I discussed the use of hydroxychloroquine and azithromycin for patients with COVID19 on my March 27th edition of This Week in Cardiology Podcast.
This is an important topic not only because of the specifics of treating patients but also vital because it shows how easily human beings can be misled.
Here is a an edited transcript of my words:
A conversation I had with my Dad this week made me realize the seriousness of this matter. My Dad is a retired insurance executive with a background in electrical engineering.
He is smart, but I could not convince him that the evidence prompting people to advocate for this drug violates the principles of eighth-grade science class.
Dr. Fred Masuodi, a thoughtful cardiologist, wrote with candor.
I am not shocked that our president can be misled about shoddy evidence but I am dismayed that so many doctors have been bamboozled.
Note, I am publishing this column two weeks after the podcast and still doctors in my city and across the world are using this drug combination. In fact, our hospital and others have had to design ways to monitor for cardiac side effects from the drugs.
See this Tweet:
A few quick words on the French study that sparked the hype of hydroxychloroquine (HQ)/azithromycin (AZ).
Originally published in preprint form, the paper was published the next day in a journal in which one of the authors was editor-in-chief.
The problems with this study are numerous:
First, it was not randomized. 26 COIVID + patients hospitalized in Marseille were in the active arm. 16 patients in surrounding areas served as controls.
Of the 26 HQ treated patients, 20 completed the study. Of these, 6 also received azithromycin.
The endpoint of the study was not clinical. Rather it was the surrogate endpoint of viral load. The duration of the study was 6 daysâ€”even though the original plan was for eval of PCR data on Days 1, 4, 7, 14.
Main result: On day 6, most of the 16 control patients vs about half of the 14 HQ treated patients, and none of the 6 HQ+AZ treated patients were PCR positive. The authors concluded the drug regimen helped clear the virus.
Here is the kicker â€“ 6 of the 26 patients in the active arm were not followed-up. But not because they were lost; they were not included because 3 worsened and required ICU care, 1 died, and 1 stopped because of nausea.
Imagine if these 5 of 6 patients were included in the results. My friend Luis Correia from Salvador called the practice of taking patients who do poorly out of your analysis an “unstudy.” An eighth-grade student would fail her science project if she removed selected subjects from the experimental arm.
My summary: So this non-randomized, surely confounded, tiny study, with surrogate endpoints, published in one day in a journal with an author who was editor in chief, with 5 of 6 active arm patients who worsened and were not included in the final analysis has caused doctors, people who are supposed to be trained in science, to promote this potentially dangerous combination boggles my mind.
In a NEJM editorial Dr. Fauci and others outline lots of potential treatments for COVID19. In a disease that well more than 90% of people survive you cannot just treat a handful of patients and say, see, they got better.
COVID19 is not heart disease. We donâ€™t need to follow patients for years to get an answer, The disease course is over in weeks. That means proper randomization will give answers by late spring and summer.
Not doing proper science would only compound the crisis this virus has caused.
That’s the end of the podcast transcript. Here is an update over the past two weeks.
The same French investigators have published another paper now with 80 patients rather than 26. Again the endpoint was not a clinical endpoint, such as not dying or not being admitted to an ICU but viral load.
From the abstract: A rapid fall of nasopharyngeal viral load tested by qPCR was noted, with 83% negative at Day7, and 93% at Day8. Virus cultures from patient respiratory samples were negative in 97.5% patients at Day5
This case series adds no useful evidence because without a control group you don’t whether the same number of patients would be virus free with no therapy.
Please note that I am not the only one with a problem with this study. Ivan Oransky writes on his blog Retraction Watch that the International Journal of Antimicrobial Agents finds that the French article
does not meet the [International Society of Antimicrobial Chemotherapyâ€™s] expected standard, especially relating to the lack of better explanations of the inclusion criteria and the triage of patients to ensure patient safety.
Other scientists have posted critiques on PubPeer.
My friends, this is not about the nitty-gritty of medical evidence. This is about basic scientific principles that underpin the modern practice of medicine.
What separates the doctor of today from the faith healers of medieval times is the use of evidence.
I understand that this virus creates fear. But that is not a reason to abandon the scientific method and evidenced-based practice.
Like you, I spent the first few weeks of this crisis anxious and fearful. Now, I worry as much about the fact that my colleagues have accepted this therapy based on what is more like an unstudy than a real study. What does it say about the medical profession that normal doctors are doing this?
For the record, I think it is highly unlikely, though not impossible, that this drug combination helps people with COVID19.
But given the toxicity of these drugs, especially in older patients with or susceptible to heart disease, the only proper action is to randomize patients to either standard care or the drug combination and follow them for the 2-4 week course.
The statistician Darren Dahly has an excellent and short video explainer on the basics of proper trial.