[JustAnotherFan]

Front page of CNN right now:
Drug touted by Trump as Covid-19 treatment linked to a greater risk of death, study finds
https://www.cnn.com/2020/05/22/health/us...index.html

Here's the actual study report.

It is pretty clear to me that the patients that were in worse shape were given the medicines (and thus wound up in this study's test groups).

One could use this study's "control group" to see what factors give better outcomes.
Indeed, based on this study, I wonder if being a former smoker is actually beneficial for COVID-19.

Let's see, what are the factors associated with worse outcomes (and reported by this study):
Older:  Control group was younger than any of the test groups.
Male:  Control group had fewer males than any of the test groups.
Overweight:  Control group had lower BMI than any of the test groups.
Race:  Control group had more whites than 3 of the 4 test groups
Comorbidities:
  Coronary artery disease: Control group was less afflicted than any of the test groups.
  Congestive heart failure: Control group was less afflicted than any of the test groups.
  Arrhythmia: Control group was in the middle of the groups.
  Diabetes: Control group was less afflicted than any of the test groups.
  Hypertension: Control group was much less afflicted than any of the test groups.
  Hyperlipidaemia: Control group was more afflicted than 3 of the 4 test groups.
  COPD: Control group was less afflicted than 3 of the 4 test groups.
  Current smoker:  Control group was less afflicted than any of the test groups.
  Former smoker: Control group was more afflicted than any of the test groups.
  Immunocompromised:  Control group was in the middle of the groups.
Disease Severity
  qSOFA:  Control group was less afflicted than any of the test groups.
  SPO2:  Control group was less afflicted than any of the test groups.

Maybe, they somehow overcame the statistical one-sidedness of their data by adjusting for it, but then I have to wonder if the conclusions are just a reflection of how over or under that adjustment was.

I do see how a particular media outlet that dumps on Trump would tout the higher deaths in the test group, but I don't think that's good science.  But it promotes the message that outlet wants to promote.  As for me, I'd file this study in the trash bin.

[This study aside (as it should be), I think the President is not behaving wisely in taking this drug.]

Thanks for your thoughts. I posted it without investigating it further because being front page on CNN I thought it was of interest to this group. I really like how you laid out the problems with it.

---

Maybe, they somehow overcame the statistical one-sidedness of their data by adjusting for it, but then I have to wonder if the conclusions are just a reflection of how over or under that adjustment was.

Indeed they did:

To minimise the effect of confounding factors, a propensity score matching analysis was done individually for each of the four treatment groups compared with a control group that received no form of that therapy. For each treatment group, a separate matched control was identified using exact and propensity-score matched criteria with a calliper of 0·001. This method was used to provide a close approximation of demographics, comorbidities, disease severity, and baseline medications between patients. The propensity score was based on the following variables: age, BMI, gender, race or ethnicity, comorbidities, use of ACE inhibitors, use of statins, use of angiotensin receptor blockers, treatment with other antivirals, qSOFA score of less than 1, and SPO2 of less than 94% on room air. The patients were well matched, with standardised mean difference estimates of less than 10% for all matched parameters.

Basically what they did is for each patient in the treated arms, they found a patient in the control arm that "matched" them based on the criteria (more specifically, their propensity score, which is the probability that they would be treated based on their baseline characteristics).  The question you pose is a good one, and basically amounts to a term called residual confounding, the idea that the adjustment might be inadequate and that some other factor(s) might be there that they failed to adjust for.  The authors note this as a limitation in their study.

Due to the observational study design, we cannot exclude the possibility of unmeasured confounding factors, although we have reassuringly noted consistency between the primary analysis and the propensity score matched analyses.

I think at this point we are at the limits of what observational data can tell us.  They bring up some red flags as they have consistently shown no benefit and possible harm, but there is no getting around the methodological limitations. So, we will await the results of the randomized, placebo controlled studies (for example, the WHO has one) before making a firm conclusion, but at this point, I wouldn't give a patient of mine CQ/HCQ outside of a clinical study.

BC

Excellent followup, BC.

[JustAnotherFan]

Front page of CNN right now:
Drug touted by Trump as Covid-19 treatment linked to a greater risk of death, study finds
https://www.cnn.com/2020/05/22/health/us...index.html

Here's the actual study report.

It is pretty clear to me that the patients that were in worse shape were given the medicines (and thus wound up in this study's test groups).

One could use this study's "control group" to see what factors give better outcomes.
Indeed, based on this study, I wonder if being a former smoker is actually beneficial for COVID-19.

Let's see, what are the factors associated with worse outcomes (and reported by this study):
Older:  Control group was younger than any of the test groups.
Male:  Control group had fewer males than any of the test groups.
Overweight:  Control group had lower BMI than any of the test groups.
Race:  Control group had more whites than 3 of the 4 test groups
Comorbidities:
  Coronary artery disease: Control group was less afflicted than any of the test groups.
  Congestive heart failure: Control group was less afflicted than any of the test groups.
  Arrhythmia: Control group was in the middle of the groups.
  Diabetes: Control group was less afflicted than any of the test groups.
  Hypertension: Control group was much less afflicted than any of the test groups.
  Hyperlipidaemia: Control group was more afflicted than 3 of the 4 test groups.
  COPD: Control group was less afflicted than 3 of the 4 test groups.
  Current smoker:  Control group was less afflicted than any of the test groups.
  Former smoker: Control group was more afflicted than any of the test groups.
  Immunocompromised:  Control group was in the middle of the groups.
Disease Severity
  qSOFA:  Control group was less afflicted than any of the test groups.
  SPO2:  Control group was less afflicted than any of the test groups.

Maybe, they somehow overcame the statistical one-sidedness of their data by adjusting for it, but then I have to wonder if the conclusions are just a reflection of how over or under that adjustment was.

