Science vs Nonscience In Understanding Hydroxychloroquine

A recently much shared video of America’s Frontline Doctors Summit shows several clinicians claiming that hydroxychloroquine (HCQ) is a "cure" for COVID-19. Their evidence for this claim is their personal experiences treating these patients with HCQ along with a cocktail of other drugs. The video was shared 14 million times before Facebook and other hosting services took it down, ostensibly for spreading inaccurate information. While the debate whether such censorship is reasonable, it is also reasonable to point out that the doctors were, in fact, sharing information that is not known to be accurate and that has a high chance to be inaccurate; in other words, scientifically unproven. They grossly overstated their conclusions from their non-scientific observations. It is irresponsible for a doctor to claim that her anecdotal experience “proves” the efficacy of any unproven therapy. If the docs had been honest, they should have said that their observations warranted further controlled clinical trials in order to prove or disprove their claims. 

One of these doctors said that she had treated 350 patients with the drug cocktail and that none died; therefore, she irresponsibly declared that she had a “cure” for the disease. While that sounds impressive, she did not do a controlled clinical trial, which means that we have no way of knowing whether the 350 also would have survived without the drug. She also claimed that her success was because the patients she treated with the cocktail were at the early stage of disease. Unfortunately, we don’t know what that means since she didn't report their clinical details as she would have been required to do in a gold standard clinical trial. This doctor also had a web page where she talked about how gynecological problems are caused by engaging in sex with demons, and that alien DNA was being used in modern medicines; all the more reason to suspect her credibility.

In other venues, I have pointed out these problems with the America’s Frontline Doctors Summit and received a lot of push-back from non-scientists. For some reason, some people bring a strong need to believe in HCQ without considering the science. So, they readily jump on ANY report that confirms their bias as proving that the drug is a cure. Several of these “experts” quickly pointed me to a recent report from the Henry Ford Health System that claimed that hydroxychloroquine saved lives. It was published earlier this month in the International Journal of Infectious Diseases. 

However, the Henry Ford report was not a clinical trial, but a much weaker retrospective chart review of more than a thousand COVID-19 patients seen in the system’s nine hospitals. In other words, the patients were not randomized, the “study” was not blinded, and patients were not treated according to a controlled, standardized study protocol. Hence, it was only marginally better than the undocumented anecdotes of America’s Frontline Doctors. At least, because the Henry Ford docs published a report on their personal observations, interested clinicians and researchers could look at the aggregate patient data. That was not the case regarding the claims by the America’s Frontline Doctors.

On Wednesday, in response to the Henry Ford report, the same journal published several scathing critiques claiming the report had serious errors. The major problem was that the patients given the HCQ cocktail regimen were healthier than the patients that were not given it. The patients not given the cocktail had more advanced disease and more frequent comorbidities that put them in a higher risk group compared to those who received the treatment. Furthermore, the HCQ treatment group was more aggressively treated and more than twice as likely to receive steroid therapy, which has been shown to help certain COVID-19 patients.

In other words, in this chart review of patient experiences, the two groups that were compared were very different and it is highly possible that the death rate difference between them would have been the same even if the HCQ protocol was not used. An important goal of randomized, blinded, controlled clinical trials is to make sure that the treatment and non-treatment comparison groups are as similar as possible in order to eliminate such bias that can skew the study’s results. This is why scientific clinical trials and not chart review reports are the gold standard for determining the best health care.

Unfortunately, people who take these anecdotal testimonies, and poorly controlled chart review reports as proof that HCQ is the panacea for COVID-19, also selectively ignore other recently reported gold standard clinical trials that show that HCQ is ineffective. This week, a randomized clinical trial in Brazil showed that hydroxychloroquine doesn't work to treat patients with Covid-19. Another randomized trial last month at the University of Minnesota showed it also doesn't help prevent infection. Other clinical trials -- one in the US, and one in the UK were halted early because interim data analysis showed the drug wasn’t working.

Like America’s Frontline Doctors emphasized, the authors of the Henry Ford report pointed out that HCQ worked for their patients because it was prescribed very early in their hospitalization. But, University of Albany researchers earlier reported that the HCQ cocktail approach was ineffective in a randomized, blinded trial employing subjects at the same point of disease as the patients in the Henry Ford report.

As reported earlier in these pages, because of similar negative clinical trials, the FDA recently pulled its approval for HCQ to treat COVID-19.

There is a very good reason why we rely on carefully designed and controlled clinical trials rather than anecdotal information or retrospective chart reviews in determining the best way to treat disease. The best the latter should contribute is to generate interest in testing the observations in controlled clinical trials to see if they are accurate.

While your humble blogger initially was enthused about the potential of HCQ to treat COVID-19, I increasingly sour on it as science continues to show it doesn’t work. So far, the reports claiming that HCQ is effective against COVID-19 are mostly based on unsubstantiated doctor’s anecdotes, or on uncontrolled retrospective chart reviews. In contrast, the reports that indicate that HCQ is ineffective are based on stronger randomized and controlled clinical trials.

Who are you going to believe?