Wednesday, April 29, 2020

The Trouble with "Evidence Based" Medicine

Is this loose, informal transmission of anecdotal findings--call it chatter, call it rumor--part of medicine?  It isn't what anyone taught in medical school; it doesn't fit in with the professional's image as a purveyor of rigorously tested interventions. But continuous, iterative clinical knowledge--the kind that can be updated minute by minute--is invaluable during this tumult, when time is of the essence and there's scant research to fall back on."
--Siddhartha Mukherjee, MD (Columbia P&S)




Daniel Griffin, MD is one of my favorite doctors and his appearances on This Week in Virology are a highlight.

But I have to dissent regarding one point he tried to make about "Evidence Based Medicine" on TWIV #606. And this is no quibble, but an essence.

Trying to analyze how we come to agree on a course of therapy for patients in the real world, whether the ICU or the ward, Griffin analyzed the way medical therapies get established--and in this he did something I have never heard but I thrilled to his observations because they were  so spot on.
Daniel Griffin, MD, a prince among physicians

It was when he tried to suggest an alternative, we parted ways.

First, we should listen to his observations, and understand it comes in the backdrop of President Trump shouting from his podium we should be using hydroxychloroquine on all COVID 19 patients because Mr. Trump had heard, IT'S JUST AMAZING. 
Most doctors hear that sort of testimonial, Hallelujah & Praise the Lord, and recoil, appropriately. 

Griffin also recounts cases of doctors who feel helpless and desperate to "just do something" to save their patients.

But, as he said this, I thought of Franklin Roosevelt saying, about the Depression, another pestilence which we did not understand, it was important to simply try something, anything, and if it did not work, try something else.  Better than the do nothing impotence of Herbert Hoover. It's fine to define the ideal, but we do not live in an ideal world, on the wards. 

Griffin says when he went to medical school there was no effort to base therapeutic practices on prospective, double blind studies and he says, "I still remember one senior physician saying 'You're telling me one well controlled prospective study of a thousand patients is better than my years of experience' and we said, 'Yes, that's what we're saying.'"

And I thought, I went to medical school a decade earlier, and just 30 blocks up the street from where Dr. Griffin went to medical school and we tried, as did our faculty, to base our therapies on the literature, but the literature was often changing and contradictory. We were constantly citing  articles to each other, and we had only a small fraction of the information available to us we have today with the internet and with podcasts and streamed conferences.

Griffin goes on to outline the types of persuasion used on the ward or the ICU to get a particular approach accepted:
1/ The "In my experience"  mode:  where you're asked to listen to  someone "who has been making the same mistakes with increasing confidence over an impressive number of years," as Griffin describes it. The 3 most dangerous words on the ward. "In my experience."  But are we to discount experience on the ward? Is "proning" the patient  from lying on his back to front not "in my experience"? 
2/ "Vehemence based medicine": Someone with a loud aggressive confident voice "confidence based" drives care by force of intimidation. This is the sort of bullying, which hurls overt or implied accusations of ignorance, fecklessness at those who might oppose it. 
3/ "Eloquence based  medicine" by which Griffin sees doctors beguiled into accepting an approach because that approach seems to make sense, to be based on solid theory in the absence of rigorously tested medicine. If we believe, in theory, the 1st week of COVID is when the viral replicates and the 2nd week is when all hell breaks loose because of an over reaction of the immune system, which causes cellular damage, endothelial cell injury and the unleashing of an avalanche of clotting, a consumptive coagulation cascade and widespread embolic bombs to cut off oxygen to brain, lung, liver and kidneys, then we should be focusing anti viral medications like remdesivir in the 1st week and glucocorticoids and anti IL6 drugs thereafter, but where are the prospective, double blind placebo controlled studies? 
4/ "Diffidence based medicine" when the doctor says, "No good evidence here."
And here Griffin is most puzzling, "There may be no good answer but there are certainly a lot of bad answers" 
5/ "Nervousness based medicine" Which Griffin describes as the attitude,  "the only bad therapy is the one you forgot to do," as in,  "I'm going to do everything because maybe one of those things is good." So "when the family calls you on the phone and asks did you try A, B and C, I say we didn't do that because we tried what we tried what we actually thought what would make your loved one better." 
But he has, implicitly admitted, we really don't know what will make your love one better; we are simply afraid of doing something to make him worse: Primum non nocere. 



