The placebo group didn't need to become unblinded. Assume they were in the study to receive free ivermectin. Due to wide use in Brazil they know that ivermectin works and works fairly rapidly. As soon as they are several days into the study and are getting worse they are going to go elsewhere for their ivermectin.
Yes. Not a scientist but remember grade school math. Even now, I compose manual calculations to check computer program work results - and my work is not life and death critical. Who does this kind of work???? Thank you, Phil. Looking forward to your next analysis of this so-called study.
The question that they could easily answer is “How many deaths were there in the Per Protocol Placebo arm?”
It’s a mystery to me why they didn’t use this as the primary metric anyway, it simply makes no sense to mix and match the results from one cohort (Per Protocol) and another.
Even then, this would not explain why so many of the other totals don’t add up. A peer-reviewed paper should already have undergone such scrutiny.
Asking patients for the estimate as to days of symptoms is not a scientifically precise measurement that can be relied upon.
Also, the important start date is not when they enrolled but how many days after first symptom did they receive the first pill?
The qualification process, before getting a pill, appears quite lengthy:
Inclusion criteria and steps to approval :
Patient says yes "within 7 days after symptom onset"
Meets standard of "at least one high-risk criterion for progression of Covid-19"
Then "trial personnel obtained written in-person informed consent and performed a rapid antigen test for SARS-CoV-2..."
Then. "Before randomization, trial personnel obtained data on ..."
Then "Participants also completed the Patient-Reported Outcomes Measurement Information System Global-10 health scale..".
Then "Patients underwent randomization by means of a block randomization procedure...:
The active-drug and placebo pills were packaged in identically shaped bottles and labeled...
Then "Patients were shown a welcome video with information on the trial, ivermectin, adverse events, and follow-up procedures... "
Then obtained medication ?
Why not list the number of days after first symptom that they took the first dose, rather than the number of days after first symptom that they applied for inclusion?
It appears that the per-protocol numbers are placebo patients who completed a 3-day placebo, as opposed to 1,7,10,14 day placebo. So the others didn't drop off, they're just a non-protocol placebo.
They shouldn't have used the full protocol at all unless they were comparing like for like. Say a medication was only given for 3 days, then the placebo arm should only be calculated for 3 days. They should also only be using the number of people that completed the protocol. You can't get real results if you use the wrong denominator.
I agree for the IVM they should have counted only the ones who took the full 3 doses for the denominator. But for the placebo group it shouldn’t make a difference since it was an inert substance that shouldn’t effect survival. What might have been helpful would have been if they gave the deaths separately for the full dose placebo subjects. Again, it shouldn’t make a difference but it would give a consistency test because then you could use the smaller 288 denominator for the placebo group.
However, I can see a scenario where it might make difference for the denominator for the placebo group. It’s reasonable to suppose all subjects knew the full dosage for the real IVM treatment should be for 3 days. Actually I think I saw that stated in the information provided to the subjects. Then those getting only 1 dose would KNOW they got the placebo. In that case since they knew they did indeed have COVID, which is potentially fatal, they would have strong incentive to get the real IVM outside the trial.
So they should have been dropped from the study. If they didn't do the protocol as prescribed then non of them can be counted. I would think that is the way a study is done.
Suspected the Together Trial would be rigged to show ivermectin failure. And…no surprise. Really appreciate your analysis Phil. It’s a shame that those who follow mainstream media will never see this.
However, the insurance industry just might be the financial and political powerhouse needed to expose pharmaceutical and public healthcare agencies corruption. Reports like this just might go into their quiver.
Phil, there were 3 different analyses done: per protocol, intention to treat, and a modified per protocol. Per protocol was the one that was supposed to omit everyone who did not (by their own admission which may not be accurate) comply with the study.
About 24 mins in he discusses the exclusion criteria showing a slide where they excluded those taking the study medications. This is misleading because, as I mentioned in another comment, as late as Nov. 2021 in a report they wrote on the trial protocol, there STILL was no mention of excluding those already taking the trial medications. This is a full two months AFTER they already submitted for publication.
