That’s why community peer-review is super important. A fraudulent study might get past journal peer-review but will most likely get caught when such study is examined by the larger community of scientists. Now Ivermectin is back to square 1.
2 Likes
Spookily accurate prediction, Michael.
2 Likes
Hee hee. That comment sure aged well.
Irrelevant to the Raoult protocol
Irrelevant to the Raoult protocol
Given than less than 90 patients were included in each of the 3 arms of the study, it is far from having the required power to tell us anything regarding the effect of HCQ+AZ on mortality, which is really what we want to know. No death has been recorded in the 3 arms!
Are you serious? How can you speak of a large RCT, when less than 90 patients were included in the 3 arms of the study. The truth is that this small RCT is underpowered to inform us about the effect of the Raoult protocol on mortality.
This is wrong. Even if one removes the Elgazzar study from the meta analyse by Bryant & al, the overall findings of a signif- icant mortality advantage in ivermectin treatment, and in prophylaxis, remain robust.
https://covid19criticalcare.com/wp-content/uploads/2021/09/Response-to-Elgazzar.pdf
I don’t know who you are going to convince with such pathetic ad-hominem arguments.
How can you draw conclusions about statistical power when you know that this is not the only RCT that analyzed HcQ + Az? We have already discussed another much larger RCT in this very thread.
But let’s examine two more recent meta-analyses that specifically address the question of HcQ + Az in the treatment of COVID-19, shall we?
Conclusion:
Mortality was not different between the standard care (SC) and HCQ groups (RR = 0.99, 95% CI 0.61-1.59, I2 = 82%), meta-regression analysis proved that mortality was significantly different across the studies from different countries. However, mortality among the HCQ + AZM was significantly higher than among the SC (RR = 1.8, 95% CI 1.19-2.27, I2 = 70%). The duration of hospital stay in days was shorter in the SC in comparison with the HCQ group (standard mean difference = 0.57, 95% CI 0.20-0.94, I2 = 92%), or the HCQ + AZM (standard mean difference = 0.77, 95% CI 0.46-1.08, I2 = 81). Overall VQR [virologic cure rate], and that at days 4, 10, and 14 among patients exposed to HCQ did not differ significantly from the SC [(RR = 0.92, 95% CI 0.69-1.23, I2 = 67%), (RR = 1.11, 95% CI 0.26-4.69, I2 = 85%), (RR = 1.21, 95% CI 0.70-2.01, I2 = 95%), and (RR = 0.98, 95% CI 0.76-1.27, I2 = 85% )] respectively. Exposure to HCQ + AZM did not improve the VQR as well (RR = 3.23, 95% CI 0.70-14.97, I2 = 58%). The need for MV [mechanical ventilation] was not significantly different between the SC and HCQ (RR = 1.5, 95% CI 0.78-2.89, I2 = 81%), or HCQ + AZM (RR = 1.27, 95% CI 0.7-2.13, I2 = 88%). Side effects were more reported in the HCQ group than in the SC (RR = 3.14, 95% CI 1.58-6.24, I2 = 0). Radiological improvement and clinical worsening were not statistically different between HCQ and SC [(RR = 1.11, 95% CI 0.74-1.65, I2 = 45%) and (RR = 1.28, 95% CI 0.33-4.99), I2 = 54%] respectively. Despite the scarcity of published data of good quality, the effectiveness and safety of either HCQ alone or in combination with AZM in treating COVID-19 cannot be assured. [emphases added]
This meta-analysis is even more recent:
Conclusions:
There was no significant effect on mortality associated with AZ plus HCQ (odds ratio [OR] = 0.562 [95% confidence interval {CI}: 0.168-1.887]), AZ alone (OR = 0.965 [95% CI: 0.865-1.077]), or HCQ alone (OR = 1.122 [95% CI: 0.995-1.266]; p = 0.06). Similarly, based on pooled effect sizes derived from direct and indirect evidence, none of the treatments had a significant benefit in decreasing the use of mechanical ventilation. No heterogeneity was identified (Cochran’s Q = 1.68; p = 0.95; τ2 = 0; I2 = 0% [95% CI: 0%-0%]). Evidence from RCTs suggests that AZ with or without HCQ was not associated with a significant effect on the mortality or mechanical ventilation rates in hospitalized patients with COVID-19. [emphases added]
I know you are a fan of Raoult, @Giltil, but please read these 2 meta-analyses with an open mind, if possible. If you read with an open mind, I think you will recognize that the high-quality evidence overwhelmingly opposes the use of HcQ + Az in treating COVID-19.
