While I agree that the numbers are small and probably not significant, I don't agree that GLP1a drugs would be unlikely to help prevent Covid mortality. Given that obesity is a significant risk factor for death by Covid, is it possible that metabolic syndrome, not obesity, is the real risk factor and that GLP1a treats metabolic syndrome?
"There were 569 events in the semaglutide arm vs 701 events in the placebo arm. That is a lot of events."
that sort of assumes that you actually believe the authors; completely; anyone reading the RIAT paper will see that behind these trials seems to be shadowy entity; appointed by pharma; that "adjudicates" results
so if a person has a heart attack in their local hospital, and a cardiologist there says it was a heart attack; there is a good chance that several weeks later, a GOBSAT committee thousands of miles away will say confidently say NO; it was not a heart attack: (GOBSAT = good old boys sat around a table); really ......
until we can access the raw data, I would urge caution in believing anything in these standard format NEJM trials; happy to be told I am completely wrong; just read the RIAT please before saying so; the Canadian authorities gave the RIAT authors full access to raw data; (after the FDA said it would take 10-15 yrs to release it, as they were SO BUSY): the canadians seemed to the SEND button on their computer for instant release; one assumes that they bitterly regret that candid error; and will regret it to the end of their days; never again will that allow that transparency; we must maintain the illusion behind the glittery curtains and mystery of marvellous medical research.
Thank you for that reference. Most people don't realize how unreliable the cause of death figures are. In many of these studies touting drug treatment it only takes a shift of a few cases to invert the result.
So we can expect a few billion $$$$'s worth of purchase by the future Pfizer/Moderna/InsertPharmaHere executives? Uhhh, I mean FDA/CDC "doctors" and "researchers"?
I agree that the treatment is disease modifying, although in 1 out of 67 patients that fulfill the inclusion criteria. If I had to pay for the drug, I would think twice… Is there any cost-effectiveness analysis done for Select?
Anytime I hear the words “miracle drug” I’m skeptical of long term outcomes. I have heard that GLP1 drugs might be implicated for addiction; we have heard “miracle drug” before - with heroin being one for opiate addiction (yes, you read that correctly!), and on and on. Any disease that can be treated with diet and exercise should be treated no other way.
It is not surprising that a group of professional researchers would grub and dig for some subgroup analysis that would generate another "peer-reviewed" publication. It is, however, disappointing to see the initial study hailed as a breakthrough in the discovery of another disease-modifying drug for heart disease. Judge for yourself whether this was "a clinically important and statistically robust finding".
Here are the percentage figures for the major endpoints:
control% rx% ARR
Composite endpoint 6.5% 8.0% 1.5%
CV Death 2.5% 3.0% 0.5%
Nonfatal MI 2.7% 3.7% 1.0%
Nonfatal stroke 1.7% 1.9% 0.2%
Total mortality 4.3% 5.2% 0.9%
The obvious conclusion is that the drug had no effect that was of any practical significance. Another example of how "statistical significance" and misleading graphs with truncated Y axes are used to fool people and promote useless drugs.
I agree. An absolute risk reduction of 1.5% in a trial of over 17,000 does not appear to be cost effective (except for PHARMA). 8000 out of 8800 on the drug had no event.
Not only not cost effective, but not effective at all. To claim that these tiny differences in composite end points of variable degrees of hardness justify "treatment" is ridiculous.
Interesting. I'd be interested to hear the group's opinion on using the GLP1 for weight loss. I become suspicious when a drug becomes a movement. I have friends and colleagues (we are CICU nurses) using this for weight loss. I'm a bit concerned about the long term effects. Myself, I'm using the calorie counting/exercise route. Very slow at my age but the more natural way to healthier living.
try going low-carb Amy; some call it keto: no sugar, no bread, no rice, no pasta .... no carbs as best you can; doctors are "baffled" how well folks do; no amount of exercise will reduce insulin levels; and insulin STOPS you burning fat; check up on Jason Fung; calorie counting is ineffective Amy; burn fat; you will be amazed how your brain works better; our brain does best on fat.
There are some really good videos on this. The big takeaway for me was that it could be a miracle drug, but that they are majorly overdosing people and that you MUST take advantage of the mental reset and change your habits while you're on it. This video is the best I've seen on it:
This was my thought exactly when I read about SELECT in Medpage Today. I'm sure that GLP1 would have significantly reduced COVID-19 mortality among naive patients in 2020 due to reduction in obesity, but not so sure about in 2021-2023 especially given that the study didn't seem properly powered to detect an effect of that size.
This substudy is insanity and stupidity wrapped into one. And an embarrassment that it’s in JACC.
That said, SELECT itself is practice-changing.
I’m no trialist but I like how hierchical analysis is used. It allows something (potentially) to be scientifically validly concluded upon beyond the primary endpoint. I’d like to understand more btw the strengths and weaknesses of such a trial design, vs a total events analysis (as opposed to time to first event) to allow more of the study data to be utilized.
Regardless, they should’ve stopped when the first hierchical analysis was non-insignificant. And this is probably a hypothesis that isn’t even worth generating.
While I agree that the numbers are small and probably not significant, I don't agree that GLP1a drugs would be unlikely to help prevent Covid mortality. Given that obesity is a significant risk factor for death by Covid, is it possible that metabolic syndrome, not obesity, is the real risk factor and that GLP1a treats metabolic syndrome?
