This is a research article published as information for health care professionals and public officials, and for an open peer review. It is not medical advice.
I reviewed the scientific literature on hydroxychloroquine (HCQ), azithromycin (AZ), and their use for COVID-19. My conclusions:
HCQ-based treatments are effective in treating COVID-19, unless started too late.
Studies, cited in opposition, have been misinterpreted, invalid, or worse.
HCQ and AZ are some of the most tested and safest prescription drugs.
Severe COVID-19 frequently causes cardiac effects, including heart arrhythmia. QTc prolonging drugs might amplify this tendency. Millions of people regularly take drugs having strong QTc prolongation effect, and neither FDA nor CDC bother to warn them. HCQ+AZ combination, probably has a mild QTc prolongation effect. Concerns over its negative effects, however minor, can be addressed by respecting contra-indications.
Effectiveness of HCQ-based treatment for COVID-19 is hampered by conditions that are presented as precautions, delaying the onset of treatment. For examples, some states require that COVID-19 patients be treated with HCQ exclusively in hospital settings.
The COVID-19 Treatment Panel of NIH evaded disclosure of the massive financial links of its members to Gilead Sciences, the manufacturer of a competing drug remdesivir. Among those who failed to disclose such links are 2 out of 3 of its co-chairs.
Despite all the attempts by certain authorities to prevent COVID-19 treatment with HCQ and HCQ+AZ, both components are approved by FDA, and doctors can prescribe them for COVID-19.
Hydroxychloroquine (HCQ) was accepted as a COVID-19 treatment by the medical community in the US and worldwide by early April. 67% of the US physicians said they would prescribe HCQ or chloroquine CQ for COVID-19 to a family member (Town Hall, 2020-04-08). An international poll of doctors rated HCQ the most effective coronavirus treatment (NY Post, 2020-04-02). On April 6, Peter Navarro told CNN that “Virtually Every COVID-19 Patient In New York Is Given Hydroxychloroquine.” This might explain decrease in COVID-19 deaths in the New York state after April 15. The time lag is because COVID-19 deaths happen on average 14 days after showing symptoms.
But on April 21, several perfectly coordinated events took place, attacking HCQ’s use for COVID-19 patients.
The COVID-19 Treatment Guidelines Panel of the National Institute of Health issued recommendations with negative-ambivalent stance regarding the use of HCQ as a COVID-19 treatment. This surprising stance was taken contrary to the ample evidence of the efficacy and safety of HCQ and despite absence evidence of its harm. The panel also strongly recommended against the use of hydroxychloroquine with azithromycin (AZ), the combination of choice among practitioners.
On the same day, a paper (Magagnoli, 2020) was posted on a pre-print server medRxiv, insinuating that HCQ is not only ineffective, but even harmful. This not-yet peer reviewed paper, by unqualified authors with conflicts of interest, received wall-to-wall media coverage, as it if were a cancer cure. It used data from Veterans Administration hospitals, spicing its effects. The paper has shown to be somewhere between junk science and fraud.
Rick Bright, a government official who was probably more responsible for the low level of preparedness to the epidemic than most others, and had been re-assigned to a lower position earlier, emerged as a “whistleblower.” He claimed he had been demoted for opposing hydroxychloroquine, the claim to be soon debunked by documents bearing his signature. The media also gave him a wall-to-wall coverage.
On April 24, the FDA struck its own blow, issuing a stern warning against use of HCQ for COVID-19 treatment.
While these warnings are not binding to doctors, they do produce a chilling effect. Consequently, either patients do not receive necessary treatment, or they receive it with a delay, sharper decreasing its effect. This allows detractors to question HCQ efficacy even more aggressively. Below, I review problems in the NIH COVID-19 Treatment Guidelines and other sources, used to wage anti-HCQ propaganda.
NIH Panel Guidelines
The relevant section of (COVID-19 Treatment Guidelines Panel, 2020) is Potential Antiviral Drugs. The antiviral treatment recommendations (more accurately, failure to provide recommendations) include:
- There are insufficient clinical data to recommend either for or against the use of the investigational antiviral agent remdesivir for the treatment of COVID-19 (AIII).