So what Griffin proclaims as his approach is "Careful observational based medicine where we are trying things" So he puts a patient in anticoagulation. "We're seeing a lot of people clotting and so we anticoagulated and the numbers went from 8 a day to much lower."  
But how is that different from "eloquence based medicine" or "diffidence based medicine' or "nervousness based medicine," except when Dr. Griffin tries it he thinks he looks more critically or carefully for the response than others do?

When he discusses clinical trials and how some studies deviate from the golden standard,  he describes Paul Merek's observation of treating septic patients with vitamin C plus steroids where he did a clinical trial and saw mortality rates from sepsis drop 70% . But there was non control group.   Then a prospective controlled trial was done and found no benefit. So what Griffin is saying is some observational/retrospective trials are misleading. 

Then Griffin describes an observational trial he did using steroids + Tosi (an IL6 receptor inhibitor) to treat cytokine storm which showed a survival rate of patients rise from 3% to 21% but it was not a controlled trial, nor was it a blinded trial which is meant to correct for physician bias. 

When he describes his own recommendations, based on his own experience in treating patients, it's hard to see how much more objective and scientific he is than doctors who he criticizes as practicing diffidence medicine. 

He mentions patients who are afraid to go to the hospital so they stay home and let their appendices burst at home.  He decries denying an appendicitis an appendectomy. But there are plenty of studies where patients are treated with antibiotics rather than surgery and seem to do as well at least in the short term. So he has bought into an experienced based therapy himself, namely, appendectomy.

Griffin is the sort of physician I admire: He is confident enough in what he knows to be humble; the really good doctors are almost always humble, more impressed by what they do not know than by what they do know. 

But he is clearly offended by the "cowboys" and bullies he has to deal with on the wards whose egos ascend over the best interests of their patients.

To hold out evidence based medicine as the true holy grail however, is to deny the impossibility of limiting therapy to  evidence from double blind prospective controlled studies. There simply are not enough of those to apply to every clinical problem, or to the complexity of multiple problems at once. 

As one of the TWiV crew asked (I think it was Rich Condit, a PhD self proclaimed "lab nerd"): You are doing so many things to the patient, how do you figure out which part of the magic sauce was important?  Condit is not a clinician, but he had the experience of learning how to do lab science and he knew he had slowly accrued experience which helped him work more efficiently.

Ward medicine is something like the problem of developing a car which drives itself, as opposed to being driven by a experience driven driver who takes in multiple cues at once and without doing a study when he sees a ball bounce along the grass toward the road, hits the brakes because he knows (even if his intelligent car never notices the ball) that balls rolling down to the road are often followed by children in pursuit.


Semmelweiss observed that women who delivered their babies  in the hospital in Hungary  got infections after delivery, but women who delivered at home almost never did. He also observed physicians going from autopsy room to bedside and examining the vaginas of post partum women without washing their hands or using gloves. (Rubber gloves came much later.) He may not have had double blind controlled prospective studies but he had observation. He did not even have solid germ theory, but he had the ability to make connections and he advocated for washing hands between examining patients. He died in 1865, well before microbiology and germ theory had taken hold in medical practice, but he saved lives by cleaving to an observing, in his experience. 


Daniel Griffin notes he's stopped using hydroxychloroquine, tosilisimab, remdesivir and azithromycin which were therapies without proper studies, but his reason for stopping them was his own observation they didn't work.  So he acts based on his observations, while he awaits better studies; he used them for a while, which is what other doctors have done. Try something until you lose faith. 

Putting patients prone, which he's observed increased oxygenation of his patients, has no evidence based studies, but as he says, "It seems to work." 

So, having stated what amounts to a practice of purity, he quickly diverges toward what seems practical. 

Still, he is trying for an ideal; you can't criticize him for trying for exalting an ideal,  and if your mother got sick with COVID19, you'd want him taking care of her. 




Sunday, April 12, 2020

COVID19: What We Know on Easter via TWIV

If you define "experts" as people with reliable, tested solutions, then COVID19 has no experts. But if you listen to the virologists on This Week in Virology podcast it will restore your faith in experts, and in humanity. There are folks who have studied viruses long before and now during COVID19. For them, this virus is just another in a long line of DNA/RNA actors. They get into depth, and it takes an hour each episode, which is beyond the attention span of most listeners, certainly of most journalists, but it's there for those who care to learn. In particular, their guest, Dr. Daniel Griffin, elucidates how physicians have learned from their mistakes to treat the COVID19 patients more successfully.

