Why do doctors not tell their patients to be aware of symptoms of shingles and, if they have them, to call and get an Rx for Valtrex and take it immediately?! It’s similar to Malarone to prevent malaria when traveling—the same drug is used to treat malaria, so why not wait and take it IF you need it? Enough with these vaccines—which have side effects for many people—when an easy to get treatment is available.
Having just yesterday got part two of my shingles vaccine I was doubly disappointed here. Apparently I am no less likely to get dementia and Dr Prasad won’t reveal the important bit. Is it good at preventing shingles? I had shingles about eight years ago and didn’t much enjoy it.
This was my feeble attempt at a joke. Clearly it failed, and probably deserved to. But he does sort of gesture at the question in the penultimate paragraph. But I agree. This was not a piece about shingles!
Well, my spouse did. Her experience with shingles was a miserable one. The shots were easy for her. I had the original vax then the first dose of the updated vax. That was miserable for me. And then Covid hit about a month later. Never got second dose but just the one is about 85-90% protective if I recall. Second does gets it up north of 95%.
That puzzled me but apparently you can get it again. Seems bizarre then that vaccines help better than getting it but I don't really understand vaccines all that well. They don't all seem to work the same way.
in the late 70's, I worked for the great modern dance choreographer Martha Graham. when she came down with shingles, I had never heard of it before. I thought it was a literary "disease" that old maids got in Dickens' novels. then in 2004, I met or heard about 3 young men, all 23, who had shingles. one was a dancer who had to miss his opening night while the crew raced to alter his costumes for his understudy in June. the second was a young stagehand who was a friend of my boyfriend's and was just recovering at a 4th of July BBQ. he said it was awful. I heard about a third 23 year old coming down with shingles in August. what did they have in common, other than their age and gender? they had all been born in 1981, and after vaccines were given liability protection in 1986 and after the varicella vaccine was routinely given to children, NONE OF THEM HAD BEEN ALLOWED TO HAVE WILD TYPE CHICKEN POX AND DEVELOP REAL IMMUNITY TO IT.
my theory was that not having chicken pox at the appropriate age because of a vaccine intervention predisposes one to getting shingles later in life. or in the case of these young men- not all that later (they were all 27 years too young to qualify for the shingles vaccine).
I subsequently read that in the UK, the varicella vaccine is not recommended because the National Health system there would prefer to avoid the shingles epidemic that follows the lack of having chicken pox as a child.
when are hubristic "scientists" going to learn that nature works best if left alone? healthy children can easily handle these infections. adults do less well with shingles. this is a man made "epidemic."
What is the bigger problem is that people are not being "reinnoculated" against chicken pox by being exposed to kids who have full-blown cases. That was another thing (or maybe the primary thing) the UK was concerned about. When you vaccinate a kid, the elders around them are more likely to get shingles. Of course, this is quite convenient for vaccine manufacturers who make a profit on boosters.
you are 100,000% correct. the immunity, superior to vaccine immunity though it is, still needs to be topped up now and then. you have children and they get measles or chicken pox and you're exposed even though you don't get sick. you're exposed to your nieces and nephews and cousins and it just "tops up" your immunity. you're exposed to your sick grandchildren.
then one day, they put you in a nursing home and your immunity wanes and they certainly don't bring your great grandchildren in to visit you when they're sick! the next thing you know, you've got shingles.
the best thing that could happen- and it will take a few generations- is for all kids to get natural measles, mumps and chickenpox at an appropriate age and to go back to exposures that reinforce natural immunity.
instead we have hysterics screaming over every case of measles like it's the end of the world.
This commentary comes pretty close to declaring this study, or at least its claims, to be fraudulent, starting with another study that was clearly so. The nicest thing it says is that the claims are “magical”. What is closer to magical thinking is this commentary. Dr. Prasad substitutes his own words – better described as unfounded accusations - for the actual claims of the authors, no less their methods. Let’s look at the article’s “claims”, taken verbatim:
“The observed reduction of risk is clinically meaningful: if confirmed in clinical trials…”
“The biological mechanisms underlying these observations remain to be clarified experimentally.”
