Thread by @DrJayVarma

In this OpEd (nytimes.com/2021/12/20/opi…), I argued that we should reduce the recommended isolation period for #COVID19 in vaccinated persons. This is how I think through a problem like this. 🧵⬇️

Call it #ThinkLikeAPublicHealthPractitioner, which is different than thinking like a virologist, epidemiologist, or clinician.

Virology helps us answer: how infectious are people with COVID-19 based on duration of infection, vaccination status, symptoms, and other factors?

Epidemiology helps us answer: what is the impact on community disease transmission for different isolation policies and different levels of adherence to those policies?

Public health helps us answer: if we start with the science (epidemiology, virology), how do we balance real-world harms, benefits, feasibility, acceptability, and costs? With Omicron, the virology and epidemiology have changed, so we need to re-think the tradeoffs.

We have data from Delta and other variants (not #Omicron ) showing that vaccinated people with COVID infections are infectious for a shorter duration of time AND the virus that is detectable later in the course of infection may not be as infectious (replication competent)

HARMS: As large numbers of people become infected, the harms from 10 day isolation are greater: more person-days missed from routine life and work.

BENEFITS: As large numbers of people become infected, the benefit of 10 day isolation is less. With so many infected people in the age of Omicron, removing one infected person from the general population will have less benefit than pre-Omicron.

FEASIBLE: Shortening isolation is also now more feasible, because we have tools now available to help with individual decision making. Antigen tests differentiate reasonably well between “likely not infectious to others” vs. “likely infectious to others.”

So, thinking like a public health practitioner, reducing duration of isolation could reduce harms, increase benefits, and be done feasibly. This is entirely a judgment call (how we start with scientific evidence then run-it through real-world tradeoffs) with lots of uncertainty.

I propose we reduce isolation from current policy (10 days after symptom onset or, if asymptomatic, date of first +ve test) if:
a. Vaccinated person
AND
b. At least 5 days have passed since first +ve test OR symptom onset, whichever is longer
AND...

c. Fever & respiratory symptoms completely resolved for at least 24 hours
AND
d. Negative rapid test on day you are going out

The impetus to make these changes is that Omicron is going to make large numbers of people ill, and many people & employers will be more likely to accept some COVID-disruptions for a longer period if we make those disruptions more manageable.

One big unknown is how many more people will return to routine life with a new policy compared with current policy. If it only applies to a small proportion of cases, then it may not be worth changing.

Another big unknown is whether the assumptions above about duration of infectivity apply to Omicron, and whether it's more prudent to wait until we learn more. I suspect this is where @CDCgov and most jurisdictions will land.

Note that TB is the perfect example of a disease in which we've traditionally used an imperfect test (AFB smears) to answer: When is someone not-so-infectious that it's OK to have them be back in public?