--Comment from a New Hampshire reader in The New York Times to a Ross Douthat article about not trusting experts when it comes to COVID19.




Episode #600 of This Week In Virology has aired and the Phantom, once again, will attempt, with apologies, to summarize, boil down and translate.

The Phantom is humbled by the task, and feels unworthy, knowing the virologists and physicians he reports may cringe to hear his simplifications.

But the Phantom hears too often from people he has pushed to tune into to this estimable podcast that:
1/ It is simply too long.
2/ Too many of the words and the concepts are unfamilier: i.e., you have to be a doctor or a biologist to understand what they are talking about.

Actually, the Phantom thinks the TWIV geeks are pretty good at explaining and "making accessible" but he will try to get it into a more digestible form.

Here is a link to the actual episode #600
https://www.microbe.tv/twiv/twiv-600/


This week the Phantom learned:

1/ Pregnant mothers who are COVID19 + are being separated from their newborns and this is a bad move. IF the mothers breast feed, they'll likely pass their antibodies on to the baby and that will be good for the baby.

2/ Hydroxychloroquine (HC) may worsen the disease if given at the wrong moment in the illness, but if given at day 7, it may be helpful.
Daniel Griffin, the physician from Long Island, who is a professor at Columbia, is launching a study, a real, double blind prospective study to answer two questions:
  A. Can hydroxychloroquine (HC) successfully "treat" COVID19?
B. Can hydroxychloroquine (HC) prevent health care workers from getting the disease?

How would you know if HC is an effective treatment?
If you treat a patient with strep throat with penicillin, you can see the patient improve with respect to throat pain, fever and malaise rapidly and you can swab their throat after the penicillin is done and find no positive throat culture.
With COVID 19 there is no convenient or practical way to culture after therapy--mostly what you've got is a PCR swab which will remain positive after therapy because it measures pieces of the virus and this flotsam may float around for days to weeks after the virus has been neutralized. If the patient gets better, if his blood oxygen improves, his fever comes down and his respiratory rate returns to normal, that's a good sign the HC worked, but all that happens in patients who are not treated, so how do you know it was the HC that did the trick?

That what science is all about--testing, designing studies to separate coincidence from effect.
Daniel Griffin, MD



Daniel Griffin, MD, who is running the studies, points out you are actually more likely to be injured by the drugs used in the studies than helped by them. If 95-99% of people recover from COVID19 and if 5% of patients receiving HC get complications (seizure, cardiac arrest, rash, liver dysfunction) then the odds of your benefiting from being a study subject are against you. He says some studies are being marketed to the public as if it's all gain--step right up, get your mother enrolled today! And he makes a really eloquent argument for science, rationality and for not falling prey to the wishful thinking which emerges in every epidemic.

3/ "Steroid" (glucocorticoid, prednisone, cortisone) therapy given at the wrong stage of the illness can cause worsening as the immune system which is trying to kill the virus is stopped in its tracks. Later on, when the immune system is carpet bombing the lungs, it may well help.

4/ Anti virals, which interfere with various actions of the virus in entering the cells, or making more virus, may be useful in some stages of the disease. There are also drugs which interfere with host enzymes which the virus uses to split it's own protein coat--very much as the Greeks (the virus particles) were contained inside the wooden horse, but they needed someone on the outside, the hosts (the Trojans) to first drag the horse inside the walls of Troy and then open up the horse so the Greeks could spill out and sack the city.

5/ Vaccines will be the ultimate answer, but there is no good way to rush their development. There are good ways to rush the production and to ramp it up quickly, but first you need to identify the right vaccine.

6/Antibody tests will be vital to figure out who has had the disease and recovered. But there are two problems: some antibody tests have been inaccurate. The false positive tests occur because antibodies to the common cold, another coronavirus, can cross react to give a false positive for antibodies to COVID19.
It is not certain that the presence of antibodies will protect against another infection with COVID19 or, if they do protect, for how long. Some immunity may last only weeks or months.



