Thread on some details of @freopp plan:
Assumptions being made:
1. Repeatedly testing the entire population may require 22 million tests/day - Paul Romer
2. Antibodies may not equal immunity
3. Antibody tests need to be highly specific (>99.5%) to be applicable broadly https://twitter.com/anish_koka/status/1252737913869209601

4. Anti-virals may take time to develop
5. A vaccine may take time to develop (The fastest vaccine Ebola took 5 years - Typically need very large RCTs to prove efficacy AND safety when giving a therapy to a healthy population)

Plan leans heavily on differential risk of COVID across ages, comorbidities https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext?utm_campaign=tlcoronavirus20&utm_source=twitter&utm_medium=social

Differential in US deaths by age is startling

Remarkably few deaths in patients with no underlying conditions (NYC Data from 4.20)

Plan highlights-
1. Reopen Schools to children who don't live with/come into contact with high risk individuals (remote schooling for those with high risk family contacts)
2. Lift stay-at-home for < 65. Risk assessment for < 65 with comorbidities

3. Incentivize employers to deploy RT-PCR @ work. ($300/test tax credit)
4. No large group gatherings, High risk locations (bars, nightclubs, gyms) closed until Ro < 1
5. Require -ve COVID test for air/train travel
6. Grant safe harbor for States that restrict travel

Issues:
1. Unable to reduce Ro -> numerator overwhelms health system -> forces economy to shut down again..
2. Deploying testing infrastructure to all places of work about as easy as ramping up to millions of tests/day (not easy)

3. Contact tracing/ App based protocols work in theory. (Did not work in Singapore after 200k Singaporeans returned. )