I watched my older sister go through menopause and witnessed the nightmare roller coaster ride she took. I swore I would not go through that. Three months after my last menstruation (15 years ago) I immediately got bioidentical hormone replacement. E2, E3 & progesterone. I’m sailing through menopause with none of the typical symptoms, not even a hot flash. My hormone levels are tested annually. My heart is great and my bones are great. Just sayin…
Were those clinical trials above only on synthetic hormones. I wonder if a comparison has been done on synthetic vs bio identical.
Women’s health and heathy aging need a lot more research to assess what really benefits individuals. For now, I will continue my estrogen, exercise and “food is medicine” approach.
Here's evidence that the allegation of smearing of hydroxychloroquine by the FDA is a conspiracy theory.
"FDA cautions against use of hydroxychloroquine or chloroquine for COVID-19 outside of the hospital setting or a clinical trial due to risk of heart rhythm problems"
It's irrelevant that primary care providers and rheumatologists have prescribed hydroxychloroquine for millions of people as outpatients, taking billions of doses for _years_, including people with heart disease.
Somehow taking HCQ for five days at clinical dosing is dangerous in the context of covid for people without heart disease as outpatients, when it's not dangerous for people with heart disease outside of the covid context (e.g., for lupus or rheumatoid arthritis) to take HCQ as outpatients for years, according to the FDA. And we know that the FDA can be trusted to be reliable and competent.
So, clearly, the allegation that the FDA smeared hydroxychloroquine is a conspiracy theory.
"Recently, the FDA commissioned a pilot project: RCT-DUPLICATE to do just this. Sadly, the result was suboptimal. Less than 50% of RCTs could be reliably emulated or replicated with this design. More work is needed."
- Without numerous details, the 50% is worse than meaningless: It's a misleading expression of overconfidence in a single and arbitrary number. What criterion was used to draw such a sharp distinction between "reliable emulation" (or replication) and not? I saw no clue here or in the linked Endpoints article.
This much I'll wager: The 50% claim is one that could be shifted all over the place by changing the criteria and analysis in defensible ways. The "more work needed" is to fight dichotomania - the compulsion to reduce continuous phenomena to arbitrary dichotomies - and to stop quoting unclear and degraded summaries as if they were revealed methodologic facts. In this case the most pressing need is to provide instead more informative summaries, such as interval estimates from emulations plotted or tabulated in detail against those from the corresponding RCTs, and direct interval estimates of the differences between the studies.
Also needed is recognition that the RCTs are not the gold standard they are treated as: Their selective enrollment guarantees that they won't be estimating the same effect as what will be seen in practice when patients are no longer carefully cherry picked to avoid adverse events and experience benefits. RCTs can and often suffer their own biases due to post-randomization withdrawals, lack of masking (blinding), and misanalyses. These problems mean that emulation meta-studies are only looking at associations of select observational analyses with trials. Those associations will be far from perfect, not only from bias and errors in both the observational data and the trials, but also from real differences in the effects or questions being targeted.
Yes, statistical adjustments for these heterogeneity sources can and should be made to the extent that the sources were measured in the studies. But, for guiding practice, even the best statistical analyses cannot by themselves tell us whether a disagreement reflects shortcomings of the emulations or of the trials or both. That is a question about what caused the disagreement, and as such requires causal assumptions that are best made explicit and realistic, rather than hidden or simplistic (like assuming the RCT is giving the "right" answer). And when more credible assumptions are made, I'll wager again that the measurement and adjustments are far from sufficient to enable any declaration remotely approaching the certainty reflected in "less than 50% of RCTs could be reliably emulated or replicated with this design". Rather, the latter kind of statement reflects the primitive state of meta-research reporting (and I fear, conduct) outside of the most advanced methodologic literature.
Sep 23, 2022·edited Sep 23, 2022Liked by John Mandrola
This article is the perfect example of what I was hoping to find in the Sensible Medicine substack! Dr Mandrola wrote an interesting article, others had salient comments, and there was a good discussion. And I gained new insights.
For me, the best part of this post was Dr. Mandrola's suggestion that all studies should have an expiration date. Am always astonished that many relatively ancient RCTs and analyses are still widely cited, and are staunchly believed to be the final law-giver in this or that domain. No matter that experts like Dr. Mandrola and Dr. Prasad frequently point out that reversals of medical protocols occur regularly. Wish that there were more - would be an indicator of an alert and ever-curious medical sector.
Most readers are probably too young to remember the super-hyped 'Feminine Forever' marketing wave. It shocked me then, and still resonates. My own view on HRT is simply that no potent treatment should ever be given anyone UNLESS the omission of that treatment clearly puts the patient in serious danger. But of course, as with liposuctions, botox, cosmetic surgery, and so many other elective interventions, it seems that societal pressure and marketing prowess rule the day.