It has been discussed that the six month delay in NEJM finally publishing it is unusually long. I think there is little doubt that the journal reviewers required them to go back and interview subjects to insure none of them were taking the IVM.
It’s difficult to believe, as the authors imply, none of them had taken IVM outside the trial since it was widely promoted and taken for COVID in Brazil, being sold in Brazil over-the-counter(no prescription required), especially since the large number of placebo subjects only getting one dose, would KNOW they were in the placebo group.
It’s not just in that separately published protocol from Nov., 2021 do they make no mention of excluding on prior IVM use, but in the PROTOCOL THEY PROVIDED WITH THE PUBLISHED NEJM ARTICLE THEY STILL MAKE NO MENTION OF EXCLUDING ON PRIOR IVM USE.
In the NEJM article they state they interviewed subjects to remove those taking IVM, but why would that not appear in the protocol?
Here’s the link to the protocol that actually accompanies the NEJM paper:
This is a 260 page pdf file, so I opened it up in Adobe and searched on the word: exclusion. There were two versions of the protocol in the file, one from Feb., 2021, and one from Jan., 2022. Neither one has any mention of excluding on prior IVM use.
It has been discussed that the 6 month delay in publishing the NEJM paper was unusually long. I think it was due to the NEJM reviewers requiring the authors to go back and interview subjects to insure they didn’t take prior IVM, since the authors, apparently, made no attempt to do this on their own.
The authors imply that none of their subjects did. This is hard to believe since IVM was so widespread in Brazil and the protocol by which they were recruited made no mention of not taking it.
REVISED A multi-center, adaptive, randomized, platform trial to evaluate the effect of repurposed medicines in outpatients with early coronavirus disease 2019 (COVID-19) and high-risk for complications: the TOGETHER master trial protocol [version 2; peer review: 1 approved, 1 approved with reservations]
You are amazing Phil - I took one look at the pdf and decided this was too complex for me to sort out - I also support Dr. Mobeen Syed on Patreon. It sounds like he is also going to do a deep dive video review on either Odysee or Rumble. - Do you know if John Campbell is also going to do a review?
The placebo group didn't need to become unblinded. Assume they were in the study to receive free ivermectin. Due to wide use in Brazil they know that ivermectin works and works fairly rapidly. As soon as they are several days into the study and are getting worse they are going to go elsewhere for their ivermectin.
So well explained, even I could understand it. A new happy monthly subscriber!
So, using the numbers for patients who completed the study...
Ivermectin, deaths, 21/624 = 3.4%
Placebo, deaths, 25/288 = 8.7%
Ivermectin reduced deaths by 61% , or said the opposite way, not taking Ivermectin increased the chances of death by 2.6 times.
As a retired scientist I find the sloppy work in these tables unacceptable. Obviously they were not well vetted. Thank you for this review.
Yes. Not a scientist but remember grade school math. Even now, I compose manual calculations to check computer program work results - and my work is not life and death critical. Who does this kind of work???? Thank you, Phil. Looking forward to your next analysis of this so-called study.
As a person formerly known as a scientist, I agree with Dianne. (It's now too embarrassing to be identified as a "scientist.")
The question that they could easily answer is “How many deaths were there in the Per Protocol Placebo arm?”
It’s a mystery to me why they didn’t use this as the primary metric anyway, it simply makes no sense to mix and match the results from one cohort (Per Protocol) and another.
Even then, this would not explain why so many of the other totals don’t add up. A peer-reviewed paper should already have undergone such scrutiny.
This stinks.
Asking patients for the estimate as to days of symptoms is not a scientifically precise measurement that can be relied upon.
Also, the important start date is not when they enrolled but how many days after first symptom did they receive the first pill?
The qualification process, before getting a pill, appears quite lengthy:
Inclusion criteria and steps to approval :
Patient says yes "within 7 days after symptom onset"
Meets standard of "at least one high-risk criterion for progression of Covid-19"
Then "trial personnel obtained written in-person informed consent and performed a rapid antigen test for SARS-CoV-2..."
Then. "Before randomization, trial personnel obtained data on ..."
