Guest post by Alec Rawls
For weeks Dr. Anthony Fauci has dismissed all of the mounting evidence for the efficacy of Hydroxychloroquine treatments for Covid-19 as “anecdotal” and “the slightest hint of evidence” because it doesn’t come from clinically controlled trials. That is just wrong. Valid statistics do need a control group but the control does not have to be clinical.
New York alone has recently amassed many thousands of case histories of Corona patients who were given various HCQ treatments (HCQ plus zinc and/or Azithromycin in various dosages) as doctors swapped information about what seemed to be working. Despite the lack of a clinical control group (doctors are not withholding what they estimate to be the best medicine from half of their gravely ill patients), there still is a massive control group in place: the hundreds of thousands of case histories of Corona patients all around the world who never received any Hydroxy treatments.
The outcomes from large samples of both groups just need to be tallied and the statistical significance of the difference in mortality will far exceed what could ever be achieved through clinical trials (unless the CDC is planning on killing an awful lot of people by withholding already known to be life-saving medicine from them).
But Dr. Fauci has no plans to look at this huge body of field data. He plans to wait instead until clinically controlled trials are completed sometime this summer. From April 9th:
It’ll probably be months, sometime in the summer—we’ll start to see which are working, which are not and to focus on those that are we developing and will work.
Forget the fantastic progress that hundreds of doctors working feverishly on thousands of patients have already made in discovering “which [treatments] are working and which are not.” What’s the hurry? We’ll just re-invent the wheel.
Denigration of trials that use non-clinical controls is a 180° turn for Dr. Fauci
It turns out that Dr. Fauci knows as well as anybody that non-clinical control groups support perfectly valid statistical evidence because his own career-making medical research was entirely dependent on historic controls (comparing his treatment outcomes to historic mortality rates), with no clinical control groups ever involved.
This bit of Fauci history was mentioned by Dr. Stephen Smith in an interview with Laura Ingraham on April 6th. Ingraham asks what Dr. Smith thinks about Fauci’s skepticism and he responds:
What he’s saying Laura is that we don’t have randomized controls of the trials and that’s true but we have plenty of diseases including the disease that doctor Wolff and Dr. Fauci [unintelligible] got famous for: Wegner’s Granulomatosis. They developed a treatment for it in the 70s and 80s that saved lives. But if memory serves their article in Annals and their article in 1973, Annals [was] in 1983 I believe, they didn’t use, it was not randomized, it was not controlled. They didn’t even have contemporaneous controls, they used historical controls, and I’m not blaming them. I believe they did great work and that the treatment DID work.
But we use the data we have, and then we build up on experience. Now the experience that we have had treating patients with a combination of Hydroxychloroquine and Azithromycin is that the data accumulated, and if you get to it, the more you are on it, the lower your chances of being intubated or being on a ventilator are. Those are observational data but very very powerful data.
Dr. Smith is right, Fauci’s own medical research is rife with non-clinical controls
FAUCI AS, WOLFF SM. Wegener’s granulomatosis: studies in eighteen patients and a review of the literature. Medicine (Baltimore). 1973;52:535-61.
On p. 551 they use a pair of past studies to establish a baseline mortality for the disease they are trying to treat:
They found that the average time of survival of these patients was 12 ½ months as compared with the untreated patients in Walton’s review (106) where survival was 5 months.
106. Walton. E.W.: Giant-cell granuloma of the respiratory tract (Wegener’s granulomatosis). Brit. Med. J., 2: 265. 1958.
Wegener’s Granulomatosis: Prospective Clinical and Therapeutic Experience With 85 Patients for 21 Years
ANTHONY S. FAUCI, M.D.; BARTON F. HAYNES, M.D.; PAUL KATZ, M.D.; SHELDON M. WOLFF, M.D.
For the mortality of an untreated control Fauci, Wolff et al. cite in this 1983 paper the same Walton review article they cited in their 1973 paper:
The mean survival of untreated Wegener’s granulomatosis was 5 months with 82% of patients dying within 1 year, and more than 90% of patients dying within 2 years (9).
9. WALTON EW. Giant-cell granuloma of the respiratory tract (Wegener’s granulomatosis). Br Med J. 1958;2:265-70.
Of course using a historical control is perfectly sound science. It is just 180 degrees opposite from what Fauci is for some reason insisting today, and the change is definitely recent.
Googling Fauci “historical control” also brings up from the New England Journal of Medicine a November 4th 2016 Fauci etal. abstract that concludes:
VRC01 slightly delayed plasma viral rebound in the trial participants, as compared with historical controls, but it did not maintain viral suppression by week 8.
So at least until November 2016 Fauci was fine with historical controls.
The CDC and Fauci’s NIAID are shirking their immediate responsibility to tally the massive existing field trial evidence
Dr. Fauci’s pretense that non-clinically controlled trials are unscientific keeps our public health bureaucracies from taking up their most immediate responsibility. Fauci is the head of the National Institute of Allergy and Infectious Diseases (NIAID). His group and the CDC are the ones who ought to sending a small part of their large federal workforces to New York and around the country to compile the massive number of existing case-histories into a data set that can be analyzed.
From the CDC’s mission statement page:
CDC in the 21st Century – confronting global disease threats through advanced computing and lab analysis of huge amounts of data to quickly find solutions.
Except they are not interested in collecting the already existing data that can quickly provide an accurate assessment of how much the HCQ treatments lower mortality for at-risk patients. Their official position under the guidance of Fauci et al. is that real-world data isn’t good enough to bother with. Instead they will wait until they can generate their own clinical data.
Listen to CDC chief Dr. Robert Redfield dismiss evidence from field trials about the effectiveness of HCQ treatments (at 1:10 here):
Yeah, I’m not going to recommend it and I’m not going to not recommend it. I think that you and I are very similar. We’re probably similar to my friend, Dr. Fauci. We-we-we we’re very comfortable to responding when we have data.