Meilleures salutations,
Christopher
2 Likes
I don’t have to make that decision. It suffices to recognize that @Giltil is probably wrong.
4 Likes
This the the current guiding matrix for Covid treatment in British Columbia.
Use of therapies in the management of COVID-19.
Hydroxychloroquine is not recommended at any level of severity. Ivermectin is not recommended at any level of severity except as part of approved clinical trials.
2 Likes
WHOA there Tex. Raoult has 10,000 patients but ONLY 16 mortalities. The statistical power in logistic regression comes from the number of events(1), which is only 16. I haven’t looked at the other paper yet (busy here) but @Chris_Falter is also well qualified to comment.
(1) Technically it is the number of events OR non-events, whichever is less. A detail that doesn’t change this discussion.
ETA: The clinical trial Michael cited is powered on length of symptoms, not mortality. Comparing power for different outcomes is sort of like comparing Apples & Oranges.
4 Likes
This survey, published within the past couple of weeks, is an clinical oriented overview of the constellation of treatment protocols for covid.
A hitchhiker’s guide through the COVID-19 galaxy
From this review…
Among its many anti-inflammatory and anti-thrombotic effects, HCQ interferes with viral uptake and intracellular transport by altering the endosomal pH. However, HCQ failed to demonstrate an impact on clinical outcome or survival in exposed presymptomatic individuals , including those with mild disease , those hospitalized with or without O2 requirement and with severe COVID-19. HCQ prolongs the QT interval, which, particularly in patients with underlying cardiac problems, is another argument against its widespread use. Two metanalyses on the effect of HQC in combination with Azithromycin demonstrated an increase in mortality among hospitalized patients.
While this paper lists several interventions found to be of no benefit, there are also several positive treatments. It is not just oxygen and good luck. It looks to me that there is enormous drive to deliver good news on the treatment front as well as vaccination. Stated otherwise, there is no impetus to suppress any promising prospective treatment, no hidebound culture of suppression against new or unconventional ideas, no mind control by big pharma, or any of the other boogie men invoked to smother maverick researchers. Hydroxychloroquine and azithromycin have received fair and even hopeful investigation, but ultimately and disappointedly proved to yield false promise.
I could somewhat understand the pushback on vaccination and conventional covid treatment were it part of good old generalized skepticism. What I cannot make sense of is that the denial of well established public health management where the data is solid appears to be coupled with fervent embrace of fringe ideas where the data is marginal to counter indicated.
Thanks for demonstrating you did not read the paper carefully and that you don’t know how statistical power relates to sample size and expected effect size. For some reason you missed this part of the paper:
The researchers were even gracious to HCQ+AZT by wanting to capture at least 50% efficacy. The bottom line is that sample size used in this study was more than enough for their study aim.
More importantly, I had earlier cited a meta-analysis of RCT studies which looked at AZ alone, HCQ alone or HCQ+AZT versus placebo and not surprisingly “Raoult’s protocol” was ineffective at reducing mortality. Read the darn paper!
https://onlinelibrary.wiley.com/doi/10.1002/jmv.27259
In the article below the authors lay out the numerous significant flaws in the Raoult’s (actually Million’s) observational study. They literally tore the Raoult paper to shreds.
You are right here. The re-analysis results alone done by Bryant et al contradict my statement on the loss of benefit when the Elgazzar study is excluded.
However, you seem to forget that the results of the TOGETHER trial which was concluded quite recently and contained efficacy data on early treatment of mild to moderately ill Covid-19 patients with ivermectin was published and ivermectin flopped big time in it. It was large (1,300 participants), double-blinded, randomized and controlled. The results have not being published yet, but the team released a graphic of the results shown below:
That’s not all. When the TOGETHER trial result and that of another recently published RCT are added to Bryant et al data, do you know what happens? Ivermectin returns to square one (that is, at the moment we can’t tell based on a combination of several RCT results if ivermectin is effective or not for Covid-19; however, the TOGETHER trial data is the highest quality data available and it deals a heavy blow to ivermectin usage). This is the revised Bryant et al meta-analysis as done by Gideon M-K incorporating the new data and showing what I described above:
https://www.google.com/amp/s/elemental.medium.com/amp/p/5e913bb49483
Thus, as at present there is nothing to support ivermectin use outside clinical trials.