Sometime back, I was pointed to the RIAT analysis by a friend; https://bmjopen.bmj.com/content/12/12/e060172
When you say John
"There were 569 events in the semaglutide arm vs 701 events in the placebo arm. That is a lot of events."
that sort of assumes that you actually believe the authors; completely; anyone reading the RIAT paper will see that behind these trials seems to be shadowy entity; appointed by pharma; that "adjudicates" results
so if a person has a heart attack in their local hospital, and a cardiologist there says it was a heart attack; there is a good chance that several weeks later, a GOBSAT committee thousands of miles away will say confidently say NO; it was not a heart attack: (GOBSAT = good old boys sat around a table); really ......
until we can access the raw data, I would urge caution in believing anything in these standard format NEJM trials; happy to be told I am completely wrong; just read the RIAT please before saying so; the Canadian authorities gave the RIAT authors full access to raw data; (after the FDA said it would take 10-15 yrs to release it, as they were SO BUSY): the canadians seemed to the SEND button on their computer for instant release; one assumes that they bitterly regret that candid error; and will regret it to the end of their days; never again will that allow that transparency; we must maintain the illusion behind the glittery curtains and mystery of marvellous medical research.
Thank you for that reference. Most people don't realize how unreliable the cause of death figures are. In many of these studies touting drug treatment it only takes a shift of a few cases to invert the result.
So we can expect a few billion $$$$'s worth of purchase by the future Pfizer/Moderna/InsertPharmaHere executives? Uhhh, I mean FDA/CDC "doctors" and "researchers"?
I agree that the treatment is disease modifying, although in 1 out of 67 patients that fulfill the inclusion criteria. If I had to pay for the drug, I would think twice… Is there any cost-effectiveness analysis done for Select?
Anytime I hear the words “miracle drug” I’m skeptical of long term outcomes. I have heard that GLP1 drugs might be implicated for addiction; we have heard “miracle drug” before - with heroin being one for opiate addiction (yes, you read that correctly!), and on and on. Any disease that can be treated with diet and exercise should be treated no other way.
It is not surprising that a group of professional researchers would grub and dig for some subgroup analysis that would generate another "peer-reviewed" publication. It is, however, disappointing to see the initial study hailed as a breakthrough in the discovery of another disease-modifying drug for heart disease. Judge for yourself whether this was "a clinically important and statistically robust finding".
Here are the percentage figures for the major endpoints:
control% rx% ARR
Composite endpoint 6.5% 8.0% 1.5%
CV Death 2.5% 3.0% 0.5%
Nonfatal MI 2.7% 3.7% 1.0%
Nonfatal stroke 1.7% 1.9% 0.2%
Total mortality 4.3% 5.2% 0.9%
The obvious conclusion is that the drug had no effect that was of any practical significance. Another example of how "statistical significance" and misleading graphs with truncated Y axes are used to fool people and promote useless drugs.
I agree. An absolute risk reduction of 1.5% in a trial of over 17,000 does not appear to be cost effective (except for PHARMA). 8000 out of 8800 on the drug had no event.
Not only not cost effective, but not effective at all. To claim that these tiny differences in composite end points of variable degrees of hardness justify "treatment" is ridiculous.
Interesting. I'd be interested to hear the group's opinion on using the GLP1 for weight loss. I become suspicious when a drug becomes a movement. I have friends and colleagues (we are CICU nurses) using this for weight loss. I'm a bit concerned about the long term effects. Myself, I'm using the calorie counting/exercise route. Very slow at my age but the more natural way to healthier living.
try going low-carb Amy; some call it keto: no sugar, no bread, no rice, no pasta .... no carbs as best you can; doctors are "baffled" how well folks do; no amount of exercise will reduce insulin levels; and insulin STOPS you burning fat; check up on Jason Fung; calorie counting is ineffective Amy; burn fat; you will be amazed how your brain works better; our brain does best on fat.
There are some really good videos on this. The big takeaway for me was that it could be a miracle drug, but that they are majorly overdosing people and that you MUST take advantage of the mental reset and change your habits while you're on it. This video is the best I've seen on it:
https://www.youtube.com/watch?v=XDcESvzX58Y&t=4s
PS: This channel's interviews are long, but I find that they are well worth listening through.
This was my thought exactly when I read about SELECT in Medpage Today. I'm sure that GLP1 would have significantly reduced COVID-19 mortality among naive patients in 2020 due to reduction in obesity, but not so sure about in 2021-2023 especially given that the study didn't seem properly powered to detect an effect of that size.
The ARR is very small here, slightly above 1%. It’s why the study was so large. Unimpressive for a composite endpoint and an expensive drug.
Money
This substudy is insanity and stupidity wrapped into one. And an embarrassment that it’s in JACC.
That said, SELECT itself is practice-changing.
I’m no trialist but I like how hierchical analysis is used. It allows something (potentially) to be scientifically validly concluded upon beyond the primary endpoint. I’d like to understand more btw the strengths and weaknesses of such a trial design, vs a total events analysis (as opposed to time to first event) to allow more of the study data to be utilized.
Regardless, they should’ve stopped when the first hierchical analysis was non-insignificant. And this is probably a hypothesis that isn’t even worth generating.
thanks for doing the heavy lifting - the highlights are indeed misleading-