Bravo David Allely! You have provided the best summary and accurate analysis of WHI. And you’re not even an MD yet. Thank you. ERT started prior to menopause in a woman with total hysterectomy has many benefits. WHI was poorly interpreted at the time.
I remember when I turned 50 in 1999 and my PCP said I had to go on HRT to prevent a heart attack and I declined saying I would rather die suddenly from a heart attack than suffer from breast cancer. She informed me the insurance company would ding her if I refused. I said take the hit for me and mark I adamantly refused, was thankful I did when the 2002 study halted this push on women with so symptoms or other benefit. 73 now and heart healthy but still needing ultrasounds every year after each 3 D mammogram,
It's important to understand the opposing side's position. Consequently, I have tried to wrap my head around the HCQ-skeptic position. Here is what it seems to be.
1. Primary care providers typically have problems with followup. They may not see negative outcomes. Physicians whose patients are in hospitals have a better opportunity to see negative outcomes (hospitalization, death, chronic morbidity caused by covid).
2. Mild covid is difficult to diagnose because of viral respiratory disease confounders which have similar symptoms (flu, syncitial virus, etc.); PCR shows exposure, not necessarily active disease; do PCPs really do lung scans or lab work to nail down the diagnosis; hence, what are they actually treating with HCQ???
3. I have tried HCQ on my (hospitalized) patients and it doesn't work. They still died. [No, I am not a physician. I am imagining what a physician HCQ-skeptic would find compelling.]
4. There is no compelling RCT proof that HCQ works anywhere. [There is controversy here, of course.]
5. HCQ was pushed because of politics. [More like the skeptical position was political in nature, but this is what some HCQ skeptics _feel_.]
Have I missed anything in the HCQ-skeptic position? Please enlighten me.
When they take a natural substance and alter it so that it can be patented and then think it will have the same effect on the human body as the natural substance does, it is a mistake. Unfortunately the studies seem to not understand that estrogen from horse urine is not the same and has components that human estrogen does not. A study on bio identical hormones might show very different results. The fact that this is never mentioned is a typical example of misleading science.
Keep the drugs with NNTT ~= 1 (NNTT = "Number Needed to Treat to get the endpoint benefit")
Pitch the rest
NB
Assuming competency in primary care (e.g., Brian Tyson and George Fareed of El Centro Clinic in Califormia and a few dozen others around the US and outside it), HCQ is a parachute for high risk covid patients when used to treat patients within 72 hours of symptom onset--treatment is based on suspicion, with lab work and lung scans being used to provide additional evidence to support the diagnosis based on suspicion. (This is exactly the protocol that the CDC recommends for high risk influenza patients--treat early with antivirals based on suspicion, so it's hardly revolutionary.)
Black Swan Challenge I (for early treatment of covid with HCQ)
Find one primary care physician who has tried treating high risk covid patients with HCQ early and has discontinued it and says it doesn't work. I have been looking since May 2020 for one and no black swans have yet been sighted. (No, hospital physicians don't count. They treat _late_ because they see patients _late_, after covid has already progressed.)
Black Swan Challenge II (to support the proposition that HCQ is dangerous for cardiac patients)
Find me one rheumatologist who won't prescribe HCQ for his or her patients with heart disease. Or find me one PCP whose patient(s) have died from heart failure while being treated for covid with HCQ.
So, anonymous MD, MPH, MAH, I may not be a clinician but I sure as hell try to understand their point of view. (MAH = "Master of Ad Hominem")
in 2013 I moved both of my parents, age 88 closer to my home after mom's stroke
I then, took them to my wonderful doctor Dr Fletcher, and we worked to get mom off of half of her meds that were piled on her by various doctors. Then, I took dad to Dr Ujevic and did the same.
Mom lived to 2016 and died of natural causes. Daddy is 98 and still going steady.
I have learned a lot over the years from taking care of my peeps
The bar should be high when "treating" otherwise healthy people with pharmaceutical drugs. Trying to work against the body's genetic programming and natural design is hubris. Menopause is there for a reason. So we don't get pregnant at 65! Reminds me of Fosamax. My mom was was told she should take both hormones and Fosamax. Thank goodness she doesn't fall for that stuff. She is "non-compliant". It seems there is a new push for hormones in the media.
I watched my older sister go through menopause and witnessed the nightmare roller coaster ride she took. I swore I would not go through that. Three months after my last menstruation (15 years ago) I immediately got bioidentical hormone replacement. E2, E3 & progesterone. I’m sailing through menopause with none of the typical symptoms, not even a hot flash. My hormone levels are tested annually. My heart is great and my bones are great. Just sayin…
Were those clinical trials above only on synthetic hormones. I wonder if a comparison has been done on synthetic vs bio identical.