3 Likes
It is interesting, isn’t it? As I was suggesting above, I think that straightforward explanations fail. What’s going on here is a kind of complex cultural process, worthy of the study of cultural anthropologists. Margaret Mead never saw anything so strange. It’s one thing for people to be wholly alien to your culture; it’s quite another for them to be superficially familiar, but underneath, to be completely incomprehensible in their bizarre orientation to reality.
We see in politics the formation of strange constellations of belief. Many of these beliefs have nothing to do with one another and yet they run in clusters. Are you a monetarist or a Keynesian? Do you support anti-discrimination laws? Should a woman have the right to abortion? Should your local schools be able to inject religious content into your child’s biology class? How much force is permissible for a police officer to use when confronting a member of the public? We all know that while there is variety among people, and any one person’s views COULD be a unique blend, their views on these things tend to cluster into ideologies.
And ideology has a gravitational force of its own, once it gets going. My father, who never had a creationist bone in his body and who definitely rejected Christian religion in all of its forms, late in life started watching a lot of Fox News, and one day he sent me an article by Jonathan Wells which he’d found persuasive. This deterioration was not a coincidence: it was the consequence of saturation in a toxic brew of ideologically-motivated modification of facts. I knew people could lose their minds due to dementia or injury; I barely suspected that a man could become stark-raving enough to believe things said by Jonathan Wells just by watching bad television.
I think in the case of COVID craziness we are just looking at that kind of historical contingency. There’s no reason in principle why the craziness on this issue shouldn’t be associated with the left more than the right, or with alternative religious beliefs more than with Talibangelicalism. But reasons in principle don’t matter – not much, anyhow – in this creation of these belief-constellations. It’s the weird random walk of cultural memes at work. Every ideology has its toolkit for force-fitting things that one wants to believe, or wants to reject. I just hope that these denialists don’t manage to kill us all – either willfully or unwittingly.
4 Likes
I don’t see an argument in there, Gil. In you, I see someone who writes about the statistical power of studies without understanding what the term means, as @Dan_Eastwood pointed out politely.
I see someone who doesn’t read the studies people cite, and then pretends that the paper had never been cited.
Speaking of ad hominems, would you kindly explain yours?
Precisely what gaps did you conclude exist in my medical understanding?
How are you qualified to judge?
2 Likes
2 Likes
[quote=“Mercer, post:56, topic:14329”]
@Giltil : Gil, first my apologies for not removing one ad hominem quickly enough. Once you quote it then it’s problematic if I remove it or not.
Second, if you had not noticed, I’ve been making something of a personal appeal to what I see as a bad decision that could cause you harm. To make this perfectly clear I have absolutely no control over your personal decisions, but I am asking you in all sincerity, please don’t jump off that ledge.
1 Like
I’m not sure there’s an ad hominem in there, Dan. I think that @Faizal_Ali is not arguing that Giltil is wrong because of what sort of person Giltil is. He is asking: given that Giltil is so disastrously and bizarrely wrong, why is that? How’d it get that way?
And this question has been asked before. I recall a lengthy conversation in which Giltil took on the wholly impossible task of trying to defend Douglas Axe’s claim that biologists now believe evolution has stopped. Nobody can witness such things without wondering just what the hell is going on here and how it is possible for a person’s head to get high-centered so often and so badly.
Now, if you feel that @Faizal_Ali was ungracious or impolite or some such thing, and that this requires moderator action, you know, I guess that’s up to you. But I do think that accusing him of making an ad hominem argument is inaccurate.
2 Likes
Yes. The argument regarding whether HcQ is an effective treatment for COVID-19 has been settled. It is not an effective treatment.
So my comment could not have been an ad hominem argument, because it was not an argument at all. Rather, it was a remark regarding the possible reasons that one of our members is not accepting the obviously correct outcome of the argument.
1 Like
AND
OK, you guys are right to call me out on this, and I’ve gotten myself into a no win scenario. The best possible outcome is raising awareness of what others see as ad hominem, which is what is happening now (and accomplishes my purpose).
3 Likes