He is following Fauci, refusing to regard field trials as having statistical significance. Because field controls are not clinical they supposedly produce no “data.” These are the heads of the two bureaucracies that need to be out processing this data and they both at least claim to believe that real-world data isn’t actually data.
We might have a substitute for social distancing in eight days
In his initial interview Dr. Smith spoke to this perverse phenomenon of health bureaucrats badmouthing the attempts of doctors in the trenches to tell each other and the world what is working for them. Instead these agencies ought to be coming in to collect and process the field data for themselves.
Here is what Smith said about Fauci denigrating the early French study as “anecdote”:
The French had the virologic study first. I think that was March 17 maybe? They showed that if you used these two drugs together [HCQ + Azithromycin] the patients stopped shedding virus, or cleared virus, by day six. American researchers — I’m a private practicing I.D. Doc [Infectious Disease] — If American researchers were worried about that [about the validity of the French findings] they could come up here to New Jersey and have that study done in eight days. I don’t understand the hand ringing. I don’t understand the hesitation. If you’re worried about getting studies done, do them.
Yes, and if steep drops in the level and duration of viral load is borne out it means we have a way to dramatically lower Corona transmission rates independent of social distancing and economic shut-down. Just ask everyone who gets flu-like symptoms to drive through for a Corona test and if they come up positive give them 5 days of Hydroxy and zinc, then we can all get back to our normal lives tomorrow.
But getting this crucial information in eight days doesn’t interest the CDC and NIAID. They only consider clinically controlled trials to be meaningful (at least since November 2016), so they aren’t going out and running virology tests on patients in the field. Neither are they collecting the existing patient data from Dr. Smith’s practice and a hundred other New York area practices and a thousand other practices around the country. What actually happens in the world is not evidence as far as they are concerned.
Doctors in the trenches are doing what bits they can of the data analysis that our AWOL CDC and NIAID should be doing
With Fauci and Redfield refusing to let NIAID and CDC do their jobs it gets left to the overworked doctors in the trenches to tally the small chunks of data from their own practices, yet even those small chunks of data show great therapeutic efficacy of strong statistical significance.
Smith says that his treatment results can be analyzed as an inter-cohort control. Out of 72 very sick patients the 20 who he had to put on ventilators had all been on Hydroxy treatments for two days or less. None of the patients who had done a full five day treatment went on a ventilator. His amateur statistician sons tell him the probability of that result being chance instead of a product of the treatment is “point zero, zero, zero, something.”
Assuming the CDC would want a more professional estimate Dr. Smith invites the CDC to come do its job and analyze his data for themselves.
“Point zero, zero, zero something” is pretty good but the CDC could really drive the uncertainty down by pulling together the case histories from hundreds of such practices.
Dr. Zelenko, an hour to the north of Dr. Smith, is up to 911 patients. As part of his ongoing effort to get the word out about the success he is having with his HCQ + Azithromycin + zinc Coronavirus treatments he tallied his own numbers in this April 4th YouTube video:
Zelenko (at 12:20): Now let’s do some statistics. Depending on which country you look at the mortality rate in the high risk patients is between five and ten percent. We’re an advanced country so let’s take the lower number of 5%. Out of those 350 patients [the number of Zelenko’s patients who he categorized as high risk] we statistically should have seen 17 dead people and a multiple of that on respirators. What did I see? Zero dead people and only three people on the respirator.
Set aside the exact statistical significance of that finding. If a full tally of all the case histories in all the practices in the New York area show similar reductions in mortality and morbidity the statistical significance/certainty of that finding will be extreme. Get it done. In the meantime Zelenko’s quick and dirty estimate is telling in itself.
Is a well controlled 1061 patient sample still “anecdote”?
Doctor Didier Raoult, the French doctor who Fauci originally discounted for running a small non-clinically controlled trial, just released a larger non-clinically controlled trial. For his control group Raoult used “patients treated with other regimens both at IHU and in all Marseille public hospitals,” which is a lot closer to Raoult’s treated-patient population than the control that Fauci used for his own research.
Raoult’s new trial involved 1061 patients and examined the effects of early intervention. His conclusion:
The HCQ-AZ combination, when started immediately after diagnosis, is a safe and efficient treatment for COVID-19, with a mortality rate of 0.5%, in elderly patients. It avoids worsening and clears virus persistence and contagiosity in most cases.
That’s the Case Fatality Rate, not the Infection Fatality rate, which is probably an order of magnitude lower. IFR for a normal flu is about .1% so the IFR of Corona when Raoult’s treatment is used should be about half that, and that is in elderly patients who have the highest Corona fatality rates.
Clinical trials don’t even know what to look at without field experience on what works
Some doctors, like Dr. Cardillo in Los Angeles, are finding that HCQ treatments are more effective with the zinc. Is the CDC planning to drag out its clinical trials by running separate trials for all the different drug combinations and dosages being tested now by doctors in the field?
They should do that, just don’t make us wait for it. Tally the field data now. It already exists and is perfectly adequately controlled and it has the numbers to yield a high level of statistical certainty.
The CDC and NIAID need to do their jobs and get on this. It is dereliction of duty that data compilation and analysis is being left for doctors in the trenches to pick up piecemeal.
For whatever reason Fauci and Redfield don’t want to analyze real world data. They need to be fired (if not court-martialed, to employ President Trump’s war metaphor) and replaced by people who will get the job done instead of blocking the job from getting done.
Only a very dishonest person can pretend that statistics are not valid unless they are based on clinical controls, which makes it not so surprising that the leading proponent of this claim is someone who professed the exact opposite until November 2016.