“Third, although our study is a natural experiment … it does not imply that associations are causal. Causality requires additional assumptions (outlined in the Methods alongside supporting evidence and discussion of remaining uncertainties).”
“Early divergence of Kaplan−Meier curves, while biologically plausible, can be a sign of unmeasured confounding..."
“…these findings, if confirmed in clinical trials and mechanistic studies, would have important implications for public health.”
In Methods, under “Assumptions for causal inference”:
“Although the above evidence supports exchangeability on observed dimensions, the influence of several unmeasured factors cannot be ruled out. … If they systematically differ between cohorts after matching, such unmeasured confounders may violate the exchangeability assumption.” They then include a variety of possibly important covariates that they could not measure, including particularly social determinants of health.
Every point raised by Dr. Prasad is discussed in the paper, some at considerable length in both the discussion and the methods. This includes the rapid divergence of the curves, whose explanation the authors concede is not evident, even though they did a variety of analyses, including 6 negative control outcomes, to explore obvious methodologic artifacts. The proposed explanations are framed as tentative; some speculative, some supported by cited evidence.
At the end of the day, the authors end up with the “claim” that the phenomenon is merely worth further exploration with stronger empirical designs (i.e. trials) and laboratory studies of mechanism. That’s it. Their claim is that the phenomenon deserves a closer and better look. Is that a universal “get out of inference free” card that protects against spurious claims? Only if the authors had not engaged in the serious methodologic efforts they did, which go unmentioned and unrebutted by Dr. Prasad.
I will add parenthetically that one of the more challenging aspects of such research is the TriNetX database itself, but that is an issue for another day. And if this turns out to be true, it's main value will be in mechanistic understanding, as the motivation to get this vaccine will always be to avoid shingles, not to improve heart health.
One could still legitimately take issue with the article's modest claims - i.e. that this hypothesis is worth further attention - and whether the many caveats are strong enough. But not using the hand-wavy, unidimensional arguments put forward by Dr. Prasad, not to mention his dismissive and almost contemptuous tone. Reasonable scientists can and will disagree as to how likely these data patterns are to represent causal relationships – which the authors do not claim and themselves may not agree on – but Dr. Prasad should engage with what the authors actually did and said, not a cartoon version. But that would take time and wouldn’t be as clickbaitable.
I would encourage every reader of this substack to read the actual paper to reflect not only on the strength of evidence provided by this study, but on the intellectual honesty of Dr. Prasad’s critique. And then decide which is more "magical".
remember when they came out with a "study" saying that covid vaccinated drivers had fewer accidents? Dr Prasad was pretty critical of that "study" too. the vaccine cartel is always trying to expand all the "miracles" that vaccines do, except when it comes to increasing autism and other developmental disorders in children. that, of course, is simply "coincidence."
“Third, although our study is a natural experiment … it does not imply that associations are causal. Causality requires additional assumptions (outlined in the Methods alongside supporting evidence and discussion of remaining uncertainties).”
Well that was big of them.
Causality requires a different study design, not more assumptions using retrospective data. This is the problem with studies that use propensity-score matching and TTE. In the end, you have associations using retrospective data from TriNetX or wherever that have been souped-up statistically to make it seem, after all the caveats & qualifiers & assumptions & corrections etc. etc., like you have a basis for causality. But listen: You can't know what you can't know, & in a non-randomized study there is plenty.
Why does everyone think retrospective observational studies are so bad? They have their place. They can be hypothesis generating. But this whole trend toward trying to dress them up as something resembling a study that can even approximate causality reminds me of people who buy tiny houses & then talk about all the stuff they did to make it seem bigger. Big windows! Tall ceilings! It even has a bathtub! yadda yadda. Look, you live in a TINYhouse, aka a "tiny HOUSE." If space was such a priority, you should have just bought a bigger house. But all this talk about how a tall ceiling makes a 200-sf place seem big, I mean, don't piss on my leg & say it's raining. Just find a way to do an RCT or else be satisfied with the modest conclusions you were able to reach & let someone who is better funded build on that.