Clinical Data to Date:
Only anecdotal data are available.“
“AIII” means a strong position based on expert opinion rather than on evidence.
“Chloroquine or Hydroxychloroquine
- There are insufficient clinical data to recommend either for or against using chloroquine or hydroxychloroquine for the treatment of COVID-19 (AIII).
- When chloroquine or hydroxychloroquine is used, clinicians should monitor the patient for adverse effects (AEs), especially prolonged QTc interval (AIII).
Clinical Data in COVID-19
The clinical data available to date on the use of chloroquine and hydroxychloroquine to treat COVID-19 have been mostly from use in patients with mild, and in some cases, moderate disease; data on use of the drugs in patients with severe and critical COVID-19 are very limited.
[Follows is a description of some studies]“
Notice that CQ and HCQ are addressed together, although these are two different drugs, and HCQ is clearly superior to CQ both in efficiency and safety.
Also notice that the basic recommendation of “insufficient clinical data to recommend either for or against” is given to both HCQ and Remdesivir. However, the recommendation for HCQ goes further to state that when using HCQ, “clinicians should monitor the patient for adverse effects (AEs), especially prolonged QTc interval”. Practically, this means that HCQ should be used only in hospital settings. No such restrictions are set for Remdesivir, for which there is no clinical data available. It goes against all logic.
The demand to use HCQ only in hospital settings means:
HCQ treatment will be delayed until a patient decides to be admitted to a hospital, thus lowering HCQ’s efficiency
Hospitals will quickly become overwhelmed with COVID-19 patients
Then the Panel nixes HCQ+AZ:
“Hydroxychloroquine plus Azithromycin
- The COVID-19 Treatment Guidelines Panel recommends against the use of hydroxychloroquine plus azithromycin for the treatment of COVID-19, except in the context of a clinical trial (AIII).“
This drug combination is the most effective and widely used treatment for COVID-19, and the Panel recommends against it!
The Panel criticizes some studies of patients’ treatment with HCQ+AZ for the absence of a control group. Stephen McIntyre tweeted about this argument long before the Panel used it: “there’s a very large control group of COVID19 patients not receiving this drug combination: hospitals and morgues are full of them.”
There are only two studies, quoted by the Panel against HCQ+AZ, (Molina, 2020) and (Chorin, 2020). Both are misinterpreted by the Panel.
Molina et al.
Despite (Molina, 2020)’s angry tone and aggressiveness, it reports no results contradicting efficiency of HCQ or HCQ+AZ. The paper describes treatment of 11 hospitalized COVID-19 patients, five of which had cancer, one had AIDS, and almost all were in a bad shape: “at the time of treatment initiation, 10 of the 11 patients had a fever and received nasal oxygen therapy.” Using HCQ+AZ, 10 of the patients’ lives were saved. The article’s point of contention is that when they tested these patients, 5-6 days after the treatment initiation, they still found CoV2 RNA in 8 out of 10. Virus RNA is a molecule. Some viral RNA remains in patients for weeks after full recovery, but it is neither harmful nor infectious. Detecting viral RNA depends on the sensitivity of the testing equipment. The study’s title is No evidence of rapid antiviral clearance or clinical benefit with the combination of hydroxychloroquine and azithromycin in patients with severe COVID-19 infection seems to be lost on the Panel.
Chorin et al.
The Panel also quotes (Chorin, 2020) as evidence that HCQ+AZ therapy causes QTc prolongation. QTc prolongation is not a health condition itself, but a warning sign that a person is at higher risk of torsades de pointes (TdP), heart arrhythmia, or tachycardia, which might lead to cardiac arrest and death (Simpson, 2020).