Then "Participants also completed the Patient-Reported Outcomes Measurement Information System Global-10 health scale..".
Then "Patients underwent randomization by means of a block randomization procedure...:
The active-drug and placebo pills were packaged in identically shaped bottles and labeled...
Then "Patients were shown a welcome video with information on the trial, ivermectin, adverse events, and follow-up procedures... "
Then obtained medication ?
Why not list the number of days after first symptom that they took the first dose, rather than the number of days after first symptom that they applied for inclusion?
It appears that the per-protocol numbers are placebo patients who completed a 3-day placebo, as opposed to 1,7,10,14 day placebo. So the others didn't drop off, they're just a non-protocol placebo.
Doesn't it still stand to reason they should have used 288 not 679 to ensure control?
Yes
They should at least have done the full per-protocol analysis like they did for fluvoxamine.
They shouldn't have used the full protocol at all unless they were comparing like for like. Say a medication was only given for 3 days, then the placebo arm should only be calculated for 3 days. They should also only be using the number of people that completed the protocol. You can't get real results if you use the wrong denominator.
I agree for the IVM they should have counted only the ones who took the full 3 doses for the denominator. But for the placebo group it shouldn’t make a difference since it was an inert substance that shouldn’t effect survival. What might have been helpful would have been if they gave the deaths separately for the full dose placebo subjects. Again, it shouldn’t make a difference but it would give a consistency test because then you could use the smaller 288 denominator for the placebo group.
However, I can see a scenario where it might make difference for the denominator for the placebo group. It’s reasonable to suppose all subjects knew the full dosage for the real IVM treatment should be for 3 days. Actually I think I saw that stated in the information provided to the subjects. Then those getting only 1 dose would KNOW they got the placebo. In that case since they knew they did indeed have COVID, which is potentially fatal, they would have strong incentive to get the real IVM outside the trial.
Robert Clark
So they should have been dropped from the study. If they didn't do the protocol as prescribed then non of them can be counted. I would think that is the way a study is done.
Suspected the Together Trial would be rigged to show ivermectin failure. And…no surprise. Really appreciate your analysis Phil. It’s a shame that those who follow mainstream media will never see this.
However, the insurance industry just might be the financial and political powerhouse needed to expose pharmaceutical and public healthcare agencies corruption. Reports like this just might go into their quiver.
Thanks Phil for your thorough analysis. Its so needed.
Indeed, another attempt to fool people with sloppy “science”. Thanks for all the digging!
Phil, there were 3 different analyses done: per protocol, intention to treat, and a modified per protocol. Per protocol was the one that was supposed to omit everyone who did not (by their own admission which may not be accurate) comply with the study.
Hello, Dr. Nass, would you expand on your comment, please?
I wrote it wrong, the modified analysis was of intention to treat, not per protocol. We could speak by phone. Too much to write here
Ed Mills discusses the TOGETHER trial in this video conference call from March 18, 2022:
Grand Rounds Rethinking Clinical Trials 03-18-2022.
https://vimeo.com/690426154
About 24 mins in he discusses the exclusion criteria showing a slide where they excluded those taking the study medications. This is misleading because, as I mentioned in another comment, as late as Nov. 2021 in a report they wrote on the trial protocol, there STILL was no mention of excluding those already taking the trial medications. This is a full two months AFTER they already submitted for publication.
It has been discussed that the six month delay in NEJM finally publishing it is unusually long. I think there is little doubt that the journal reviewers required them to go back and interview subjects to insure none of them were taking the IVM.
It’s difficult to believe, as the authors imply, none of them had taken IVM outside the trial since it was widely promoted and taken for COVID in Brazil, being sold in Brazil over-the-counter(no prescription required), especially since the large number of placebo subjects only getting one dose, would KNOW they were in the placebo group.
Robert Clark
Do you have a link to the report on the trial protocol? This is interesting.
It’s not just in that separately published protocol from Nov., 2021 do they make no mention of excluding on prior IVM use, but in the PROTOCOL THEY PROVIDED WITH THE PUBLISHED NEJM ARTICLE THEY STILL MAKE NO MENTION OF EXCLUDING ON PRIOR IVM USE.