I do see how a particular media outlet that dumps on Trump would tout the higher deaths in the test group, but I don't think that's good science.  But it promotes the message that outlet wants to promote.  As for me, I'd file this study in the trash bin.

[This study aside (as it should be), I think the President is not behaving wisely in taking this drug.]

Thanks for your thoughts. I posted it without investigating it further because being front page on CNN I thought it was of interest to this group. I really like how you laid out the problems with it.

---

Maybe, they somehow overcame the statistical one-sidedness of their data by adjusting for it, but then I have to wonder if the conclusions are just a reflection of how over or under that adjustment was.

Indeed they did:

To minimise the effect of confounding factors, a propensity score matching analysis was done individually for each of the four treatment groups compared with a control group that received no form of that therapy. For each treatment group, a separate matched control was identified using exact and propensity-score matched criteria with a calliper of 0·001. This method was used to provide a close approximation of demographics, comorbidities, disease severity, and baseline medications between patients. The propensity score was based on the following variables: age, BMI, gender, race or ethnicity, comorbidities, use of ACE inhibitors, use of statins, use of angiotensin receptor blockers, treatment with other antivirals, qSOFA score of less than 1, and SPO2 of less than 94% on room air. The patients were well matched, with standardised mean difference estimates of less than 10% for all matched parameters.

Basically what they did is for each patient in the treated arms, they found a patient in the control arm that "matched" them based on the criteria (more specifically, their propensity score, which is the probability that they would be treated based on their baseline characteristics).  The question you pose is a good one, and basically amounts to a term called residual confounding, the idea that the adjustment might be inadequate and that some other factor(s) might be there that they failed to adjust for.  The authors note this as a limitation in their study.

Due to the observational study design, we cannot exclude the possibility of unmeasured confounding factors, although we have reassuringly noted consistency between the primary analysis and the propensity score matched analyses.

I think at this point we are at the limits of what observational data can tell us.  They bring up some red flags as they have consistently shown no benefit and possible harm, but there is no getting around the methodological limitations. So, we will await the results of the randomized, placebo controlled studies (for example, the WHO has one) before making a firm conclusion, but at this point, I wouldn't give a patient of mine CQ/HCQ outside of a clinical study.

BC

Excellent followup, BC.

[BostonCard]

We may have to wait a bit longer for that randomized, controlled (but open-label) study of hydroxychloroquine.

https://www.npr.org/sections/coronavirus...y-concerns

The WHO, which is running the largest trial of hydroxychloroquine, is temporarily halting the HCQ arm of the trial, in light of the new paper published in the Lancet.

"The steering committee met over the weekend, in the light of this uncertainty," Swaminathan says. "We decided we should be proactive, err on the side of caution and suspend enrollment temporarily into the hydroxychloroquine arm [of the Solidarity trial]."

The other arms of the trial (local standard of care, remdesivir, lopinavir/ritonavir, and lopinavir/ritonavir + interferon ß) are continuing.  My guess is that the WHO investigators will allow those who got hydroxychloroquine to make it to the primary endpoint (all cause mortality after 28 days), though it might do a partial data cut with the patients who have already made the endpoint.  It will then do a safety unblinding to see if mortality is higher in this group than in the standard treatment group.  If it is, then they will stop that arm for good; if it is not, they will re-open enrollment in the arm.  Depending on how many patients they have in that arm, they may also do a futility analysis and could stop that arm for futility.  There's an outside chance that they would also look for superiority, but the context makes me think that's unlikely.

My guess is that they will eventually resume the trial, but stay tuned.

BC

[BostonCard]

We may have to wait a bit longer for that randomized, controlled (but open-label) study of hydroxychloroquine.

https://www.npr.org/sections/coronavirus...y-concerns

The WHO, which is running the largest trial of hydroxychloroquine, is temporarily halting the HCQ arm of the trial, in light of the new paper published in the Lancet.

"The steering committee met over the weekend, in the light of this uncertainty," Swaminathan says. "We decided we should be proactive, err on the side of caution and suspend enrollment temporarily into the hydroxychloroquine arm [of the Solidarity trial]."

The other arms of the trial (local standard of care, remdesivir, lopinavir/ritonavir, and lopinavir/ritonavir + interferon ß) are continuing.  My guess is that the WHO investigators will allow those who got hydroxychloroquine to make it to the primary endpoint (all cause mortality after 28 days), though it might do a partial data cut with the patients who have already made the endpoint.  It will then do a safety unblinding to see if mortality is higher in this group than in the standard treatment group.  If it is, then they will stop that arm for good; if it is not, they will re-open enrollment in the arm.  Depending on how many patients they have in that arm, they may also do a futility analysis and could stop that arm for futility.  There's an outside chance that they would also look for superiority, but the context makes me think that's unlikely.

My guess is that they will eventually resume the trial, but stay tuned.

BC

[BostonCard]

Now some people are asking questions about the Lancet study.

https://www.msn.com/en-us/health/health-...r-BB14MxZw

I very much want to see the results of the WHO study, because in the end only a randomized, controlled study will answer the question of whether hydroxychloroquine is helpful, harmful or makes no difference (I think "makes no difference" is the most likely outcome).  If they halted the HCQ arm of the trial at 3500 patients, almost 900 will have gotten hydroxychloroquine, which is a pretty good sample size, which should be pretty well powered to detect a mortality benefit of 5 percentage points (assuming that 20% of patients in the standard of care arm die).  The study should be even more well-powered to detect a reduction in hospitalization time.