Nevertheless, none of the patients, treated with HCQ+AZ, suffered TdP or arrhythmia. Four patients died, but none of them had an arrhythmia. Other studies, in which COVID-19 patients are treated with HCQ+AZ, reported taking patients off this medicine after QTc exceeds 500ms. But the treatment may have already had its effect at that time or later, while HCQ remained in the bloodstream.
This study has no control group. It provides no information on whether QTc prolongation was caused by the disease or the therapy.
(FDA WARNING, 2020), issued on April 24, piggybacks on the COVID-19 Panel Guidelines. It says
Hydroxychloroquine and chloroquine can cause abnormal heart rhythms such as QT interval prolongation and a dangerously rapid heart rate called ventricular tachycardia.
This statement is confused, and probably not true about hydroxychloroquine. See below.
Be aware that there are no proven treatments for COVID-19 …
I think that HCQ+AZ is a proven treatment for COVID-19. There is a difference between proven treatment and approved treatment. HCQ+AZ is not approved but proven, because many patients have been treated with this combination and have recovered.
We have reviewed case reports … concerning serious heart-related adverse events and death in patients with COVID-19 receiving hydroxychloroquine and chloroquine, either alone or combined with azithromycin or other QT prolonging medicines. These adverse events were reported from the hospital and outpatient settings for treating or preventing COVID-19, and included QT interval prolongation, ventricular tachycardia and ventricular fibrillation, and in some cases death.
These are manifestations of COVID-19! See (Bansal, 2020) and (Wang, et al., 2020). The media hysteria played its role, too. The articles about the supposed dangers of HCQ, with detailed description of the symptoms, triggered complaints even before the April 24 warning. And there are people who tried to self-medicate – in the situation when authorities make it difficult to obtain prescription for HCQ – and took the wrong drug or overdosed. Also, QT interval prolongation is not an event, but an early warning.
To help FDA track safety issues with medicines, we urge patients and health care professionals to report side effects involving hydroxychloroquine and chloroquine or other medicines to the FDA MedWatch program, using the information in the “Contact FDA” box at the bottom of the page.
Such an urging and advertisement guarantee that the FDA will receive mountains of complaints.
HCQ and AZ Safety
HCQ, CQ, and AZ
HCQ & CQ are two different drugs. HCQ is clearly superior to CQ. HCQ has already been selected over CQ. Discussing these two drugs as if they were co-equal in COVID-19 treatment is misleading and a sign of bad faith.
HCQ and AZ are some of the most widely prescribed drugs and have been prescribed for decades. HCQ is as safe as a prescription drug can be. AZ is an antibiotic, and it is as safe as an antibiotic can be.
Because these drugs have been prescribed so widely, their adverse effects have been studied. A few adverse events associated with them have been reported. Combining these few anecdotal cases, some medical researchers have raised some concern, as a precaution. Doctors understand this. Statisticians understand this. But unscrupulous media uses this information to mislead the naïve public and even public figures
Remdisivir is the opposite. It has been developed very recently and has been scarcely used. There is little information about its adverse effects. The corrupt news networks present this lack of evidence of adverse effects as evidence of the absence of adverse effects.
The leading objection against HCQ / HCQ+AZ is possible QTc prolongation. Most professionals refer to (CredibleMeds.org, 2020) which puts both HCQ and AZ in the category of Known Risk of TdP (KR).
I think that HCQ was listed in that category by mistake. A review of the literature reveals only few anecdotal cases. Some of them are poisoning by large overdoses of HCQ. Then there are patients who were on HCQ for years, suddenly got sick and recovered when HCQ was withdrawn. While there are millions of people continuously taking HCQ, only a few cases of cardiac events have been reported. Even if HCQ was the cause of these rare cases, which is usually unknown, it is still statistically insignificant. It is much safer than driving. Other antivirals are known to cause QTc prolongation too but are not being pulled from practice. In the case of HCQ, it seems that a precaution principle has prevailed over statistical reasoning and common sense.
AZ is in the KR category, just like many other antibiotics, including Erythromycin. I have never heard of patients requiring QTc monitoring, when taking Erythromycin.
