After you have had a chance to go over the Lancet study that was released, I'd appreciate hearing if it has changed your position on the neutrality of administering CQ and/or HCQ.
I haven’t dug too deep into it, but on first glance of peeking at the summary, my biggest concern was that this, like other analyses, is still largely observational and naturalistic (as any will be at this point), which makes it harder to truly control for underlying severity differences between the groups. That said, the actual body of the article expands on this. One thing to note is the nature of this being data-mined from EMRs, rather than being from specific treatment protocol (again, that’s not really possible, and in reality unethical at this point, so this is not a surprise and should be expected, but also has to be considered). So we do have the division of people into groups based not on randomization but on clinical decision, so it’s hard to account for what drove those clinical decisions. They do a pretty good job, however, of catching a number of data points that helps paint a generalized clinical picture of who’s getting treatment and who isn’t. Based on qualitative (but not quantitative) criteria regarding comorbidities, all groups are quite similar. The data points they use regarding illness severity is an initial (presumably, at least) qSOFA (risk stratification relating to mortality and sepsis/infection in the hospital) and oxygen saturation. Those variables don’t show any significant group differences, either. So while it is possible that an argument can still be made that the sicker patients were the ones getting treatment, it is hard to say that none of that was captured in part by any of those measures. There are other (clinical) indicators that wouldn’t be captured by those measures, but it would be hard to not see a lot of the above being corollaries to that. One important one would be the presence of an altered baseline of mental status (which one could argue is captured by the qSOFA encompassing a Glasgow Coma Score, but ultimately doesn’t touch the nuances of delirium), as people who are high-risk would fit into a health and age demographic where altered mental status is a more sensitive criteria for underlying poor general health condition, even in the absence more concrete identifiers.
That said, it’s important to keep context of what thresholds of evidence we’re seeking and for what purpose. COVID-19 wasn’t on any radar 6 months ago. Our knowledge of it and potential treatments is limited. So there’s a lower threshold of evidence required to justify treatment decisions, and thus the hurdle that would be the required burden of proof to cross regarding altering practice of conventional approaches becomes much smaller. The argument largely regarding hydroxychloroquine is that there was some observed level of benefit clinically and anecdotally (I’ll ignore potential mechanisms of action as justification because [in my strong bias] those are significantly misleading to clinical application) that potentially could yet to be adequately captured by more rigorous observation. But I do believe that, though there can still be a number of unaccounted for variables that may change the equation, there’s enough data here that it’s hard to justify that there’s a benefit here that’s only capturable in the “advanced analytics,” if you will, but is large enough to be seen by the “eyeball test.” The difference in mortality is certainly interesting. I think people could make arguments regarding different ideas, like “hey, this isn’t hydroxychloroquine with zinc,” and that’s true. Perhaps there is something there, but the only thing with that is an appeal to theoretical mechanisms of action, of which I’ve already expressed my exhaustion for (universally, not this specific instance) and also was not part of the original impetus for the use of hydroxychloroquine (though it had been discussed for quite some time with zinc, though I’m not aware of zinc actually being used with it in any widespread fashion).
Tl;dr if there are benefits that have yet to be captured by what we’ve seen in any published literature, the odds of them being clinically meaningful would be significantly smaller, and as more has unfolded that changes the equation on ‘potential benefits,’ we haven’t seen the dire picture perceived (rationing of ventilators, healthcare system beyond capacity nationally), so the urgency with which interventions can or should be implemented at lower evidence thresholds has also significantly diminished.