"Reasonable scientists can and will disagree as to how likely these data patterns are to represent causal relationships" - what an enormous waste of time. Do the study & get it overwith. Oh, no money for an RCT. Well I'll sell you a TTE real cheap!
Thanks Vinay. Couple of things - 1. where possible can you include more info on the charts/graphs you present. I can mostly figure them out, but it takes a lot of work. (And I guess I don't really understand what SMD is in table 1 - I plugged the numbers in for difference in % white 81% vs 76% and get a z of 21.4 on the test of two sample independent proportions, making p<.00000 but SMD is reported at .13. 2. Whatever SMD is, I can see that there are (significant?) differences in the two populations, but most of them are really really small and at the ends of the % scale where there is a lot of power to detect differences. And, while it would certainly be fair to say the groups are different, the variables presented are not the ones that I think of as really confounding these kinds of comparison i.e., SES type variables or those related to neuroticism or conscientiousness.
I realize that it is probably an undue burden to expect the critic to come up with the "story" of the confounding, but you have done it in other cases and I think that ease and plausibility of generating the story is useful in many cases (e.g, conscientiousness is the third variable explaining vaccination and auto accidents).
There is only one shingles vaccine available in the US since 2021 -Shingrix which is recombinant. Before 2021 there was also a live, attenuated vaccine but it has not been available for years. The vaccine is quite effective, increasingly so as one ages which is unusual. But most people get some kind of side effect when they take it (not major except for very rare Guillain-Barré syndrome, but not pleasant either) and it takes two shots three to six months apart to actually be decently effective. But most would agree better than getting shingles in the final analysis.
Do you know the % of these shingle shots working? I’ve had 2 friends and my brother all get the shingles during covid. All 3 had the double shots. I’m hoping it was just coincidental and the % is high.
My observational, non-medical take is that “spiking” the immune system per COVID shots led to a rise in Shingles. A vaxxed and boosted 50 year old friend came down with Shingles. Her primary care doctor told her to quit the boosters. He was seeing a bunch of Shingles cases.
Shingrix in ZOE-50 and ZOE-70 produced 97.2% efficacy against zoster in adults 50 and older and 89.8% in those 70 and older, with about 91% efficacy against PHN (Post herpetic neuralgia...the really crummy part); the ZOE long-term follow-up cohort still showed roughly 79% to 82% efficacy at ten to eleven years post-vaccination.
thank you for this. Honest question, what does the 97% efficacy rate mean? percentage of people who dont get shingles compared to same age cohort who dont get the vaccine? i understand ARR, RRR and NNT but in this case im confused on how efficacy is defined. thank you again. ( Guillain Barre would be terrible.)
Vaccine efficacy is simly a relative risk reduction, calculated on incidence rates rather than cumulative proportions: VE = 1 − (rate in vaccinated / rate in placebo), where rate is confirmed zoster cases per person-year of follow-up in the randomized trial.
In ZOE-50 the numbers were 6 cases in about 7,300 vaccinees versus 210 in about 7,400 placebo recipients over a mean 3.2 years of follow-up, giving incidence rates of roughly 0.3 versus 9.1 per 1,000 person-years; 1 − 0.3/9.1 ≈ 0.97, hence 97.2%. So yes, it is the vaccinated cohort compared against a same-age randomized placebo cohort, but it tells you about the proportion of expected cases that were prevented, not the proportion of vaccinees protected.
The absolute framing is what you would expect from the RRR: ARR of about 8.8 cases per 1,000 person-years, so an NNT of roughly 114 person-years, or about 35 people vaccinated over the 3.2-year trial window to prevent one case of zoster (larger for PHN, since that is a fraction of cases).
Note that this is per-protocol efficacy in trial conditions, and "effectiveness" is the corresponding term for real-world observational estimates, which for Shingrix have run in the 80s to low 90s rather than 97%. But still impressive as these things go.
Why do doctors not tell their patients to be aware of symptoms of shingles and, if they have them, to call and get an Rx for Valtrex and take it immediately?! It’s similar to Malarone to prevent malaria when traveling—the same drug is used to treat malaria, so why not wait and take it IF you need it? Enough with these vaccines—which have side effects for many people—when an easy to get treatment is available.
