be quite heteregeneous (addressed in source 18)
4) countries with low numbers of deaths give lower quality data

Despite these limitations, this way of looking at data can provide some very interesting insight into the response of various countries for comparatively very little effort, particularly for the “first wave” where data are much more complete.

An excel macro for automatically extracting data by country, performing the above mentioned calculations and plotting this data into a graph, as well as a file containing data from 16/08 with the macro embedded in a button are included at the end of this post after the references.

Most countries massively underdiagnosed in the initial wave of infections, but have now brought their testing up to par, some examples follow.

Estimated “real” cases of COVID-19 by country derived from official death counts from COVID-19. 7-dma = 7-day moving average. Depicted are UK (top left), USA (top right), Germany (bottom left) and Italy (bottom right).

These estimations show that for example, in the UK, based on deaths per day, the infection rate reached 140,000 people per day. In the US peak infection numbers top 350,000 per day.

By looking at data in such a way it is also possible to pick out certain countries where both the response for containment is poor, as well as testing. Brazil follows as an example of such a country.

Cases per day derived from deaths for Brazil. 7-dma = 7-day moving average.

Readers are encouraged to download the data provided at the end (included are the simple three step instructions on how to use them) to look at countries of interest.

A different IFR may be applied if so desired, though with any realistic value provided (see source 18) the results do not differ drastically.

In the future, an updated macro that normalises all data per capita, compares two or more countries and shows the % discrepancy between reported cases and cases derived from deaths may be included.


A team of researchers in the US have developed nanobodies against SARS-CoV-2 that are extremely potent (below the detection limit of the assays used for measuring binding!). [19]

Derek Lowe at In the Pipeline has an excellent write-up of this research, highly recommended. There, the potential of these nanobodies to be used to create anti-SARS-CoV-2 filters or nasal sprays is discussed and is very interesting! [20]



Data and Macro for Finding Death-Derived Cases

Link to data file with macro embedded:!An2l_O4Ok5Dwg7p5z-Zx46RI1_64Bg?e=qbGjSe

Instructions for use, after downloading and opening the file:
1) Enable macros when prompted
2) Enter the name of the country of choice in cell O2 in sheet “COVID-19-geographic-disbtribution”
3) Press the “Show me the real cases!” button

Script of macro:
Sub Country_fancy()

‘ Country_fancy Macro

‘ Keyboard Shortcut: Ctrl+q

ActiveWindow.ScrollColumn = 9
ActiveWindow.ScrollColumn = 8
ActiveWindow.ScrollColumn = 7
ActiveWindow.ScrollColumn = 6
ActiveWindow.ScrollColumn = 5
ActiveWindow.ScrollColumn = 4
ActiveWindow.ScrollColumn = 3
ActiveWindow.ScrollColumn = 2
ActiveWindow.ScrollColumn = 1
Application.CutCopyMode = False
Application.CutCopyMode = False
Application.CutCopyMode = False
Application.CutCopyMode = False
Application.CutCopyMode = False
Range(“A1:L36613”).AdvancedFilter Action:=xlFilterCopy, CriteriaRange:= _
Range(“O1:O2”), CopyToRange:=Range(“O3:Z3”), Unique:=False
ActiveWindow.ScrollColumn = 2
ActiveWindow.ScrollColumn = 3
ActiveWindow.ScrollColumn = 4
ActiveWindow.ScrollColumn = 5
ActiveWindow.ScrollColumn = 6
ActiveWindow.ScrollColumn = 7
ActiveWindow.ScrollColumn = 8
ActiveWindow.ScrollColumn = 9
ActiveWindow.ScrollColumn = 10
Sheets(“Extraction of data sheet”).Select
ActiveWindow.SmallScroll Down:=15
ActiveWindow.SmallScroll Down:=-219
Columns(“N:N”).ColumnWidth = 16.71
Columns(“M:M”).ColumnWidth = 15.14
Columns(“O:O”).ColumnWidth = 20.57
ActiveWindow.SmallScroll Down:=-207
ActiveSheet.Shapes.AddChart2(240, xlXYScatter).Select
ActiveChart.SetSourceData Source:=Range( _
“‘Extraction of data sheet’!$M$4:$N$219,’Extraction of data sheet’!$O$18:$O$228,’Extraction of data sheet’!$E$18:$E$228,’Extraction of data sheet’!$A$18:$A$228” _
ActiveSheet.Shapes(“Chart 6”).IncrementLeft 435
ActiveSheet.Shapes(“Chart 6”).IncrementTop -147
ActiveWindow.SmallScroll Down:=-6
Application.CutCopyMode = False
Application.CutCopyMode = False
Application.CutCopyMode = False
Application.CutCopyMode = False
ActiveChart.FullSeriesCollection(1).Name = “=””Real cases”””
ActiveChart.FullSeriesCollection(1).Values = _
“=’Extraction of data sheet’!$M$4:$M$219”
ActiveChart.FullSeriesCollection(1).XValues = _
“=’Extraction of data sheet’!$A$18:$A$228”
ActiveChart.FullSeriesCollection(1).XValues = _
“=’Extraction of data sheet’!$A$18:$A$228”
ActiveChart.FullSeriesCollection(2).XValues = _
“=’Extraction of data sheet’!$A$18:$A$228”
ActiveChart.FullSeriesCollection(2).Name = “=””Reported Cases 7dma”””
ActiveChart.FullSeriesCollection(3).XValues = _
“=’Extraction of data sheet’!$A$18:$A$228”
ActiveChart.FullSeriesCollection(3).Name = “=””Reported Cases”””
ActiveChart.FullSeriesCollection(4).XValues = _
“=’Extraction of data sheet’!$A$18:$A$228”
ActiveChart.FullSeriesCollection(4).Values = _
“=’Extraction of data sheet’!$N$4:$N$219”
ActiveChart.FullSeriesCollection(4).Name = “=””Real Cases 7dma”””
ActiveSheet.Shapes(“Chart 6”).IncrementTop -96.75
End Sub