Attention of the Trump Derangement Syndrome crowd: many widely used psycho-active drugs are also listed in the KR category. That includes anti-psychotic Haloperidol, anti-depressants Escitalopram (Cipralex, Lexapro) and Citalopram (Celexa).
American College of Cardiology
The most reliable source of information about arrhythmia risks is the American College of Cardiology. (Simpson, 2020) in the Cardiology Magazine:
Chloroquine, and its more contemporary derivative hydroxychloroquine, have remained in clinical use for more than a half-century as an effective therapy for treatment of some malarias, lupus, and rheumatoid arthritis. … Despite these suggestive findings, several hundred million courses of chloroquine have been used worldwide making it one of the most widely used drugs in history, without reports of arrhythmic death under World Health Organization surveillance.
HCQ is even milder than CQ.
Azithromycin, a frequently used macrolide antibiotics lacks strong pharmacodynamic evidence of iKr inhibition [associated with QT prolongation]. Epidemiologic studies have estimated an excess of 47 cardiovascular deaths which are presumed arrhythmic per 1 million completed courses, although recent studies suggest this may be overestimated.
In other words, after over 50 years of effective use, HCQ and AZ have proven their safety and efficacy. There is no reason for fear, except the fear itself. But some people might be vulnerable, so the article explains how to calculate an individual Risk Score for QTc prolongers. Individuals with higher Risk Score might need QTc monitoring. Also, the authors suggest avoiding other QTc prolonging medications in the time of HCQ+AZ treatment.
The cardiologists who wrote this article did not dismiss the concern. They explained the science pertaining to it and suggest proper mitigation measures.
Other literature also suggests low risk of HCQ and AZ. (Prutkin, 2020):
Limited data on hydroxychloroquine suggest it has a low risk of causing TdP, based on its use for rheumatoid arthritis, systemic lupus erythematosus, and antimalarial therapy. … For these medications [HCQ and AZ], their time window of use is short duration, which is another reason the risk of TdP may be lower
HCQ and AZ have other known contra-indications, but they are out of the scope here.
COVID-19 caused Arrhythmia
Many studies show that COVID-19 causes heart arrhythmia. Cardiac arrest, not directly caused by respiratory damage, is one of the leading direct causes of COVID-19 deaths.
(Bansal, 2020) is a review. It finds that
COVID-19 is primarily a respiratory illness but cardiovascular involvement can occur through several mechanisms.
Acute cardiac injury is the most reported cardiovascular abnormality in COVID-19, with average incidence 8-12%
Both tachy- and brady-arrhythmias are known to occur in COVID-19. A study describing clinical profile and outcomes in 138 Chinese patients with COVID-19 reported 16.7% incidence of arrhythmia. The incidence was much higher (44.4%) in those requiring ICU admission …
It also notes that CoV2 virus might cause cardiac injury directly or indirectly. The possibility of a treatment impact is mentioned as a less likely one.
(Wang, et al., 2020) finds that 44% of the patients transferred to ICU developed arrhythmia. None of them received HCQ or CQ. Most of the patients received an unrelated anti-viral and an antibiotic. Only in 18% of the patients the antibiotic was AZ. At least some of the patients developed an arrhythmia before the treatment.
Doctors have found that the infection can mimic a heart attack. They have taken patients to the cardiac catheterization lab to clear a suspected blockage, only to find the patient wasn’t really experiencing a heart attack but had COVID-19.
Thus, the hypothesis that CVOID-19 patients experience QTc prolongation and arrhythmia because of the disease, rather than due to HCQ+AZ treatment, is well founded. AZ may increase the odds of QTc prolongation in COVID-19 patients, who would otherwise die from cardiac arrest or multiple organs failure.
The media and professional publications report a sharp increase of mortality from cardiac arrest at home in the last few weeks. Some of these cases are known to be COVID-19, but most of them are not tested. Could many of them be happening due to the cardiac damage caused by COVID-19? Can the cardiac impact of COVID-19 be aggravated by strong QTc prolongers that many people take regularly? There are countless variables confounding this statistic. There is an especially sharp increase in home cardiac arrests in New York, which is usually explained by people’s reluctance to call an ambulance or ER.