BC

[BostonCard]

Now some people are asking questions about the Lancet study.

https://www.msn.com/en-us/health/health-...r-BB14MxZw

I very much want to see the results of the WHO study, because in the end only a randomized, controlled study will answer the question of whether hydroxychloroquine is helpful, harmful or makes no difference (I think "makes no difference" is the most likely outcome).  If they halted the HCQ arm of the trial at 3500 patients, almost 900 will have gotten hydroxychloroquine, which is a pretty good sample size, which should be pretty well powered to detect a mortality benefit of 5 percentage points (assuming that 20% of patients in the standard of care arm die).  The study should be even more well-powered to detect a reduction in hospitalization time.

BC

[BostonCard]

The Lancet has issued an "Expression of Concern" about the study.

https://www.thelancet.com/lancet/article...3620312903

Important  scientific  questions  have  been  raised  about  data  reported  in  the  paper  by  Mandeep  Mehra  et  al—Hydroxychloroquine  or  chloroquine  with  or  without  a  macrolide  for  treatment  of  COVID-19:  a  multinational  registry  analysis1—published  in  The  Lancet  on  May  22,  2020. Although an independent audit of the provenance and  validity  of  the  data  has  been  commissioned  by  the  authors  not  affiliated  with  Surgisphere  and  is  ongoing,  with  results  expected  very  shortly,  we  are  issuing  an Expression  of  Concern  to  alert  readers  to  the  fact  that  serious  scientific  questions  have  been  brought  to  our  attention. We will update this notice as soon as we have further information.

The Lancet Editors

A good summary of the issues:

https://www.sciencemag.org/news/2020/06/...unraveling

But just as quickly, the Lancet results have begun to unravel—and Surgisphere, which provided patient data for two other high-profile COVID-19 papers, has come under withering online scrutiny from researchers and amateur sleuths. They have pointed out many red flags in the Lancet paper, including the astonishing number of patients involved and details about their demographics and prescribed dosing that seem implausible. “It began to stretch and stretch and stretch credulity,” says Nicholas White, a malaria researcher at Mahidol University in Bangkok.

I default to my previous position.  This can only be answered for sure with a randomized, controlled study, of which there are several ongoing, and while the observational studies raise concerns about the safety of hydroxychloroquine, they have methodological limitations (aside from the questions that are being raised about the study in the Lancet) that can only be overcome with a randomized trial.

Still, this is not, as they say, "a good look".

BC

[BostonCard]

The Lancet has issued an "Expression of Concern" about the study.

https://www.thelancet.com/lancet/article...3620312903

Important  scientific  questions  have  been  raised  about  data  reported  in  the  paper  by  Mandeep  Mehra  et  al—Hydroxychloroquine  or  chloroquine  with  or  without  a  macrolide  for  treatment  of  COVID-19:  a  multinational  registry  analysis1—published  in  The  Lancet  on  May  22,  2020. Although an independent audit of the provenance and  validity  of  the  data  has  been  commissioned  by  the  authors  not  affiliated  with  Surgisphere  and  is  ongoing,  with  results  expected  very  shortly,  we  are  issuing  an Expression  of  Concern  to  alert  readers  to  the  fact  that  serious  scientific  questions  have  been  brought  to  our  attention. We will update this notice as soon as we have further information.

The Lancet Editors

A good summary of the issues:

https://www.sciencemag.org/news/2020/06/...unraveling

But just as quickly, the Lancet results have begun to unravel—and Surgisphere, which provided patient data for two other high-profile COVID-19 papers, has come under withering online scrutiny from researchers and amateur sleuths. They have pointed out many red flags in the Lancet paper, including the astonishing number of patients involved and details about their demographics and prescribed dosing that seem implausible. “It began to stretch and stretch and stretch credulity,” says Nicholas White, a malaria researcher at Mahidol University in Bangkok.

I default to my previous position.  This can only be answered for sure with a randomized, controlled study, of which there are several ongoing, and while the observational studies raise concerns about the safety of hydroxychloroquine, they have methodological limitations (aside from the questions that are being raised about the study in the Lancet) that can only be overcome with a randomized trial.

Still, this is not, as they say, "a good look".

BC

[akiddoc]

Also, some confusion over asymptomatic carriers:

https://www.sciencemag.org/news/2020/02/...irus-wrong

[akiddoc]

Also, some confusion over asymptomatic carriers:

https://www.sciencemag.org/news/2020/02/...irus-wrong

[M T]

Also, some confusion over asymptomatic carriers:

https://www.sciencemag.org/news/2020/02/...irus-wrong

I think it would be better phrased to use the term "asymptomatic transmission" to emphasize the question as to whether a person can be asymptomatic and spread the disease to others.

[M T]

Also, some confusion over asymptomatic carriers:

https://www.sciencemag.org/news/2020/02/...irus-wrong

I think it would be better phrased to use the term "asymptomatic transmission" to emphasize the question as to whether a person can be asymptomatic and spread the disease to others.

[BostonCard]

And here it is... the first randomized, placebo-controlled trial of hydroxychloroquine, this in use as post-exposure prophylaxis

https://www.nejm.org/doi/full/10.1056/NE...?query=TOC

Unsurprisingly, hydroxychloroquine doesn't prevent people from getting infected.

Results

We enrolled 821 asymptomatic participants. Overall, 87.6% of the participants (719 of 821) reported a high-risk exposure to a confirmed Covid-19 contact. The incidence of new illness compatible with Covid-19 did not differ significantly between participants receiving hydroxychloroquine (49 of 414 [11.8%]) and those receiving placebo (58 of 407 [14.3%]); the absolute difference was −2.4 percentage points (95% confidence interval, −7.0 to 2.2; P=0.35). Side effects were more common with hydroxychloroquine than with placebo (40.1% vs. 16.8%), but no serious adverse reactions were reported.

More trial results to come in the future, which will look at different patient populations, including those with confirmed disease.  But at the least, if you have been exposed to COVID-19 (defined as adults with household or occupational exposure to someone with confirmed Covid-19 at a distance of less than 6 ft for more than 10 minutes while wearing neither a face mask nor an eye shield (high-risk exposure) or while wearing a face mask but no eye shield (moderate-risk exposure).