Women’s health and heathy aging need a lot more research to assess what really benefits individuals. For now, I will continue my estrogen, exercise and “food is medicine” approach.
Conspiracy theorist Dr. Andrew Huff, former VP of Ecohealth Alliance, claims that NIH funded development of SARS-COV-2.
https://myemail.constantcontact.com/EcoHealth-Alliance-Newsletter.html?soid=1109104170770&aid=B8ZzloGYn_A
Clearly, Dr. Huff couldn't have known anything about what happened at the Wuhan lab. He's a conspiracy theorist. /sarc
Here's evidence that the allegation of smearing of hydroxychloroquine by the FDA is a conspiracy theory.
"FDA cautions against use of hydroxychloroquine or chloroquine for COVID-19 outside of the hospital setting or a clinical trial due to risk of heart rhythm problems"
https://www.fda.gov/drugs/drug-safety-and-availability/fda-cautions-against-use-hydroxychloroquine-or-chloroquine-covid-19-outside-hospital-setting-or
It's irrelevant that primary care providers and rheumatologists have prescribed hydroxychloroquine for millions of people as outpatients, taking billions of doses for _years_, including people with heart disease.
Somehow taking HCQ for five days at clinical dosing is dangerous in the context of covid for people without heart disease as outpatients, when it's not dangerous for people with heart disease outside of the covid context (e.g., for lupus or rheumatoid arthritis) to take HCQ as outpatients for years, according to the FDA. And we know that the FDA can be trusted to be reliable and competent.
So, clearly, the allegation that the FDA smeared hydroxychloroquine is a conspiracy theory.
Sorry, I cringe when I see comments like this:
"Recently, the FDA commissioned a pilot project: RCT-DUPLICATE to do just this. Sadly, the result was suboptimal. Less than 50% of RCTs could be reliably emulated or replicated with this design. More work is needed."
- Without numerous details, the 50% is worse than meaningless: It's a misleading expression of overconfidence in a single and arbitrary number. What criterion was used to draw such a sharp distinction between "reliable emulation" (or replication) and not? I saw no clue here or in the linked Endpoints article.
This much I'll wager: The 50% claim is one that could be shifted all over the place by changing the criteria and analysis in defensible ways. The "more work needed" is to fight dichotomania - the compulsion to reduce continuous phenomena to arbitrary dichotomies - and to stop quoting unclear and degraded summaries as if they were revealed methodologic facts. In this case the most pressing need is to provide instead more informative summaries, such as interval estimates from emulations plotted or tabulated in detail against those from the corresponding RCTs, and direct interval estimates of the differences between the studies.
Also needed is recognition that the RCTs are not the gold standard they are treated as: Their selective enrollment guarantees that they won't be estimating the same effect as what will be seen in practice when patients are no longer carefully cherry picked to avoid adverse events and experience benefits. RCTs can and often suffer their own biases due to post-randomization withdrawals, lack of masking (blinding), and misanalyses. These problems mean that emulation meta-studies are only looking at associations of select observational analyses with trials. Those associations will be far from perfect, not only from bias and errors in both the observational data and the trials, but also from real differences in the effects or questions being targeted.
Yes, statistical adjustments for these heterogeneity sources can and should be made to the extent that the sources were measured in the studies. But, for guiding practice, even the best statistical analyses cannot by themselves tell us whether a disagreement reflects shortcomings of the emulations or of the trials or both. That is a question about what caused the disagreement, and as such requires causal assumptions that are best made explicit and realistic, rather than hidden or simplistic (like assuming the RCT is giving the "right" answer). And when more credible assumptions are made, I'll wager again that the measurement and adjustments are far from sufficient to enable any declaration remotely approaching the certainty reflected in "less than 50% of RCTs could be reliably emulated or replicated with this design". Rather, the latter kind of statement reflects the primitive state of meta-research reporting (and I fear, conduct) outside of the most advanced methodologic literature.
This article is the perfect example of what I was hoping to find in the Sensible Medicine substack! Dr Mandrola wrote an interesting article, others had salient comments, and there was a good discussion. And I gained new insights.
Well done!
For me, the best part of this post was Dr. Mandrola's suggestion that all studies should have an expiration date. Am always astonished that many relatively ancient RCTs and analyses are still widely cited, and are staunchly believed to be the final law-giver in this or that domain. No matter that experts like Dr. Mandrola and Dr. Prasad frequently point out that reversals of medical protocols occur regularly. Wish that there were more - would be an indicator of an alert and ever-curious medical sector.