Sunday, April 5, 2020

COVID19: TWIV: The Role of the Microbiome



Jonathan Yewdell, speaking on This Week in Viriology (#597) about the human system, B cells, T cells, antibodies, memory cells, lymph nodes, mentioned, just in passing, something which may help explain the answer to the question in the prior blog post: Why do some people get deathly ill or even dead, from COVID19, where other people hardly even know they have it?

Jon Yewdell


He tells a story about  a graduate student's work in Maryland, where he went around collecting mouse feces, and he brought these samples back and gave some to lab mice, whose guts were colonized by lab microbiomes.

These lab mice were notoriously sensitive to influenza virus, and died quickly whenever they ran across it.  But the lab mice who ingested the microbiome of the wild field mice now  just shrugged off the influenza virus.

This suggests, or possibly proves, the microbiome of the mouse, and possibly of human beings might protect against viral infections.

Is it possible that human beings who have asymptomatic intercourse with COVID19 may have been impervious because of their gut microbiomes?

Just another little pearl from TWIV.


COVID19 for Dummies: Meaning, Just About All Of Us



A   disclaimer.  The Phantom is not an Infectious Disease specialist and has no more than a reading/podcast listening  knowledge of COVID19.



But one thing the Phantom can do is to detect quality, or lack of it,  in the folks who want to tell you about COVID19.

That means the Phantom knows that Dr. Nicole Saphier of FOX & Friends and Dr. Oz, of same and Jared Kushner, President Trump, Sean Hannity, Alex Jones, Rush Limbaugh all fall into the same category:  sources not to be trusted on the topic of COVID19. 
They are not credible sources, with respect to virology.
They are, as our President would say: "Incredible."
An Excellent Source of Reliable Thought and Information

But don't despair:  there are good sources.
 Anthony Fauci, MD has been reliable, open, and, as most really good doctors are: he is more humbled by what he does not know than he is impressed with his own mastery.




Then there are the folks who have done a podcast for geeks called "This Week in Virology" which is the best single source of information, science and clarity about COVID19 of which the Phantom is aware.



The trouble is, this is a podcast by and for geeks, and they use a vocabulary which likely will lose the average lawyer, journalist, plumber, school teacher. 

So the Phantom has undertaken a project in translation. This is not something the Phantom has been trained for, but for now, he tries to fill a void. The Phantom has listened to the most salient episodes (#595 & # 598) multiple times and they are so densely packed, he knows he has more listening to do.


Dickson Despommier


For those who want real information, the Phantom urges the reader to go to these podcasts: But be warned, each is an hour long. On the other hand, perfectly timed for walking the dog.

Here are some questions:
1. How many people get a dose of COVID19 and never become symptomatic?
2. Can these people shed virus and infect their family members, coworkers or other more casual contacts?
3. How long does the carrier shed?
4. Is it shed by droplets, urine, stools?
5. Once on board, what factors determine whether someone shrugs it off or gets sick or dies?
6. How much of a risk for carrier state are cats and dogs?
7. Is hydroxy chloroquine the answer? By what mechanism does it work? 
8. How effective are masks?
9. What is the prevalence, the mortality rate of COVID19?
10. How do you know if you have COVID19  or have had it?
11. What progress has been made on treatment?
12. What is "Cytokine storm"?
Kathy Spindler

Some tentative answers:
1. Asymptomatic people: We cannot know this without a certain type of study, which as yet has not yet been done. A study of people in Washington State of people who showed up at a drive in testing center because they were either worried or had symptoms showed that people with no symptoms were +10% for virus, whereas people with fever, cough, aches, fatigue, abdominal pain were +75% of the time. 
The more symptoms you had, the more likely you were to be + on PCR testing.

PCR looks for molecular material and indicates the presence of the virus.
To know if someone who is negative for virus on PCR has already fought off the disease, you need antibody testing. 
There are two types of antibodies: IgM, the first antibodies within 7 days of arrival of infection. IgG the antibodies which rise 7-14 days later and may confer long term immunity.

2. Silent shedders: are clearly a reality. Typhoid Mary did not know she was shedding typhoid because she was asymptomatic.

3. How long does the carrier shed?  
Depends on what you measure. Present in droplets sprayed out by patients for 7-14 days after cough ends. In stool, maybe 3 weeks. On surfaces, depends on the surface, from 3 weeks to 4 days.