There are some limitations: only about 2% of patients had laboratory confirmed COVID-19 (most of the rest of the "positives" were based on symptoms).  Secondly, a very small minority of patients were hospitalized (one in each arm) so if hydroxychloroquine works to reduce the risk of serious disease, you wouldn't have seen it in this study.  Third, patients had up to 4 days from the exposure to start prophylaxis; it's possible that this time period was too late.  Fourth, some 65 patients didn't take their medication; that's a pretty high non-adherence rate.

Since they had the data, I wish they had analyzed the effect of mask wearing on the likelihood of developing COVID-19.

Still, I would not take hydroxychloroquine if I were exposed to COVID-19.

BC

[BostonCard]

And here it is... the first randomized, placebo-controlled trial of hydroxychloroquine, this in use as post-exposure prophylaxis

https://www.nejm.org/doi/full/10.1056/NE...?query=TOC

Unsurprisingly, hydroxychloroquine doesn't prevent people from getting infected.

Results

We enrolled 821 asymptomatic participants. Overall, 87.6% of the participants (719 of 821) reported a high-risk exposure to a confirmed Covid-19 contact. The incidence of new illness compatible with Covid-19 did not differ significantly between participants receiving hydroxychloroquine (49 of 414 [11.8%]) and those receiving placebo (58 of 407 [14.3%]); the absolute difference was −2.4 percentage points (95% confidence interval, −7.0 to 2.2; P=0.35). Side effects were more common with hydroxychloroquine than with placebo (40.1% vs. 16.8%), but no serious adverse reactions were reported.

More trial results to come in the future, which will look at different patient populations, including those with confirmed disease.  But at the least, if you have been exposed to COVID-19 (defined as adults with household or occupational exposure to someone with confirmed Covid-19 at a distance of less than 6 ft for more than 10 minutes while wearing neither a face mask nor an eye shield (high-risk exposure) or while wearing a face mask but no eye shield (moderate-risk exposure).

There are some limitations: only about 2% of patients had laboratory confirmed COVID-19 (most of the rest of the "positives" were based on symptoms).  Secondly, a very small minority of patients were hospitalized (one in each arm) so if hydroxychloroquine works to reduce the risk of serious disease, you wouldn't have seen it in this study.  Third, patients had up to 4 days from the exposure to start prophylaxis; it's possible that this time period was too late.  Fourth, some 65 patients didn't take their medication; that's a pretty high non-adherence rate.

Since they had the data, I wish they had analyzed the effect of mask wearing on the likelihood of developing COVID-19.

Still, I would not take hydroxychloroquine if I were exposed to COVID-19.

BC

[burger]

And here it is... the first randomized, placebo-controlled trial of hydroxychloroquine, this in use as post-exposure prophylaxis

I say give the authors a Nobel Prize just for doing a randomized placebo-controlled study on covid-19.  Pubmed only lists 12 such studies on any drug for covid-19 (there may be more preprints out there).

[burger]

And here it is... the first randomized, placebo-controlled trial of hydroxychloroquine, this in use as post-exposure prophylaxis

I say give the authors a Nobel Prize just for doing a randomized placebo-controlled study on covid-19.  Pubmed only lists 12 such studies on any drug for covid-19 (there may be more preprints out there).

[BostonCard]

There are many other such trials in progress.  Prospective trials always take longer to set up, because you have to do all the planning in advance.  Observational and retrospective trials are always going to be easier to set up and run.

There are currently a lot of trials "out in the field" as it were.

BC

[BostonCard]

There are many other such trials in progress.  Prospective trials always take longer to set up, because you have to do all the planning in advance.  Observational and retrospective trials are always going to be easier to set up and run.

There are currently a lot of trials "out in the field" as it were.

BC

[BostonCard]

The Lancet has retracted the hydroxychloroquine article:

https://www.thelancet.com/journals/lance...6/fulltext

On the one hand, the possibility (likelihood?) that the data would have been manufactured and gotten into the Lancet (along with the NEJM and JAMA, one of the big three general medical journals) is a major black eye for science.  On the other hand, science is very self-correcting and the problems were identified by other researchers, and the process worked.

The sheer sample size was a bit of a siren's song in that article in a way that might have made the article very alluring.

BC

[BostonCard]

The Lancet has retracted the hydroxychloroquine article:

https://www.thelancet.com/journals/lance...6/fulltext

On the one hand, the possibility (likelihood?) that the data would have been manufactured and gotten into the Lancet (along with the NEJM and JAMA, one of the big three general medical journals) is a major black eye for science.  On the other hand, science is very self-correcting and the problems were identified by other researchers, and the process worked.

The sheer sample size was a bit of a siren's song in that article in a way that might have made the article very alluring.

BC

[oldalum]

a major black eye for science.  

and in particular the Harvard lead author and the other authors (one of whom was the source of the data), who could have discovered the same problems in the data that the preprint reviewers did, had they looked questioningly enough.

So how come this isn't called the "Harvard Study" while we're stuck with the controversial but clearly not fabricated "Stanford Study"? Sheesh

[oldalum]

a major black eye for science.  

and in particular the Harvard lead author and the other authors (one of whom was the source of the data), who could have discovered the same problems in the data that the preprint reviewers did, had they looked questioningly enough.

So how come this isn't called the "Harvard Study" while we're stuck with the controversial but clearly not fabricated "Stanford Study"? Sheesh

[76lsjumb]

The Lancet has retracted the hydroxychloroquine article:

https://www.thelancet.com/journals/lance...6/fulltext

On the one hand, the possibility (likelihood?) that the data would have been manufactured and gotten into the Lancet (along with the NEJM and JAMA, one of the big three general medical journals) is a major black eye for science.  On the other hand, science is very self-correcting and the problems were identified by other researchers, and the process worked.

