"Yet the use of this evidence can never be as algorithmic as guidelines or quality measures or news stories make it seem." That statement is the essence of your message! There are so many applications to your message - it could be at the very least a book.
We try to teach students, that guidelines and algorhithms are not mandates, but must be adapted to the needs of the patient, family, and individual situation. Government mandates are even worse! Doctors must have the freedom to make judgements in concert with the patient. There will never be one size fits all in medicine.
I really liked your conclusion. No patient is the same as the "normal" patient. The clinician's job is to translate results as it relates to your current patient. So many people, especially the government, believe that you can tell how the patient will react to a treatment just by what the computer tells you. This is not true. Every patient is different, and a robot should never replace a clinician in the analysis of your patient.
If we can disregard drug trials because different people might react differently, we might as well not do drug trials at all. Each person is uniquely different from all other people in many different ways . The only way we can be investigate whether a drug will work on a given individual is to do a trial with many people who are identical to the person we want to use the drug on. This is impossible. Drug trials with a large number of subjects, all different from one another, can only tell us if the drug works on a population of subjects. It has public health significance but not necessarily relevant to any single individual. If we can't accept this limitation, then we should. just abandon all drug trials rather than conduct them and later point out their obvious uncertain relevance to any given individual.
I imagine there are distinctions related to population that matter. I suppose that is the real dilemma on genetic based drugs that are intended for a small fraction of individuals. We can see via VAERS that the mRNA vaccines affected some people adversely. Without a detailed investigation which I understand has not been done, we don't fully understand the parameters. It would useful to know in advance who might be at risk from the mRNA vaccine and why.
There are too many parameters to to know if we really want to know about any individual.
Sure we can study all those who "identify as black" or who were born with blue eyes. That is a population study. Bur within that population no matter how narrowly defined, there still will be individuals and they all will differ from each other in many, many, .many ways, some ways that are inconceivable to us now but will be. discovered. in the future.
"too many parameters". My hope has been with "big data" in developing an analysis framework. I would hope that the paid public health researchers would be involved in establishing suitable biomarkers that might give insight into immune system responses. Sadly, I don't see how we get to that "in the future" when we haven't bothered to start the analysis. But such work seems critical if we intend to continue technologies like mRNA that involve a complex dance inside a human.
But population studies for genetic drugs intended for a small subset of individuals seems impossible. I assume the authorities have alternate ways to determine effectiveness of proposed treatments for such subsets.
What you have to do is treat each patient. with humility. knowing that. each person is different. Science. can never deal with. all the variables due to changing. external
While I can quite agree with individual attention and care, how does advocating a vaccine for all young children even arrive in the game? Much of the pandemic response has been one-size-all.
But I was expecting that analysts would be teasing out data from VAERS, from reported deaths, etc to see what factors can be postulated as common and start to collect data to understand the illness along with the "cure".
I'm a tech nerd, with no medical/science training, but have a curious mind and a love for data.
Seems like comprehensive RCT's are almost impossible to do with the type of confidence we'd hope for.
Humans are unique in their DNA, their range of physical conditions, etc....but what strikes me as a tough thing to factor in is the consideration of the patient's current pharma and med treatments.
Again, I'm not a practitioner, but trying to understand the interaction(s) of various pharma combinations, in a wide range of human condition seems like a daunting task.
Accepting this, RCTs must be done, but their conclusiveness will not be as strong as we'd like. That's why I think intelligent patient + intelligent holistically oriented medical professional are required. Need both to provide the best assessment and treatment approach. Industrial medicine does not promote this.
Also, the COVID has made me aware that there are multiple dimensions to good medical practice. Patient treatments, public health, socioeconomics, cultural consideration all must be considered.
In COVID, extreme "locking down" might have saved some lives, but the knock-on negative impacts would have been worse (IMHO). We need a holistic approach when dealing with big issues like this.
The nurses, doctors, OTs, PTs, and research scientists in my family (two who also have their MPH) appreciate those like yourself willing to opine and provoke critical thinking and debate regarding medical practice and public health policy/actions. We recognize the complexities and nuances of providing best practice medicine, but we try our best to “check” any arrogance and bias during our debates. We often disagree on EBM interpretations and extrapolations as well as public health decisions. We all agree that trust in public health institutions and regulatory agencies have plummeted and with good reason.
Nefarious is always a little difficult to prove but quite often the enrollment criteria for a trial can indicate intention. For example did Pfizer not enroll vaccinated patients due to availability or were they concerned with performance in this population?