Most readers are probably too young to remember the super-hyped 'Feminine Forever' marketing wave. It shocked me then, and still resonates. My own view on HRT is simply that no potent treatment should ever be given anyone UNLESS the omission of that treatment clearly puts the patient in serious danger. But of course, as with liposuctions, botox, cosmetic surgery, and so many other elective interventions, it seems that societal pressure and marketing prowess rule the day.
Bravo David Allely! You have provided the best summary and accurate analysis of WHI. And you’re not even an MD yet. Thank you. ERT started prior to menopause in a woman with total hysterectomy has many benefits. WHI was poorly interpreted at the time.
I remember when I turned 50 in 1999 and my PCP said I had to go on HRT to prevent a heart attack and I declined saying I would rather die suddenly from a heart attack than suffer from breast cancer. She informed me the insurance company would ding her if I refused. I said take the hit for me and mark I adamantly refused, was thankful I did when the 2002 study halted this push on women with so symptoms or other benefit. 73 now and heart healthy but still needing ultrasounds every year after each 3 D mammogram,
It's important to understand the opposing side's position. Consequently, I have tried to wrap my head around the HCQ-skeptic position. Here is what it seems to be.
1. Primary care providers typically have problems with followup. They may not see negative outcomes. Physicians whose patients are in hospitals have a better opportunity to see negative outcomes (hospitalization, death, chronic morbidity caused by covid).
2. Mild covid is difficult to diagnose because of viral respiratory disease confounders which have similar symptoms (flu, syncitial virus, etc.); PCR shows exposure, not necessarily active disease; do PCPs really do lung scans or lab work to nail down the diagnosis; hence, what are they actually treating with HCQ???
3. I have tried HCQ on my (hospitalized) patients and it doesn't work. They still died. [No, I am not a physician. I am imagining what a physician HCQ-skeptic would find compelling.]
4. There is no compelling RCT proof that HCQ works anywhere. [There is controversy here, of course.]
5. HCQ was pushed because of politics. [More like the skeptical position was political in nature, but this is what some HCQ skeptics _feel_.]
Have I missed anything in the HCQ-skeptic position? Please enlighten me.
When they take a natural substance and alter it so that it can be patented and then think it will have the same effect on the human body as the natural substance does, it is a mistake. Unfortunately the studies seem to not understand that estrogen from horse urine is not the same and has components that human estrogen does not. A study on bio identical hormones might show very different results. The fact that this is never mentioned is a typical example of misleading science.
These comments are great! I like them better than the posts.
Keep the drugs with NNTT ~= 1 (NNTT = "Number Needed to Treat to get the endpoint benefit")
Pitch the rest
NB
Assuming competency in primary care (e.g., Brian Tyson and George Fareed of El Centro Clinic in Califormia and a few dozen others around the US and outside it), HCQ is a parachute for high risk covid patients when used to treat patients within 72 hours of symptom onset--treatment is based on suspicion, with lab work and lung scans being used to provide additional evidence to support the diagnosis based on suspicion. (This is exactly the protocol that the CDC recommends for high risk influenza patients--treat early with antivirals based on suspicion, so it's hardly revolutionary.)
Black Swan Challenge I (for early treatment of covid with HCQ)
Find one primary care physician who has tried treating high risk covid patients with HCQ early and has discontinued it and says it doesn't work. I have been looking since May 2020 for one and no black swans have yet been sighted. (No, hospital physicians don't count. They treat _late_ because they see patients _late_, after covid has already progressed.)
Black Swan Challenge II (to support the proposition that HCQ is dangerous for cardiac patients)
Find me one rheumatologist who won't prescribe HCQ for his or her patients with heart disease. Or find me one PCP whose patient(s) have died from heart failure while being treated for covid with HCQ.
So, anonymous MD, MPH, MAH, I may not be a clinician but I sure as hell try to understand their point of view. (MAH = "Master of Ad Hominem")
less pills, and life is less complicated.
in 2013 I moved both of my parents, age 88 closer to my home after mom's stroke
I then, took them to my wonderful doctor Dr Fletcher, and we worked to get mom off of half of her meds that were piled on her by various doctors. Then, I took dad to Dr Ujevic and did the same.
Mom lived to 2016 and died of natural causes. Daddy is 98 and still going steady.
I have learned a lot over the years from taking care of my peeps
The bar should be high when "treating" otherwise healthy people with pharmaceutical drugs. Trying to work against the body's genetic programming and natural design is hubris. Menopause is there for a reason. So we don't get pregnant at 65! Reminds me of Fosamax. My mom was was told she should take both hormones and Fosamax. Thank goodness she doesn't fall for that stuff. She is "non-compliant". It seems there is a new push for hormones in the media.