Having just yesterday got part two of my shingles vaccine I was doubly disappointed here. Apparently I am no less likely to get dementia and Dr Prasad won’t reveal the important bit. Is it good at preventing shingles? I had shingles about eight years ago and didn’t much enjoy it.
This was not a piece about shingles. You can research the studies on efficacy of the shot for shingles prevention.
This was my feeble attempt at a joke. Clearly it failed, and probably deserved to. But he does sort of gesture at the question in the penultimate paragraph. But I agree. This was not a piece about shingles!
Is it really worth getting shingles vaccines AFTER having had shingles though?!
Well, my spouse did. Her experience with shingles was a miserable one. The shots were easy for her. I had the original vax then the first dose of the updated vax. That was miserable for me. And then Covid hit about a month later. Never got second dose but just the one is about 85-90% protective if I recall. Second does gets it up north of 95%.
That puzzled me but apparently you can get it again. Seems bizarre then that vaccines help better than getting it but I don't really understand vaccines all that well. They don't all seem to work the same way.
There's no such thing as a vax that works better than nature-we're talking future immunity here, not symptoms in the moment.
Well the chicken pox virus is weird. Only folks who got that virus are at risk of shingles. Folks who got the vax for chicken pox do not get shingles.
Interesting. I probably missed by only a few years.
in the late 70's, I worked for the great modern dance choreographer Martha Graham. when she came down with shingles, I had never heard of it before. I thought it was a literary "disease" that old maids got in Dickens' novels. then in 2004, I met or heard about 3 young men, all 23, who had shingles. one was a dancer who had to miss his opening night while the crew raced to alter his costumes for his understudy in June. the second was a young stagehand who was a friend of my boyfriend's and was just recovering at a 4th of July BBQ. he said it was awful. I heard about a third 23 year old coming down with shingles in August. what did they have in common, other than their age and gender? they had all been born in 1981, and after vaccines were given liability protection in 1986 and after the varicella vaccine was routinely given to children, NONE OF THEM HAD BEEN ALLOWED TO HAVE WILD TYPE CHICKEN POX AND DEVELOP REAL IMMUNITY TO IT.
my theory was that not having chicken pox at the appropriate age because of a vaccine intervention predisposes one to getting shingles later in life. or in the case of these young men- not all that later (they were all 27 years too young to qualify for the shingles vaccine).
I subsequently read that in the UK, the varicella vaccine is not recommended because the National Health system there would prefer to avoid the shingles epidemic that follows the lack of having chicken pox as a child.
when are hubristic "scientists" going to learn that nature works best if left alone? healthy children can easily handle these infections. adults do less well with shingles. this is a man made "epidemic."
What is the bigger problem is that people are not being "reinnoculated" against chicken pox by being exposed to kids who have full-blown cases. That was another thing (or maybe the primary thing) the UK was concerned about. When you vaccinate a kid, the elders around them are more likely to get shingles. Of course, this is quite convenient for vaccine manufacturers who make a profit on boosters.
you are 100,000% correct. the immunity, superior to vaccine immunity though it is, still needs to be topped up now and then. you have children and they get measles or chicken pox and you're exposed even though you don't get sick. you're exposed to your nieces and nephews and cousins and it just "tops up" your immunity. you're exposed to your sick grandchildren.
then one day, they put you in a nursing home and your immunity wanes and they certainly don't bring your great grandchildren in to visit you when they're sick! the next thing you know, you've got shingles.
the best thing that could happen- and it will take a few generations- is for all kids to get natural measles, mumps and chickenpox at an appropriate age and to go back to exposures that reinforce natural immunity.
instead we have hysterics screaming over every case of measles like it's the end of the world.
This commentary comes pretty close to declaring this study, or at least its claims, to be fraudulent, starting with another study that was clearly so. The nicest thing it says is that the claims are “magical”. What is closer to magical thinking is this commentary. Dr. Prasad substitutes his own words – better described as unfounded accusations - for the actual claims of the authors, no less their methods. Let’s look at the article’s “claims”, taken verbatim:
“The observed reduction of risk is clinically meaningful: if confirmed in clinical trials…”
“The biological mechanisms underlying these observations remain to be clarified experimentally.”