I was part of one of the first BNP tests product launches in the US. We had very narrow and strict enrollment criteria which gave us a ROC curve of .90. When doing a meta analysis of all BNP tests years later you saw a average ROC curve of .83, still a good diagnostic test but below our original study. Enrollment criteria matters and can give a false impression. The VA study for hydroxy that caused it to loose the EUA was due to a bad enrollment criteria and this was absolutely nefarious in my opinion. They knew this would cause the drug to have little benefit. The drug was administered late and to mostly critically ill patients.
This article is spot on. Look at the “troponin” develop in very specific set of patients and circumstances and now drawn on everyone. How many people has a false troponin lead to procedures and complications. The “d-dimer” how much damage has that done when released broadly. PFO……we see studies looking for that in 88 year old patients……so sad how things that may help a few folks, when released get generalized and weaponized against patients, providers i guess in mainly the fear of “not doing enough”
This, to me, is not a good thing. It is partly why we have so much experimental polypharmacy. Doctors prescribe all of these meds I would suppose from studies that do not reflect the real world. Patients need the real world results, not profiteering theatrics. The research nurses and pharmacists should know about safety signals and contraindications within the trials. And doctors should get all of the data. The pfizer vaccine data dump at PHMPT.ORG is a perfect example of important data withheld from doctors, that should never be allowed. And yet here we are...
if those doctors do not seem to know about the adverse effects of statins, statins being about the most prescribed drug there is, imagine what they know about the other more obscure drugs, or about ANYTHING for that matter...
I'm healthy as a horse. One blood test I forgot to skip the cream in my coffee that morning and my HDL came in a bit higher than the "acceptable" range. Looking at my results through the portal I saw the boxed tip to the GP to "discuss statins" with patient.
My husband has bp that gets him lectured about beta blockers etc: 140/80. No sale: those drugs cause dementia over the long term My BP is 125/60, and now I'm reading that diastolic should be above 70 to avoid dementia from low blood flow. You either learn to kick the tires or you end up a lab rat.
Thank for this extremely insightful analysis. It is what I would assume and you have explained well with eveidence. I will save this for future reference.
"Yet the use of this evidence can never be as algorithmic as guidelines or quality measures or news stories make it seem." That statement is the essence of your message! There are so many applications to your message - it could be at the very least a book.
We try to teach students, that guidelines and algorhithms are not mandates, but must be adapted to the needs of the patient, family, and individual situation. Government mandates are even worse! Doctors must have the freedom to make judgements in concert with the patient. There will never be one size fits all in medicine.
Wonderful piece. I appreciated the examples from other areas of medicine as well.
I really liked your conclusion. No patient is the same as the "normal" patient. The clinician's job is to translate results as it relates to your current patient. So many people, especially the government, believe that you can tell how the patient will react to a treatment just by what the computer tells you. This is not true. Every patient is different, and a robot should never replace a clinician in the analysis of your patient.
interesting, thank you.
p.s. note extraneous word in; They were young (age 64), mostly male, non-obese, patients who **were** could tolerate medical therapy of heart failure.
And yet our Government spent 1.6 billion on Paxlovid and still advocating its use.
If we can disregard drug trials because different people might react differently, we might as well not do drug trials at all. Each person is uniquely different from all other people in many different ways . The only way we can be investigate whether a drug will work on a given individual is to do a trial with many people who are identical to the person we want to use the drug on. This is impossible. Drug trials with a large number of subjects, all different from one another, can only tell us if the drug works on a population of subjects. It has public health significance but not necessarily relevant to any single individual. If we can't accept this limitation, then we should. just abandon all drug trials rather than conduct them and later point out their obvious uncertain relevance to any given individual.
I imagine there are distinctions related to population that matter. I suppose that is the real dilemma on genetic based drugs that are intended for a small fraction of individuals. We can see via VAERS that the mRNA vaccines affected some people adversely. Without a detailed investigation which I understand has not been done, we don't fully understand the parameters. It would useful to know in advance who might be at risk from the mRNA vaccine and why.
There are too many parameters to to know if we really want to know about any individual.
Sure we can study all those who "identify as black" or who were born with blue eyes. That is a population study. Bur within that population no matter how narrowly defined, there still will be individuals and they all will differ from each other in many, many, .many ways, some ways that are inconceivable to us now but will be. discovered. in the future.
"too many parameters". My hope has been with "big data" in developing an analysis framework. I would hope that the paid public health researchers would be involved in establishing suitable biomarkers that might give insight into immune system responses. Sadly, I don't see how we get to that "in the future" when we haven't bothered to start the analysis. But such work seems critical if we intend to continue technologies like mRNA that involve a complex dance inside a human.
But population studies for genetic drugs intended for a small subset of individuals seems impossible. I assume the authorities have alternate ways to determine effectiveness of proposed treatments for such subsets.