The sheer sample size was a bit of a siren's song in that article in a way that might have made the article very alluring.

BC

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

I know the mantra oft repeated is that one side should be believed because it is relying on "science", while the other side should be disregarded [or, worse, chastised] for relying only on "politics."  However, "scientists" are people, just like politicians, and people -- scientists, politicians, and all -- have their own viewpoints, biases, and prejudices.  In fact, scientists are quite often quite political.  See, e.g., The Lancet [May 16, 2020].

Science is only reliably "self-correcting" if scientists make it so.  That process is jeopardized, though, when "science" is used by one side not to investigate and reveal facts, irrespective of the "political" side that might benefit from those facts, but as a means to silence an opposing political view.

As for the process working, maybe so, but I think the reporting of the study's dubious negative -- read: anti-Trump -- conclusions are likely to far outstrip the news of their retraction.  So, in that sense, it [the article, not so much the process] probably worked from The Lancet's perspective.

[76lsjumb]

The Lancet has retracted the hydroxychloroquine article:

https://www.thelancet.com/journals/lance...6/fulltext

On the one hand, the possibility (likelihood?) that the data would have been manufactured and gotten into the Lancet (along with the NEJM and JAMA, one of the big three general medical journals) is a major black eye for science.  On the other hand, science is very self-correcting and the problems were identified by other researchers, and the process worked.

The sheer sample size was a bit of a siren's song in that article in a way that might have made the article very alluring.

BC

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

I know the mantra oft repeated is that one side should be believed because it is relying on "science", while the other side should be disregarded [or, worse, chastised] for relying only on "politics."  However, "scientists" are people, just like politicians, and people -- scientists, politicians, and all -- have their own viewpoints, biases, and prejudices.  In fact, scientists are quite often quite political.  See, e.g., The Lancet [May 16, 2020].

Science is only reliably "self-correcting" if scientists make it so.  That process is jeopardized, though, when "science" is used by one side not to investigate and reveal facts, irrespective of the "political" side that might benefit from those facts, but as a means to silence an opposing political view.

As for the process working, maybe so, but I think the reporting of the study's dubious negative -- read: anti-Trump -- conclusions are likely to far outstrip the news of their retraction.  So, in that sense, it [the article, not so much the process] probably worked from The Lancet's perspective.

[2006alum]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

I know the mantra oft repeated is that one side should be believed because it is relying on "science", while the other side should be disregarded [or, worse, chastised] for relying only on "politics."  However, "scientists" are people, just like politicians, and people -- scientists, politicians, and all -- have their own viewpoints, biases, and prejudices.  In fact, scientists are quite often quite political.  See, e.g., The Lancet [May 16, 2020].

Science is only reliably "self-correcting" if scientists make it so.  That process is jeopardized, though, when "science" is used by one side not to investigate and reveal facts, irrespective of the "political" side that might benefit from those facts, but as a means to silence an opposing political view.

As for the process working, maybe so, but I think the reporting of the study's dubious negative -- read: anti-Trump -- conclusions are likely to far outstrip the news of their retraction.  So, in that sense, it [the article, not so much the process] probably worked from The Lancet's perspective.

This entire issue would be moot if we didn't have a snake oil salesman for a President who hawks anything that makes him seem large and in charge and insinuates himself into literally every issue. I'm sorry, but the "science" only got politicized because he has something to say about literally every issue, which inherently makes it political, including hydroxychloroquine. As someone said today, he's the bride at every wedding and the corpse at every funeral. 

I also blame the media for hyping each of these preliminary studies so much - the science doesn't even have time to work before the court of public opinion has already rendered a verdict.

And, incidentally, that's what galls me so much about John Ioannidis - he is actively promoting his preliminary findings to alter public policy choices where lives are on the line. The reason it's known as the "Stanford Study" is because he and his co-authors have been leveraging their Stanford credentials to get anyone who will pay attention to listen. 

Maybe one day soon we'll have leadership that has the instinct to deescalate rather than escalate.

[2006alum]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

I know the mantra oft repeated is that one side should be believed because it is relying on "science", while the other side should be disregarded [or, worse, chastised] for relying only on "politics."  However, "scientists" are people, just like politicians, and people -- scientists, politicians, and all -- have their own viewpoints, biases, and prejudices.  In fact, scientists are quite often quite political.  See, e.g., The Lancet [May 16, 2020].

Science is only reliably "self-correcting" if scientists make it so.  That process is jeopardized, though, when "science" is used by one side not to investigate and reveal facts, irrespective of the "political" side that might benefit from those facts, but as a means to silence an opposing political view.

As for the process working, maybe so, but I think the reporting of the study's dubious negative -- read: anti-Trump -- conclusions are likely to far outstrip the news of their retraction.  So, in that sense, it [the article, not so much the process] probably worked from The Lancet's perspective.

This entire issue would be moot if we didn't have a snake oil salesman for a President who hawks anything that makes him seem large and in charge and insinuates himself into literally every issue. I'm sorry, but the "science" only got politicized because he has something to say about literally every issue, which inherently makes it political, including hydroxychloroquine. As someone said today, he's the bride at every wedding and the corpse at every funeral. 

I also blame the media for hyping each of these preliminary studies so much - the science doesn't even have time to work before the court of public opinion has already rendered a verdict.

And, incidentally, that's what galls me so much about John Ioannidis - he is actively promoting his preliminary findings to alter public policy choices where lives are on the line. The reason it's known as the "Stanford Study" is because he and his co-authors have been leveraging their Stanford credentials to get anyone who will pay attention to listen. 

Maybe one day soon we'll have leadership that has the instinct to deescalate rather than escalate.

