A few final points I want to address:
1️⃣ Missing / non-perfect data does not mean we have no data and can't conclude anything.
2️⃣ Potential bias vs multiple independent data sources.
3️⃣ Is the "Lab Leak" a conspiracy theory and those believing it "conspiracy theorists"? (hint: no, but...).
🧵
On the day Dr. Fauci - who dedicated his life to public health, saving millions - is dragged in front of congress and told he belongs in prison, @nytimes.com pushed pseudoscience that served to further degrade trust in science.
I wrote 2️⃣ threads to highlight many issue-links in 🧵👇 for easy reading.
1️⃣ Missing / non-perfect data does not mean we have no data and can't conclude anything.
This is a common misconception and often brought up - because the data aren't perfect and we don't have _all_ the data, we can't conclude anything until we do.
This is obviously false.
1/
We very rarely have 'perfect' data and there is always missing data - after all, science is a process, not an end goal.
Now, during a rapidly evolving and highly chaotic situation, like the beginning of a pandemic, data will *always* be incomplete and non-perfect.
2/
Literally *anybody* who has worked directly as part of outbreak response will tell you this. Do you think our data were complete and perfect when we went to Sierra Leone in March, 2014? Or when we worked with the DPH in Florida during the Zika epidemic? Or SARS-CoV-2 in San Diego?
Nope. Never.
However, this does not translate to "so we have to wait for such data before making any conclusions". Obviously - and anybody, especially scientists, should recognize this.
Because we *do* have data and as I outlined in my first threads, the data are rich and detailed.
4/
For one, we have home addresses for many of the first detected cases!! I mean, holy hell - that's definitely a first! We also have genomes, isolates, details about source and much more from the very market this all started. We have epi, we have serology, we have excess death data, etc., etc.
5/
We can't just discount all this data because we find the conclusions from them inconvenient or because we WANT better data.
We all want more data, but that's not how this works.
What we have, while incomplete, is already rich.
Further versions of this go something like this: "But we can't trust any of the data because, China".
Can you see why some of us have highlighted the issue about xenophobia creeping into the discourse?
Further, many of the main conclusions can be reached independently using data outside China.
6/
A similar argument: "I believe it started in October, so unless I get data from earlier than December*, I'm going to discount everything we have".
A) That data does't exist, and B) for one, because it didn't start in October.
*a common extension goes "and they're hiding it". Now we're deep.
7/
A final point I want to make is that a lot of the data we have that clearly points directly to the market come from scientists and public health responders during the very early stages of the pandemic. You can find much of this on Pubmed and if you read it, you'll notice that it's unfiltered.
8/
While we always have to consider censorship, which certainly became (and is) prevalent - and not just in China - much of that early data, from independent sources, are raw and unfiltered.
Does this mean that we should just fully trust every single data point?
Of course not, which brings me to:
9/
2️⃣ Potential bias vs multiple independent data sources.
When working with messy data - which 'outbreak' data will always be - we have to be concerned about potential errors and biases.
For example, many of the early SARS-CoV-2 genomes contained informatics errors, which had to be corrected.
10/
In addition to outright errors, potential biases also need to be considered and, for example, during January 2020, multiple different criteria were introduced for case ascertainment, introducing bias.
One needs to consider that during downstream analyses.
11/
In our studies, we do that. First, we assess potential bias, second we use data _prior_ to the bias being introduced, and third, we assess the effect of potential bias.
Most importantly, though, fourth, we use multiple independent data sources to reach the same conclusions.
12/
The latter point is critically important.
For example, the way we collect hospital data is different from the way we collect case data, which is different from how excess death data is collected, which is different from how serological data is collected.
All of these are independent.
13/
Because these are all collected differently, by different people, for different reasons, hence, even if you are concerned about potential bias in one of the data sources, they can't all be affected (in the same way).
And they all point to the market.
14/
And so does the data from inside the market - which, again, is a completely different type and source of data.
For the genomic data - that allow us to estimate timing of the pandemic - you get the same result whether you use data from the market, China, or outside China. Again, independent.
15/
That timing, if you use early epi data, again you get to the same conclusion.
You can't just discount all of these data in one fell swoop by shouting "I wan't more data!", "All the data are biased!!, and "Because, China!!!).
(bonus point - see the separate issue there?).
16/
Which brings me to the final point.
3️⃣ Is the "Lab Leak" a conspiracy theory and those believing it "conspiracy theorists"?
The answer is "no". The idea of a "Lab Leak" itself is not a "conspiracy theory" - in the absence of the evidence base we have today, it was an important hypothesis.
17/
Further, the vast majority of people who believe in the "Lab Leak" are not "conspiracy theorists" - in fact, few are.
However, the vast majority of Lab Leak rhetoric, speculation, and accusations are overwhelmingly conspiracists - including the NYTimes piece.
18/
The reason being that the Lab Leak idea is entirely devoid of actual evidence in support of it being true.
Hence, you see very common conspiracist tropes being constantly pushed:
19/
1️⃣ All the data can be dismissed!
2️⃣ All the experts are conflicted, so you can't trust them!
3️⃣ All the published studies are wrong!
4️⃣ Evidence missing for "zoonosis" is evidence of a Lab Leak!
5️⃣ They're hiding stuff = evidence for Lab Leak!
6️⃣ There's a giant cover-up!
7️⃣ NYTimes published on it!
20/
I get it. To the general reader, the "Lab Leak" certainly seem plausible, but as I pointed out in my thread, the devil is in the detail - but getting to that level of detail requires an awful lot of background knowledge of the underlying topic and data.
Most people don't have that.
21/
And for a very good reason - it's a complex topic, so why would they? And the Lab Leak certainly seems enticing on the surface - I myself was there in the beginning of the pandemic. As in, full on there, but I got around - and because of working with colleagues and expertise, rather quickly.
22/
Which is why it is so incredibly dishonest of @nytimes.com to publish this piece in the first place. They have a responsibility in ensuring that what their are platforming is evidence-based and not conspiratorial pseudoscience.
As I pointed out in my first two threads, they failed at that.
/end
Thank you for pointing this out. It's a reminder that when this topic comes up in casual conversation that the people who believe in the "Lab Leak" might just need to be presented with more information. Can't believe you are still having to do this, but thanks for taking time to do so. Tusind tak!