What you have to do is treat each patient. with humility. knowing that. each person is different. Science. can never deal with. all the variables due to changing. external
circumstances
While I can quite agree with individual attention and care, how does advocating a vaccine for all young children even arrive in the game? Much of the pandemic response has been one-size-all.
But I was expecting that analysts would be teasing out data from VAERS, from reported deaths, etc to see what factors can be postulated as common and start to collect data to understand the illness along with the "cure".
Great article and discussion -thank you Dr. John.
I'm a tech nerd, with no medical/science training, but have a curious mind and a love for data.
Seems like comprehensive RCT's are almost impossible to do with the type of confidence we'd hope for.
Humans are unique in their DNA, their range of physical conditions, etc....but what strikes me as a tough thing to factor in is the consideration of the patient's current pharma and med treatments.
Again, I'm not a practitioner, but trying to understand the interaction(s) of various pharma combinations, in a wide range of human condition seems like a daunting task.
Accepting this, RCTs must be done, but their conclusiveness will not be as strong as we'd like. That's why I think intelligent patient + intelligent holistically oriented medical professional are required. Need both to provide the best assessment and treatment approach. Industrial medicine does not promote this.
Also, the COVID has made me aware that there are multiple dimensions to good medical practice. Patient treatments, public health, socioeconomics, cultural consideration all must be considered.
In COVID, extreme "locking down" might have saved some lives, but the knock-on negative impacts would have been worse (IMHO). We need a holistic approach when dealing with big issues like this.
The nurses, doctors, OTs, PTs, and research scientists in my family (two who also have their MPH) appreciate those like yourself willing to opine and provoke critical thinking and debate regarding medical practice and public health policy/actions. We recognize the complexities and nuances of providing best practice medicine, but we try our best to “check” any arrogance and bias during our debates. We often disagree on EBM interpretations and extrapolations as well as public health decisions. We all agree that trust in public health institutions and regulatory agencies have plummeted and with good reason.
Nefarious is always a little difficult to prove but quite often the enrollment criteria for a trial can indicate intention. For example did Pfizer not enroll vaccinated patients due to availability or were they concerned with performance in this population?
I was part of one of the first BNP tests product launches in the US. We had very narrow and strict enrollment criteria which gave us a ROC curve of .90. When doing a meta analysis of all BNP tests years later you saw a average ROC curve of .83, still a good diagnostic test but below our original study. Enrollment criteria matters and can give a false impression. The VA study for hydroxy that caused it to loose the EUA was due to a bad enrollment criteria and this was absolutely nefarious in my opinion. They knew this would cause the drug to have little benefit. The drug was administered late and to mostly critically ill patients.
This article is spot on. Look at the “troponin” develop in very specific set of patients and circumstances and now drawn on everyone. How many people has a false troponin lead to procedures and complications. The “d-dimer” how much damage has that done when released broadly. PFO……we see studies looking for that in 88 year old patients……so sad how things that may help a few folks, when released get generalized and weaponized against patients, providers i guess in mainly the fear of “not doing enough”
Brilliant! Absolutely brilliant - the right tone and pitch to share with my family and friends to awaken curiosity. Thank you 😊🙏
never stop learning, (now at the age of 77)
Again underscores the importance of evidence based practice and practitioners.
This, to me, is not a good thing. It is partly why we have so much experimental polypharmacy. Doctors prescribe all of these meds I would suppose from studies that do not reflect the real world. Patients need the real world results, not profiteering theatrics. The research nurses and pharmacists should know about safety signals and contraindications within the trials. And doctors should get all of the data. The pfizer vaccine data dump at PHMPT.ORG is a perfect example of important data withheld from doctors, that should never be allowed. And yet here we are...
like with the statins.
the chance of benefit is near to negligible, the risk of side effects is 100%.
if those doctors do not seem to know about the adverse effects of statins, statins being about the most prescribed drug there is, imagine what they know about the other more obscure drugs, or about ANYTHING for that matter...
I'm healthy as a horse. One blood test I forgot to skip the cream in my coffee that morning and my HDL came in a bit higher than the "acceptable" range. Looking at my results through the portal I saw the boxed tip to the GP to "discuss statins" with patient.
My husband has bp that gets him lectured about beta blockers etc: 140/80. No sale: those drugs cause dementia over the long term My BP is 125/60, and now I'm reading that diastolic should be above 70 to avoid dementia from low blood flow. You either learn to kick the tires or you end up a lab rat.
Thank for this extremely insightful analysis. It is what I would assume and you have explained well with eveidence. I will save this for future reference.