[BostonCard]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

Unfortunately this is a testable hypothesis, because another high profile paper, by many of the same authors using the same source of data, that had nothing to do with anything that the President talked about, was published in a high profile journal.

https://www.nejm.org/doi/full/10.1056/NEJMoa2007621

Using the dataset provided by Surgisphere, Mandeep Mehra published a high profile paper about whether ACE inhibitors and Angiotensin Receptor Blockers worsened outcomes of COVID-19.  You will recall this was a topic of debate here, as some provocative editorials wondered aloud whether the use of these drugs would lead to increased expression of ACE2, and easier entry of COVID-19 into cells.  In the paper, the authors claimed that ACE inhibitors/ARBs did not worsen COVID-19 outcomes.  And since the president did not say "stop your blood pressure medicines, people are saying they will kill you", you will have to find another theory as to why the paper got into the New England Journal than the journal had it in for the President.

Here's the retraction, FWIW:

https://www.nejm.org/doi/full/10.1056/NE...tured_home

The papers were "sexy" because they were very large observational datasets in a fast moving disease for which we have a very large hunger for knowledge, not because they contradict something the President said.  Had the underlying data been sound, they would have been the next best thing to a randomized trial.  And for better or for worse, they exploited a flaw in the underlying peer review system, which is that while it can identify analytical weaknesses request additional analyses, and push back against conclusions that are not fully supported by the results, they generally have a hard time catching problems with the underlying dataset.  Having peer reviewed many articles in my career, I know it would have been easy to slip fraudulent or sloppily collected data past me.  In fact, though I would have been within my rights to request the data and the underlying analytical code, I never have, nor have I been asked for it when I published.  You just sort of assume that the people publishing the paper are above board.  Unfortunately, that is not always the case.

BC

[BostonCard]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

Unfortunately this is a testable hypothesis, because another high profile paper, by many of the same authors using the same source of data, that had nothing to do with anything that the President talked about, was published in a high profile journal.

https://www.nejm.org/doi/full/10.1056/NEJMoa2007621

Using the dataset provided by Surgisphere, Mandeep Mehra published a high profile paper about whether ACE inhibitors and Angiotensin Receptor Blockers worsened outcomes of COVID-19.  You will recall this was a topic of debate here, as some provocative editorials wondered aloud whether the use of these drugs would lead to increased expression of ACE2, and easier entry of COVID-19 into cells.  In the paper, the authors claimed that ACE inhibitors/ARBs did not worsen COVID-19 outcomes.  And since the president did not say "stop your blood pressure medicines, people are saying they will kill you", you will have to find another theory as to why the paper got into the New England Journal than the journal had it in for the President.

Here's the retraction, FWIW:

https://www.nejm.org/doi/full/10.1056/NE...tured_home

The papers were "sexy" because they were very large observational datasets in a fast moving disease for which we have a very large hunger for knowledge, not because they contradict something the President said.  Had the underlying data been sound, they would have been the next best thing to a randomized trial.  And for better or for worse, they exploited a flaw in the underlying peer review system, which is that while it can identify analytical weaknesses request additional analyses, and push back against conclusions that are not fully supported by the results, they generally have a hard time catching problems with the underlying dataset.  Having peer reviewed many articles in my career, I know it would have been easy to slip fraudulent or sloppily collected data past me.  In fact, though I would have been within my rights to request the data and the underlying analytical code, I never have, nor have I been asked for it when I published.  You just sort of assume that the people publishing the paper are above board.  Unfortunately, that is not always the case.

BC

[76lsjumb]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

Unfortunately this is a testable hypothesis, because another high profile paper, by many of the same authors using the same source of data, that had nothing to do with anything that the President talked about, was published in a high profile journal.

https://www.nejm.org/doi/full/10.1056/NEJMoa2007621

Using the dataset provided by Surgisphere, Mandeep Mehra published a high profile paper about whether ACE inhibitors and Angiotensin Receptor Blockers worsened outcomes of COVID-19.  You will recall this was a topic of debate here, as some provocative editorials wondered aloud whether the use of these drugs would lead to increased expression of ACE2, and easier entry of COVID-19 into cells.  In the paper, the authors claimed that ACE inhibitors/ARBs did not worsen COVID-19 outcomes.  And since the president did not say "stop your blood pressure medicines, people are saying they will kill you", you will have to find another theory as to why the paper got into the New England Journal than the journal had it in for the President.

Here's the retraction, FWIW:

https://www.nejm.org/doi/full/10.1056/NE...tured_home

The papers were "sexy" because they were very large observational datasets in a fast moving disease for which we have a very large hunger for knowledge, not because they contradict something the President said.  Had the underlying data been sound, they would have been the next best thing to a randomized trial.  And for better or for worse, they exploited a flaw in the underlying peer review system, which is that while it can identify analytical weaknesses request additional analyses, and push back against conclusions that are not fully supported by the results, they generally have a hard time catching problems with the underlying dataset.  Having peer reviewed many articles in my career, I know it would have been easy to slip fraudulent or sloppily collected data past me.  In fact, though I would have been within my rights to request the data and the underlying analytical code, I never have, nor have I been asked for it when I published.  You just sort of assume that the people publishing the paper are above board.  Unfortunately, that is not always the case.

BC

I'm sorry, but I honestly don't follow your argument.  As I understand it, you think my hypothesis is incorrect because another medical journal than the one I'm talking about, which, at least as far as I can tell from a fairly quick scan, did not run a highly political editorial criticizing Trump, also published a high profile paper about something that had nothing to do with Trump [as opposed to the one drug that has been so heavily linked to Trump] but had to retract it as well, so, therefore, ...what?  It seems to me that doesn't prove a single thing about what The Lancet did or why it did it, unless, I suppose, you offer it as an indictment of scientists as a whole.  I'm assuming that isn't your intent, though, since that would also undercut the argument that "science" and scientists should trump [no pun intended] politics and politicians because, well, scientists can always be trusted to get things right.  Except when they don't.  Which apparently is happening a lot these days.

