Hacker News Reader

Comment by pcrh | Hacker News Reader

Figure 1A shows 120 cases from 2019 with no known link to the market, yet they cluster about the market. So the allegation of ascertainment bias does not appear to be substantiated.

Further, the criticism of rewording "these analyses provide dispositive evidence" to "these results provide incontrovertible evidence" is bizarre. The term "dispositive" in this context is equivalent to "the last and final word" which in a scientific context was rightfully removed, though it was replaced by a strong assertion.

With these two matters, I did not think the author has sufficient credibility for me to further investigate their claims.

View on HN

Thanks for the feedback.

On the unlinked cases - if you read my medium post, it explains that even the cases with no connections to the market had been identified with ascertainment bias. Local investigators had searched hospitals and the neighborhood near the market for cases even if they had no link to the market.

On your 2nd point, I've clarified that it was not a rewording but a removal of both claims of dispositive and incontrovertible evidence from the preprint. This means that peer review flagged both of these strong assertions in the manuscript and the authors had to remove both of them.

You state below that the requirement for cases to be linked to the market was removed after only a few days. Is it possible that all 120 "unlinked" cases were in fact identified within those few days?

Removal of a claim of "dispositive" evidence isn't weakening the case; there could be further incontrovertible evidence that would make the present evidence no longer dispositive.

By the time the case definition requirement for a link to the market was removed (Jan 18, 2020), nearly 200 cases had been confirmed. This is not including the number of cases that were suspected/clinically diagnosed.

The 174 cases with onset in Dec 2019 considered by Worobey et al. include both confirmed cases and clinically diagnosed cases. Therefore, most, if not all, of these cases had been identified using the criteria that the patients had to either be linked to the market or, if unlinked to the market, have been identified at hospitals near the market or in the neighborhood of the market.

See: https://archive.ph/LVU6z

Re: weakening of their claim, not only did Worobey et al. have to remove these instances of unscientific language, but they also had to insert a new study limitations section and this clarification in their article:

“However, the observation that the preponderance of early cases were linked to the Huanan market does not establish that the pandemic originated there.”

"Study limitations" is standard language in epidemiological papers, so that's not a surprise, nor a condemnation.

Further, to my knowledge it is known that "case zero" (i.e. the zoonotic event) did not occur at the Wuhan market, specifically because there was already sequence divergence in early cases.

However, that doesn't prevent the market from being the main source of the outbreak. There's no evidence for any other source.

Reply on news.ycombinator.com