Thread by @yoongkhean

Seems one of the debate last few days is whether factory clusters r causing the high C19 numbers in S’gor or it’s mostly community cases now. My thoughts in this thread & will try to correct some misconceptions too.

Also, because @nadsjalil ask to. She ask, u dont say no 😂

1st of all, categorising betw factory vs sporadic cases is wrong. Sporadic ≠ community cases. Sporadic simply means unlinked, where the source of infection cant be determined. So sporadic cases could mean from community or from factory clusters.

Only way to know the source of infection for the sporadic cases is thru contact tracing, specifically backward tracing, which the PH system is overwhelmed now & cant trace back in time (more on this later). So we cant exactly know.

But what do we know about factories and C19?

First, we know factory clusters are the dominant in the new clusters. 10/21 of recent clusters are from factories.

And we know factories & living conditions of workers, especially migrant workers are high risk for transmission due to crowding.

https://twitter.com/kkmputrajaya/status/1410566324950953990

Spillover to community is inevitable. I’ve said this since late last year during the earlier peak. If we don’t ring fence it quick enough, factory clusters will act as a super spread event/location & spread to the community.

malaymail.com/news/malaysia/…

YB Dr Afif mentioned the dominant cases now are sporadic (but we established sporadic ≠ community cases). The comparison is inaccurate.

Because we cannot look at current numbers & make a conclusion of the source of infection w/out adequate contact tracing.

https://twitter.com/afifbahardin/status/1410594676646637575

For a disease like C19, we know transmissibility is uneven & our ?obsession w R0/Rt is pointing us at the wrong direction.

R is the potential transmissibility & it’s the average of the total infections. Consider this 2 situations of a R0 of 1 (1 person will infect 1 more person)

1. In a group of 100 infected person, R0 of 1 means 100 will infect 1 each = another 100 infected.

But,

2. It could be 1 person infecting 100 person while 99 dont.

Both are R0 =1 , but implications are different.

And from what we observe, situation 2 (or similar) is more likely than 1. It’s almost always a super spreader event/location or environments like close, crowded areas which becomes a cluster & spreads.

This is why factory clusters will seed the infection elsewhere.

That is why we need to backward trace, to find source of infection which lead to = clusters. Forward tracing will give u the R0, but b/ward tracing will give u the K, dispersion factor, which we know is more important to assess spread.

But b/ward tracing is resource intensive.

Point is, if u want to find the source of infection, look for dispersion factor. If u know where are your super spreader events/locations, u can deploy better/targeted PH measures.

Because that’s how C19 spreads.

Relying on R0/Rt to implement lockdowns wont close the spread fast enough. U dont use a number based on the average potential transmissibility to stop spread. Unless u lockdown forever. Which we know what are the costs.

/end.