(Kochi, 2020) provides in-depth explanation of the cardiac effects of respiratory infections and interaction with QTc prolongation medications.
Positive Cardiac Effects of HCQ
Gone unmentioned are HCQ’s positive cardiac effects. They were widely reported before HCQ had misfortune of being mentioned by President Trump. For example, Taking Hydroxychloroquine for RA or Lupus Can Reduce Heart Risk by 17%
If you take the anti-malarial drug hydroxychloroquine (Plaquenil) as part of your treatment for lupus or rheumatoid arthritis (RA), you may be getting cardiovascular protection as an added bonus.
The article is based on (Jorge, 2019). These findings might be applicable only to long term taking of HCQ, not a 5-day course for COVID-19, but the same can be said about the alleged negative cardiac effects.
Articles/Studies criticizing HCQ
Listed here are several other papers, influential in the media, but not in the science. These papers span the range from erroneous to … non-existent.
Magagnoli et al.
(Magagnoli, 2020) is a not peer-reviewed pre-print. It makes a retrospective statistical comparison of the outcome in COVID-19 patients, who received HCQ or HCQ+AZ treatment prior to April 11, in Veterans Affairs hospitals. In the Abstract, it claims that a larger percentage of HCQ treated patients died compared to untreated patients. This ignores the fact that HCQ or HCQ+AZ treatment was given only in the most desperate cases, frequently as compassionate care. Deep inside of the manuscript, it does acknowledge that initial conditions of the HCQ and HCQ+AZ groups was much worse than those of the untreated group, but then ignores it
The original version (archived) of the “study” was published on April 21. It received crushing criticism in the comments and was replaced with another one on April 23, hiding those comments. Casting even further doubt on the credibility of this study, one of the authors disclosed Gilead funding for another research. This work was funded by a NIH grant.
Despite its multiple flaws, lack of peer review, and obscurity of the authors, this pre-print immediately received wall-to-wall media coverage. Given these circumstances, this work looks like a criminal fraud, rather than a scientific one.
Tang et al.
(Tang, 2020) is a not peer-reviewed pre-print. It reports results of a clinical trial in China, in which HCQ was given to patients 16-17 days after onset of the disease. This is too late for an anti-viral to work. Thus, this study describes the incorrect use of HCQ, rather than efficacy or safety of the drug. From the comments:
With an average delay of 16 days from symptom onset to enrollment and treatment in this trial, those patients are pretty much past the viral phase of the disease, where an antiviral treatment would have the most value, and are well on their way to pneumonia and a cytokine storm problem, which is ultimately what kills.
Once again, despite its obvious errors, the study was widely covered, including the New York Times and LA Times. Neither headline nor article addresses the obvious lateness of the drug’s application.
Mahevas et al.
(Mahevas, 2020) is another not peer-reviewed pre-print. Didier Raoult and his colleagues replied to it with a bluntness, rare in scientific journals: Scientific fraud to demonstrate the lack of efficacy of hydroxychloroquine compared to placebo in a non-randomized retrospective cohort of patients with Covid: Response to MAHEVAS et al. , MedRxiv, 2020. (Brouqui, et al., 2020). (Mahevas, 2020) also gathered many negative comments on MedRxiv.
Oral Statements of Holtgrave & Cuomo
A study of 600 patients at 22 hospitals in New York is being conducted by the University at Albany School of Public Health under the management of dean David Holtgrave. Although the study was not finished, Mr. Holtgrave already announced that the results are negative: “We don’t see a statistically significant difference between patients who took the drugs [HCQ, HCQ+AZ] and those who did not,” according to CNN. New York Governor Andrew Cuomo referred to the results as neither positive nor negative, per CNN and ABC.