In the NEJM article they state they interviewed subjects to remove those taking IVM, but why would that not appear in the protocol?
Here’s the link to the protocol that actually accompanies the NEJM paper:
https://www.nejm.org/doi/suppl/10.1056/NEJMoa2115869/suppl_file/nejmoa2115869_protocol.pdf
This is a 260 page pdf file, so I opened it up in Adobe and searched on the word: exclusion. There were two versions of the protocol in the file, one from Feb., 2021, and one from Jan., 2022. Neither one has any mention of excluding on prior IVM use.
It has been discussed that the 6 month delay in publishing the NEJM paper was unusually long. I think it was due to the NEJM reviewers requiring the authors to go back and interview subjects to insure they didn’t take prior IVM, since the authors, apparently, made no attempt to do this on their own.
The authors imply that none of their subjects did. This is hard to believe since IVM was so widespread in Brazil and the protocol by which they were recruited made no mention of not taking it.
Robert Clark
Yes. Here:
REVISED A multi-center, adaptive, randomized, platform trial to evaluate the effect of repurposed medicines in outpatients with early coronavirus disease 2019 (COVID-19) and high-risk for complications: the TOGETHER master trial protocol [version 2; peer review: 1 approved, 1 approved with reservations]
https://gatesopenresearch.org/articles/5-117/v2
This revised protocol was published Nov., 2021, and still no mention of excluding on prior IVM use.
Robert Clark
You are amazing Phil - I took one look at the pdf and decided this was too complex for me to sort out - I also support Dr. Mobeen Syed on Patreon. It sounds like he is also going to do a deep dive video review on either Odysee or Rumble. - Do you know if John Campbell is also going to do a review?
Thanks for all you do!!
Yes Dr John said he is planning to.
Another odd aspect of the TOGETHER trial is they did not exclude for prior OR EVEN CURRENT IVM USE:
Robert Clark
@RGregoryClark
Replying to
@omnichord
and
@boulware_dr
As late as Nov., 2021 the trial authors *still* said no exclusions for prior IVM use: 2 mos. AFTER it was already submitted for publication!
https://gatesopenresearch.org/articles/5-117
Only now *after* it’s published are they claiming they excluded on IVM use. MUST do INDEPENDENT review on prior use
https://twitter.com/rgregoryclark/status/1510731349476401160?s=21&t=XAZw5fMpMAnfW9ww5xxftg
From the study protocol published Nov. 2021, two months after submission to NEJM:
===================================================
Patients who meet any of the following criteria are excluded:
1. Patients with acute respiratory condition compatible with COVID-19 treated in the primary care and with hospitalization need.
2. Patients with acute respiratory condition due to other causes.
3. Severe terminal illness irrespective of type or etiology.
4. Acute flu showing at least ONE of the criteria below:
I. Respiratory Rate > 28 / min
II. SaO2 < 90% or < 93% on nasal oxygen therapy at 10 L / min;
III. PaO2 / FIO2 < 300 mmHg
5. Use of the following medications in the last 14 days:
Monoamine Oxide Inhibitors (MAOIs): Phenelzine, Tranylcypromine, Selegiline, Isocarboxazide, moclobemide
Alpha-1 antagonists, Sotalol, Clonidine, Phosphodiesterase 5 inhibitors, Methyldopa, Prazosin, terasozin, Doxazosin)
Use of antiretroviral agents
6. Patients using serotonin reception inhibitors (Donepezil, Sertraline).
7. Pregnant or breastfeeding patients.
8. Surgical procedure or use of contrast planned to occur during treatment or up to 05 days after the last dose of the study medication.
9. Inability of the patient or representative to give informed consent or adhere to the procedures proposed in the protocol.
10. Known hypersensitivity and / or intolerance to IPs, or taking medications contraindicated by IPs.
11. Inability to follow protocol-related procedures.
===================================================
https://gatesopenresearch.org/articles/5-117
Robert Clark
Thank you Phil!