If I've missed something, please clarify.

[76lsjumb]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

Unfortunately this is a testable hypothesis, because another high profile paper, by many of the same authors using the same source of data, that had nothing to do with anything that the President talked about, was published in a high profile journal.

https://www.nejm.org/doi/full/10.1056/NEJMoa2007621

Using the dataset provided by Surgisphere, Mandeep Mehra published a high profile paper about whether ACE inhibitors and Angiotensin Receptor Blockers worsened outcomes of COVID-19.  You will recall this was a topic of debate here, as some provocative editorials wondered aloud whether the use of these drugs would lead to increased expression of ACE2, and easier entry of COVID-19 into cells.  In the paper, the authors claimed that ACE inhibitors/ARBs did not worsen COVID-19 outcomes.  And since the president did not say "stop your blood pressure medicines, people are saying they will kill you", you will have to find another theory as to why the paper got into the New England Journal than the journal had it in for the President.

Here's the retraction, FWIW:

https://www.nejm.org/doi/full/10.1056/NE...tured_home

The papers were "sexy" because they were very large observational datasets in a fast moving disease for which we have a very large hunger for knowledge, not because they contradict something the President said.  Had the underlying data been sound, they would have been the next best thing to a randomized trial.  And for better or for worse, they exploited a flaw in the underlying peer review system, which is that while it can identify analytical weaknesses request additional analyses, and push back against conclusions that are not fully supported by the results, they generally have a hard time catching problems with the underlying dataset.  Having peer reviewed many articles in my career, I know it would have been easy to slip fraudulent or sloppily collected data past me.  In fact, though I would have been within my rights to request the data and the underlying analytical code, I never have, nor have I been asked for it when I published.  You just sort of assume that the people publishing the paper are above board.  Unfortunately, that is not always the case.

BC

I'm sorry, but I honestly don't follow your argument.  As I understand it, you think my hypothesis is incorrect because another medical journal than the one I'm talking about, which, at least as far as I can tell from a fairly quick scan, did not run a highly political editorial criticizing Trump, also published a high profile paper about something that had nothing to do with Trump [as opposed to the one drug that has been so heavily linked to Trump] but had to retract it as well, so, therefore, ...what?  It seems to me that doesn't prove a single thing about what The Lancet did or why it did it, unless, I suppose, you offer it as an indictment of scientists as a whole.  I'm assuming that isn't your intent, though, since that would also undercut the argument that "science" and scientists should trump [no pun intended] politics and politicians because, well, scientists can always be trusted to get things right.  Except when they don't.  Which apparently is happening a lot these days.

If I've missed something, please clarify.

[2006alum]

Seems pretty obvious what BC"s point was to me: you're selecting on the dependent variable. If one selects for "withdrawal of high profile medical study in a well regarded medical journal," there's no obvious correlation with anti-Trump bias or effect. There was, however, a plausible alternative explanation: a correlation with reliance on Surgisphere datasets.

As for the relative utility of science, when you offer up a better alternative, I'll be eager to hear it.

[2006alum]

Seems pretty obvious what BC"s point was to me: you're selecting on the dependent variable. If one selects for "withdrawal of high profile medical study in a well regarded medical journal," there's no obvious correlation with anti-Trump bias or effect. There was, however, a plausible alternative explanation: a correlation with reliance on Surgisphere datasets.

As for the relative utility of science, when you offer up a better alternative, I'll be eager to hear it.

[76lsjumb]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

I know the mantra oft repeated is that one side should be believed because it is relying on "science", while the other side should be disregarded [or, worse, chastised] for relying only on "politics."  However, "scientists" are people, just like politicians, and people -- scientists, politicians, and all -- have their own viewpoints, biases, and prejudices.  In fact, scientists are quite often quite political.  See, e.g., The Lancet [May 16, 2020].

Science is only reliably "self-correcting" if scientists make it so.  That process is jeopardized, though, when "science" is used by one side not to investigate and reveal facts, irrespective of the "political" side that might benefit from those facts, but as a means to silence an opposing political view.

As for the process working, maybe so, but I think the reporting of the study's dubious negative -- read: anti-Trump -- conclusions are likely to far outstrip the news of their retraction.  So, in that sense, it [the article, not so much the process] probably worked from The Lancet's perspective.

This entire issue would be moot if we didn't have a snake oil salesman for a President who hawks anything that makes him seem large and in charge and insinuates himself into literally every issue. I'm sorry, but the "science" only got politicized because he has something to say about literally every issue, which inherently makes it political, including hydroxychloroquine. As someone said today, he's the bride at every wedding and the corpse at every funeral. 

I also blame the media for hyping each of these preliminary studies so much - the science doesn't even have time to work before the court of public opinion has already rendered a verdict.

And, incidentally, that's what galls me so much about John Ioannidis - he is actively promoting his preliminary findings to alter public policy choices where lives are on the line. The reason it's known as the "Stanford Study" is because he and his co-authors have been leveraging their Stanford credentials to get anyone who will pay attention to listen. 

Maybe one day soon we'll have leadership that has the instinct to deescalate rather than escalate.

Let me see if I've got this right.  You blame Trump for everything.  You blame the media for hyping this article.  You blame people generally [that is, "the court of public opinion"].  The one group I don't see you blaming anywhere in your post are the folks who committed scientific journalistic malpractice by publishing this thing.