No paper, or even pre-print, reporting these results, has been published, as of April 29 (searches on Google Scholar, PubMed, and medRxiv were conducted for Holtgrave hydroxychloroquine; Holtgrave COVID-19).
New York and other “resistance” states make patients jump through hoops to obtain HCQ. As an anti-viral, it should be taken as soon as possible. Dr. Vladimir Zelenko explained that in his letter, which is worth reading in its entirety:
It is essential to start treatment against Covid-19 immediately upon clinical suspicion of infection and not to wait for confirmatory testing. There is a very narrow window of opportunity to eliminate the virus before pulmonary complications begin. The waiting to treat is the essence of the problem.
He refers to patients in the high-risk category – older than 60, having certain health conditions, or shortness of breath. The resistance states established onerous requirements that delay HCQ treatment for days. This sharply lowers the efficiency of the treatment, and possibly increases TdP risks. The mixed results, promised by Mr. Holtgrave, might be caused by this delay.
On March 28, Russia announced a COVID-19 treatment based on Mefloquine. Mefloquine, invented in the US in 1970s, is another anti-malaria drug, similar to HCQ. In the West, Mefloquine was withdrawn from use after a controversy about its long-term effects. Russia might also use HCQ. From a Russian brochure (Nikiforov, 2020):
These drugs have a comprehensive negative effect on the coronavirus. It may take years of scientific experimentation to understand how and what exactly they affect. Now the fact of a positive effect has been established, and the drugs should and will be used.
The mechanisms of HCQ and HCQ+AZ action are explained (Hache & Raoult, 2020).
On March 27, WHO erected another roadblock to treating COVID-19 patients with HCQ. WHO stated that HCQ was not only insufficiently tested (which was true at that time), but that it was considered for COVID-19 at much higher doses than for malaria.
In the context of the COVID-19 response, the dosage and treatment schedules for chloroquine and hydroxychloroquine that are currently under consideration do not reflect those used for treating patients with malaria. The ingestion of high doses of these medicines may be associated with adverse or seriously adverse health outcomes.
This is dangerous misinformation. HCQ dosage for COVID-19 is the same or lower than for malaria (Drugs.com, 2019). WHO was aware of this, because it was already conducting clinical trials including HCQ and a number of other Big Pharma drugs. Yet, as of April 29, this paragraph still appears there. This act alone justifies not only defunding but ignoring WHO.
Google and Facebook adhered to WHO on everything related to COVID-19. Together with Twitter, they purged information favorable to HCQ. This is outrageous behavior for telecommunications and computational services providers.
It seems that the main contra-indication for HCQ treatment of COVID-19 is that no treatment is needed for healthy individuals below age 50.
Persons in the President’s circle were claiming that HCQ / HCQ+AZ are unproven treatments. That might have been true a month ago, but not now. These drugs are proven by practice and by failure of its opponents to disprove their efficacy and relative safety.
The Guidelines are accompanied by a financial disclosure of the panel members. Weirdly, this disclosure covers a period of 11 months: May 1, 2019 to March 31, 2020. The latest three weeks were excluded for some reason. Nevertheless, 9 out of 50 members of the panel disclosed financial ties to Gilead. Gilead’s Remdesivir is an inferior competitor to HCQ – more expensive, almost untested, and less efficient (as far as the little testing with it has shown). HCQ is a generic drug with low profit margin. Gilead Sciences directly participates in WHO trials of Remdesivir as a COVID-19 treatment.
HCQ / HCQ+AZ are prescribed by a doctor. They are not OTC and should not be used for self-medication.
HCQ+AZ is the most common treatment. HCQ acts on its own but is much more effective with Zinc; AZ is an antibiotic and a source of Zinc. See Dr. Zelenko’s regimen is HCQ+AZ+Zinc.
There is a live document by Michael J. A. Robb, M.D., tracking effectiveness of HCQ-based treatments https://drive.google.com/file/d/1w6p_HqRXCrW0_wYNK7m_zpQLbBVYcvVU/view