This is a very confused thread arguing testing isn't perfect and therefore it should only be done where probability is high. But it's fundamentally flawed because you don't know where probability is high unless you test enough and accurately. https://twitter.com/karthiks/status/1250362089585033216?s=20

There are two concepts sensitivity and specificity) that apply to any microbiological test. The RT-PCR will catch around 70% of positives with the appropriate sampling. What that means is if you test positive early on, you are almost certainly positive for SARS-CoV-2.

What this also means is that 30% of positives will not get caught for the disease. This is a limitation of the test. Plain and simple. No one is arguing this point.

But just how this argues for less or limited testing doesn't many any sense logically.

1. 25-50% of those infected with SARS-CoV-2 are asymptomatic but spread the disease.
2. In those who are symptomatic up to 50% of transmission might occur before symptoms show up.
3. In symptomatic people, the most infectious period is before or around the time symptoms show up.

Please read my last tweet on the mode of transmission as known this morning and then read the central point of the original poster who who is a management consultant who provides this rationale: https://twitter.com/karthiks/status/1250362120039878657?s=20

In brief, a pandemic is a dynamic event.

The premise that high-probability people who are identified in the early stages of a pandemic are still the high probability people when 1) the mode of transmission is known and 2) the virus is spreading through a population is silly.

Given the remarkable new information that a major way the novel coronavirus is spread is in the early stages of infection among those who have no symptoms, it is clear that contact tracing alone with not stop the pandemic.

This is a fact based on logic, not on any personal hope

You know the torch works. You also know it is not perfect. Until you can make a better one, you shine the light instead of stumbling in the dark.

Ultimately, it is this light (even with its limitations) which will tell you more than preconceived notions.

What this pandemic has taught me already is that context-specific information based on current information is important.

It is presumptuous and arrogant of people in other fields to come in and think that they're the only ones who have taken courses in statistics.