I'm a lifelong conservative Republican that happens to agree that Trump is horrible.  I think he has done immeasurable damage to the Republican Party, the public perception of conservative principles [since I think he has no principles], and many aspects of the country as a whole.  What I find so frustrating is that much of the opposition to him mirrors many of his worst traits.  Put simply, opposition that is based on the principle that anything Trump does is wrong and anything done in opposition to Trump is okay [simply by virtue of the fact that it is against Trump] is, in my mind, no better than Trump.  Trump's a snake oil salesman so it's not a problem to publish false scientific data.  Trump's a liar so it's okay to claim you have irrefutable direct evidence of collusion with the Russians.  And the list goes on.  At some point something changed from "when they go low, we go high" to "when they go low, we get down there with them."

Maybe I'm wrong about this, though.  Maybe the opposition [yours, at least] to Trump isn't so overwhelming as to be irrational, as it seems to me to be.  Can you come up with three good things he has done as President over the past three years for which you would give him credit?  And, if that is deemed to unrelatedly political for the Board, I would amend it to three good things has has done with respect to the COVID-19 situation for which you would give him credit?

Just wondering...

[76lsjumb]

On May 16, 2020, The Lancet ran an editorial highly critical of Donald Trump.  Suppose this study had contained findings that were favorable toward this drug, not unfavorable.  Do you believe it would have been ushered through to print without substantially more review that, presumably, would have avoided such a problem?  I certainly do.  But, given its unfavorable findings, it was [to repeat myself from another thread] obviously "too good to check."

I know the mantra oft repeated is that one side should be believed because it is relying on "science", while the other side should be disregarded [or, worse, chastised] for relying only on "politics."  However, "scientists" are people, just like politicians, and people -- scientists, politicians, and all -- have their own viewpoints, biases, and prejudices.  In fact, scientists are quite often quite political.  See, e.g., The Lancet [May 16, 2020].

Science is only reliably "self-correcting" if scientists make it so.  That process is jeopardized, though, when "science" is used by one side not to investigate and reveal facts, irrespective of the "political" side that might benefit from those facts, but as a means to silence an opposing political view.

As for the process working, maybe so, but I think the reporting of the study's dubious negative -- read: anti-Trump -- conclusions are likely to far outstrip the news of their retraction.  So, in that sense, it [the article, not so much the process] probably worked from The Lancet's perspective.

This entire issue would be moot if we didn't have a snake oil salesman for a President who hawks anything that makes him seem large and in charge and insinuates himself into literally every issue. I'm sorry, but the "science" only got politicized because he has something to say about literally every issue, which inherently makes it political, including hydroxychloroquine. As someone said today, he's the bride at every wedding and the corpse at every funeral. 

I also blame the media for hyping each of these preliminary studies so much - the science doesn't even have time to work before the court of public opinion has already rendered a verdict.

And, incidentally, that's what galls me so much about John Ioannidis - he is actively promoting his preliminary findings to alter public policy choices where lives are on the line. The reason it's known as the "Stanford Study" is because he and his co-authors have been leveraging their Stanford credentials to get anyone who will pay attention to listen. 

Maybe one day soon we'll have leadership that has the instinct to deescalate rather than escalate.

Let me see if I've got this right.  You blame Trump for everything.  You blame the media for hyping this article.  You blame people generally [that is, "the court of public opinion"].  The one group I don't see you blaming anywhere in your post are the folks who committed scientific journalistic malpractice by publishing this thing.

I'm a lifelong conservative Republican that happens to agree that Trump is horrible.  I think he has done immeasurable damage to the Republican Party, the public perception of conservative principles [since I think he has no principles], and many aspects of the country as a whole.  What I find so frustrating is that much of the opposition to him mirrors many of his worst traits.  Put simply, opposition that is based on the principle that anything Trump does is wrong and anything done in opposition to Trump is okay [simply by virtue of the fact that it is against Trump] is, in my mind, no better than Trump.  Trump's a snake oil salesman so it's not a problem to publish false scientific data.  Trump's a liar so it's okay to claim you have irrefutable direct evidence of collusion with the Russians.  And the list goes on.  At some point something changed from "when they go low, we go high" to "when they go low, we get down there with them."

Maybe I'm wrong about this, though.  Maybe the opposition [yours, at least] to Trump isn't so overwhelming as to be irrational, as it seems to me to be.  Can you come up with three good things he has done as President over the past three years for which you would give him credit?  And, if that is deemed to unrelatedly political for the Board, I would amend it to three good things has has done with respect to the COVID-19 situation for which you would give him credit?

Just wondering...

[BostonCard]

Unfortunately, also not the first time the Lancet has retracted a controversial paper.

https://www.thelancet.com/journals/lance...4/fulltext

Pretty sure that for all his faults, Trump is not an anti-vaxxer (this was one of the papers purporting to find a link between vaccines and autism, and it remained in the Lancet for 12 years before being retracted).

BC

[BostonCard]

Unfortunately, also not the first time the Lancet has retracted a controversial paper.

https://www.thelancet.com/journals/lance...4/fulltext

Pretty sure that for all his faults, Trump is not an anti-vaxxer (this was one of the papers purporting to find a link between vaccines and autism, and it remained in the Lancet for 12 years before being retracted).

BC

[76lsjumb]

As for the relative utility of science, when you offer up a better alternative, I'll be eager to hear it.

Well, if we're talking about the governance of society, I'd offer up the Constitution and the First Amendment as at least as good an alternative to "science" imposed by those in power against the will of the people.  and, in light of the past week or so, I'm thinking you might even be willing to agree with me on that.

[76lsjumb]

As for the relative utility of science, when you offer up a better alternative, I'll be eager to hear it.

Well, if we're talking about the governance of society, I'd offer up the Constitution and the First Amendment as at least as good an alternative to "science" imposed by those in power against the will of the people.  and, in light of the past week or so, I'm thinking you might even be willing to agree with me on that.

[martyup]

Reportedly one of the few employees of Surgisphere was a female stripper.  So . . . there's that.

[martyup]

Reportedly one of the few employees of Surgisphere was a female stripper.  So . . . there's that.