“Third, although our study is a natural experiment … it does not imply that associations are causal. Causality requires additional assumptions (outlined in the Methods alongside supporting evidence and discussion of remaining uncertainties).”
“Early divergence of Kaplan−Meier curves, while biologically plausible, can be a sign of unmeasured confounding..."
“…these findings, if confirmed in clinical trials and mechanistic studies, would have important implications for public health.”
In Methods, under “Assumptions for causal inference”:
“Although the above evidence supports exchangeability on observed dimensions, the influence of several unmeasured factors cannot be ruled out. … If they systematically differ between cohorts after matching, such unmeasured confounders may violate the exchangeability assumption.” They then include a variety of possibly important covariates that they could not measure, including particularly social determinants of health.
Every point raised by Dr. Prasad is discussed in the paper, some at considerable length in both the discussion and the methods. This includes the rapid divergence of the curves, whose explanation the authors concede is not evident, even though they did a variety of analyses, including 6 negative control outcomes, to explore obvious methodologic artifacts. The proposed explanations are framed as tentative; some speculative, some supported by cited evidence.
At the end of the day, the authors end up with the “claim” that the phenomenon is merely worth further exploration with stronger empirical designs (i.e. trials) and laboratory studies of mechanism. That’s it. Their claim is that the phenomenon deserves a closer and better look. Is that a universal “get out of inference free” card that protects against spurious claims? Only if the authors had not engaged in the serious methodologic efforts they did, which go unmentioned and unrebutted by Dr. Prasad.
I will add parenthetically that one of the more challenging aspects of such research is the TriNetX database itself, but that is an issue for another day. And if this turns out to be true, it's main value will be in mechanistic understanding, as the motivation to get this vaccine will always be to avoid shingles, not to improve heart health.
One could still legitimately take issue with the article's modest claims - i.e. that this hypothesis is worth further attention - and whether the many caveats are strong enough. But not using the hand-wavy, unidimensional arguments put forward by Dr. Prasad, not to mention his dismissive and almost contemptuous tone. Reasonable scientists can and will disagree as to how likely these data patterns are to represent causal relationships – which the authors do not claim and themselves may not agree on – but Dr. Prasad should engage with what the authors actually did and said, not a cartoon version. But that would take time and wouldn’t be as clickbaitable.
I would encourage every reader of this substack to read the actual paper to reflect not only on the strength of evidence provided by this study, but on the intellectual honesty of Dr. Prasad’s critique. And then decide which is more "magical".
remember when they came out with a "study" saying that covid vaccinated drivers had fewer accidents? Dr Prasad was pretty critical of that "study" too. the vaccine cartel is always trying to expand all the "miracles" that vaccines do, except when it comes to increasing autism and other developmental disorders in children. that, of course, is simply "coincidence."
It should have not have been published at all. That's the point! People will generally read only the title and get a false impression.
“Third, although our study is a natural experiment … it does not imply that associations are causal. Causality requires additional assumptions (outlined in the Methods alongside supporting evidence and discussion of remaining uncertainties).”
Well that was big of them.
Causality requires a different study design, not more assumptions using retrospective data. This is the problem with studies that use propensity-score matching and TTE. In the end, you have associations using retrospective data from TriNetX or wherever that have been souped-up statistically to make it seem, after all the caveats & qualifiers & assumptions & corrections etc. etc., like you have a basis for causality. But listen: You can't know what you can't know, & in a non-randomized study there is plenty.
Why does everyone think retrospective observational studies are so bad? They have their place. They can be hypothesis generating. But this whole trend toward trying to dress them up as something resembling a study that can even approximate causality reminds me of people who buy tiny houses & then talk about all the stuff they did to make it seem bigger. Big windows! Tall ceilings! It even has a bathtub! yadda yadda. Look, you live in a TINYhouse, aka a "tiny HOUSE." If space was such a priority, you should have just bought a bigger house. But all this talk about how a tall ceiling makes a 200-sf place seem big, I mean, don't piss on my leg & say it's raining. Just find a way to do an RCT or else be satisfied with the modest conclusions you were able to reach & let someone who is better funded build on that.