4. Virus has been found in all these sources, and gets in via eye, mouth ports of entry.

5. Who fights it off and who dies? Why?
Probably there are "host factors" and "virus factors."
Virus factors include the dose of virus particles, virions. This is likely why the people intubating (putting tubes in tracheas) in the hospitals have over a 30% chance of getting COVID19 week by week. They are being sprayed by the patients. Once intubated, the doctors around the patients are less likely to catch the disease.
There are rare folks with mutations in the genes for cell receptors to which virus attaches who are resistant

6. Cats carry it, not so much dogs
It's not clear if you can get it from your cat.
Your dog, likely not a vector.
{***SINCE THIS WAS ORIGINALLY PUBLISHED A REPORT OF A TIGER BECOMING COVID19+ AT THE BRONX ZOO APPEARED IN THE NYTIMES: APPARENTLY A CASE OF HUMAN-->CAT TRANSMISSION.)

7. Hydroxychloroquine: 
Works by affecting the way endosomes inside cells which help kill virus. Likely there are other, unknown mechanisms of action as well.
Malaria can get resistant to this drug and the virus may well as well. If given at just the right time, this drug may make a difference and Dr. Daniel Griffin talks clearly about how ICU docs have learned to time the sequence of the drugs they give. They try to start it within the first 48 hours after symptoms begin.
Dr. Griffin mentioned he looked at the French study which suggested a great benefit for hydroxychloroquine and it was nothing more than a string of anecdotes woven together to suggest this is an effective drug.  Dr. Fauci called the study, "not robust" which was likely a generous description. Nevertheless, all COVID19 patients are now getting this + zithromycin on admission at some NYC hospitals. 

8. Masks:  used properly, may protect others against you and may protect you, in part by preventing you from getting your fingers in your nose or mouth. The TWIV guys admit they were less persuaded of the efficacy of masks in protecting against acquiring infection, but there are now some studies which suggest they can be effect not just to prevent spreading it to others but against acquiring it.

9. Mortality/ Prevalence: Because we do not have studies of the population for the presence of virus (current infection) or antibodies (past infection) we do not know.

10. How do you know if you've had it?  
Blood tests for antibodies are required. You can even tell how recently you have had it. If your IgM is negative and your IGG positive, you're probably 7-14 day past your innoculation.
Dr. Griffin

11. Daniel Griffin, MD who features in two of the key episodes on TWIV (#595 and #598) details exactly how the doctors have learned to treat this, correcting initial mistakes and capitalizing on what they have seen in their patients. Likely, your chances of dying if you get admitted to hospital are improved now because of what doctors have learned to do and not to do. But if you wind up on a respirator, your chances are only 50/50.




12. "Cytokine storm" :
Is what happens when your immune system starts carpet bombing your own body, trying to kill of the virus.

The best way to understand this is through an analogy to that episode from "Band of Brothers" in which Captain Winters has his platoon attack the Germans and they fire single shots mostly, some machine gun fire, and they shoot individual German soldiers.  In doing this, they kill only the enemy, and not their own troops. 

This is analogous to what is called "adaptive" immunity in biology. 
The B cells make specific antibodies to a specific virus and kill off the virions specifically, aiming at specific targets.

But then the artillery starts firing, and anyone in the vicinity gets blown to smithereens, the American shelling killing Americans and Germans indiscriminately. 
This is analogous to what is called "innate" immunity.  It involves things like IL6, cytokines, all sorts of stuff which just burns the village down, blows away Jeeps, troops, anything.

https://www.youtube.com/watch?v=DOSvLWK5Z2A

During the first 7 days of infection, the virus load increases and the patient is sick. 
At this point, antivirals like "Tosi" may help. This is when the B cells are trying to shoot the virions with antibodies.

But when the patient crashes, has trouble breathing and the chest X rays show fluid flooding the lungs, that is the "innate" system, the artillery, kicking in and the patient dies from the indiscriminant mayhem wrought by the innate system.

Doctors have learned to treat this with a variety of medications, including cortisone drugs, but also interferon.

You can google TWIV episode #598 or try this link: https://player.fm/series/this-week-in-virology-with-vincent-racaniello/twiv-598-who-was-that-masked-man-coronavirus-update-with-daniel-griffin