COVID-19 Resources
Disclaimer: I am a computer scientist, not an epidemiologist. I am familiar with the math, but might make subject-matter mistakes.
The Big Picture
The death rate is the metric most likely to be accurate, so lets graph it for the world and various countries.
Here is the data from Our World in Data with the death rate per million people. I have here a few countries that I’m interested in, but feel free to add more since the chart is interactive.
The USA is currently (April 24th) at about 6x the global average in infection incidence (150 per million vs 24 per million).
Estimating the Number of Cases
Many western countries are not testing adequately. In Tomas Pueyo’s article Coronavirus: Why You Must Act Now he discusses a way to estimate the number of infections given the numbers of deaths.
Before hospitals are overwhelmed, we can assume the death rate is around 1%. We can get this rate by looking at statistics from places that have extensive testing and contact tracing such as China and South Korea. This means that infections on a certain date are 100x more than the resulting deaths.
We must also take into the account the latency from infection to death. That gives us another factor 50x if the doubling time is 3 days (early March) or 4x if the doubling time is 8 days (late April).
In sum, if we see a certain number of deaths on a certain date, and given the assumptions above, we can estimate 400-5000x actual infections on that date.
Other estimation resources:
- Imperial College of London estimates % infected.
- NEW The Economist on excess deaths and under-counting. England under-counts deaths by 30% and Lombardy and Netherlands by 50%. Belgium and New York are on track at less than 10% under-count.
- NEW The Financial Times updates their coronavirus tracking page, including excess death graphs on a regular bais
Estimating Test Coverage
Given the ratio above, we can estimate how many infections are missed in a region, and estimate the “under-testing factor” on 2020-03-21:
- For the USA, 244 deaths result in an estimate of 244,000 cases, compared to 19,100 actual. So we see an under-testing factor of 13x. This article has a similar expert estimate of under-testing in the USA.
- For Canada, also a factor of 12x
- Italy’s hospitals are overwhelmed, so we use 3% as the death rate, and therefore get an estimate of 1.2 million infected and a 25x under-testing factor.
Strategy
Here are some good articles about strategy (newer and/or more important at top):
- NEW A Roadmap for Reopening by Scott Gottlieb (ex FDA) andothers
- Massive testing and contact tracing is needed. We also need antibody serum tests to verify immunity in previously infected individuals. NEW Tomas Pueyo’s fourth article has a good exposition.
- Everybody should wear masks in public, as this will greatly reduce transmission. How to make your own. Explanatory videos on masks4all.org and masks4all.co. Collected resources.
- Estimates of CFR and IFR (actual fatality risk if infected) from CEBM / Oxford
- The Hammer and the Dance – Tomas Pueyo – suggests a strategy of an initial lockdown (the hammer) and a more relaxed followup (dance) to keep the cases low while treatments and vaccines are developed
- Coronavirus: Why You Must Act Now – Tomas Pueyo
- Weather may have some effect on virus spread, but likely not decisive by itself
- Low dose inoculation (AKA “variolation”) proposed by Robin Hanson
- There appears no trade-off between economic performance and social distancing in the 1918 pandemic – see the “So Which One Is Better, Mitigation or Suppression” section of Tomas Pueyo’s third article
People to Follow
- Bob Wachter from UCSF
- NEW Eric Topol
- NEW Trevor Bedford – researches SARS-COV-2 mutations to see how the pandemic spread (see also https://nextstrain.org/ncov/global advised by Trevor)
- NEW Balaji S. Srinivasan
- NEW Zeynep Tufekci
Tools
- NEW CMMID group global, national and regional transmission factor over time and other epidemiological factors
- NEW rt.live – estimate of the transmission factor (R_t) – if it’s well below 1 then the curve is being crushed. USA only at this point, and no state is crushing it.
- Our World in Data
- USA state-level projection of ICU bed shortage
- Metaculus predictions market pandemic dashboard (and pandemic questions)
- Stanford interactive model for Hammer/Dance dynamics
- Health Weather Map
- Infection Disease Experts – estimation of epidemic parameters
- Local (US county/Canada province level) case counts
- Oura Ring and UCSF team up for early-detection of symptoms. NEW they will supply antibody tests.
- UCSF Citizen Science Study
- Epidemic Calculator
Video conferencing tools are important for remote work. There are some serious concerns out there about Zoom. Consider using Signal, Jitsi, Google Duo or Skype Private mode.
Personal Strategy
Personally, I’m taking the following supplements:
- Vitamin D
- Zinc
- Vitamin C
- Vitamin A
- Selenium
- Multivitamin
- 5-HTP, which is converted in the body into Melatonin
See also the Life Extension page.
I’m also making sure to reduce stress by sleeping enough and through recreational activities.
Other things you should do:
- Wash your hands
- Sanitize items brought in to your home
- Initial viral load may affect outcome, so social distancing may reduce the chance of infection and severity
Potential Treatments
- JCI – Convalescent Serum, also preliminary results published in JAMA, and now in PNAS
- Nature letter – Remdesivir and chloroquine
- Techcrunch – Chloroquine and azithromycin
Risk Factors
Some potential risk factors:
- Smoking (quit now!)
- Blood type other than O
- Being male
- Diabetes, CVD and other chronic conditions
- Immune system problems
- Old age (but perhaps this is mostly due to correlation with chronic conditions)
If you have significant risk factors, consider taking additional steps to reduce your chance of infection.