"Reasonable scientists can and will disagree as to how likely these data patterns are to represent causal relationships" - what an enormous waste of time. Do the study & get it overwith. Oh, no money for an RCT. Well I'll sell you a TTE real cheap!
Thanks Vinay. Couple of things - 1. where possible can you include more info on the charts/graphs you present. I can mostly figure them out, but it takes a lot of work. (And I guess I don't really understand what SMD is in table 1 - I plugged the numbers in for difference in % white 81% vs 76% and get a z of 21.4 on the test of two sample independent proportions, making p<.00000 but SMD is reported at .13. 2. Whatever SMD is, I can see that there are (significant?) differences in the two populations, but most of them are really really small and at the ends of the % scale where there is a lot of power to detect differences. And, while it would certainly be fair to say the groups are different, the variables presented are not the ones that I think of as really confounding these kinds of comparison i.e., SES type variables or those related to neuroticism or conscientiousness.
I realize that it is probably an undue burden to expect the critic to come up with the "story" of the confounding, but you have done it in other cases and I think that ease and plausibility of generating the story is useful in many cases (e.g, conscientiousness is the third variable explaining vaccination and auto accidents).
I would love an analysis on shingles…i think there are two types live and attenuated(?). 🙏
There is only one shingles vaccine available in the US since 2021 -Shingrix which is recombinant. Before 2021 there was also a live, attenuated vaccine but it has not been available for years. The vaccine is quite effective, increasingly so as one ages which is unusual. But most people get some kind of side effect when they take it (not major except for very rare Guillain-Barré syndrome, but not pleasant either) and it takes two shots three to six months apart to actually be decently effective. But most would agree better than getting shingles in the final analysis.
Do you know the % of these shingle shots working? I’ve had 2 friends and my brother all get the shingles during covid. All 3 had the double shots. I’m hoping it was just coincidental and the % is high.
My observational, non-medical take is that “spiking” the immune system per COVID shots led to a rise in Shingles. A vaxxed and boosted 50 year old friend came down with Shingles. Her primary care doctor told her to quit the boosters. He was seeing a bunch of Shingles cases.
My husband also got shingles after the covid vaccination despite being inoculated prior.
Shingrix in ZOE-50 and ZOE-70 produced 97.2% efficacy against zoster in adults 50 and older and 89.8% in those 70 and older, with about 91% efficacy against PHN (Post herpetic neuralgia...the really crummy part); the ZOE long-term follow-up cohort still showed roughly 79% to 82% efficacy at ten to eleven years post-vaccination.
thank you for this. Honest question, what does the 97% efficacy rate mean? percentage of people who dont get shingles compared to same age cohort who dont get the vaccine? i understand ARR, RRR and NNT but in this case im confused on how efficacy is defined. thank you again. ( Guillain Barre would be terrible.)
Vaccine efficacy is simly a relative risk reduction, calculated on incidence rates rather than cumulative proportions: VE = 1 − (rate in vaccinated / rate in placebo), where rate is confirmed zoster cases per person-year of follow-up in the randomized trial.
In ZOE-50 the numbers were 6 cases in about 7,300 vaccinees versus 210 in about 7,400 placebo recipients over a mean 3.2 years of follow-up, giving incidence rates of roughly 0.3 versus 9.1 per 1,000 person-years; 1 − 0.3/9.1 ≈ 0.97, hence 97.2%. So yes, it is the vaccinated cohort compared against a same-age randomized placebo cohort, but it tells you about the proportion of expected cases that were prevented, not the proportion of vaccinees protected.
The absolute framing is what you would expect from the RRR: ARR of about 8.8 cases per 1,000 person-years, so an NNT of roughly 114 person-years, or about 35 people vaccinated over the 3.2-year trial window to prevent one case of zoster (larger for PHN, since that is a fraction of cases).
Note that this is per-protocol efficacy in trial conditions, and "effectiveness" is the corresponding term for real-world observational estimates, which for Shingrix have run in the 80s to low 90s rather than 97%. But still impressive as these things go.