Skip to main content

Dr Silverberg Shares His Thoughts on Evidence-based Medicine and Hydroxychloroquine During the Pandemic

Jonathan Silverberg, MD, PhD, MPH, discusses rise and fall of hydroxychloroquine as the “wonder drug” for COVID-19, and the importance of evidence-based medicine in this podcast.

Dr Silverberg is an associate professor of dermatology and director of clinical research and contact dermatitis at George Washington University School of Medicine and Health Sciences.


1. Gao J, Tian Z, Yang X. Breakthrough: Chloroquine phosphate has shown apparent efficacy in treatment of COVID‐19 associated pneumonia in clinical studies. Biosci Trends. 2020;14(1):72- 2. doi:10.5582/bst.2020.01047

3. Gautret P, Lagier JC, Parola P, et al. Clinical and microbiological effect of a combination of hydroxychloroquine and azithromycin in 80 COVID-19 patients with at least a six-day follow up: A pilot observational study [published online ahead of print, 2020 Apr 11]. Travel Med Infect Dis. doi:10.1016/j.tmaid.2020.101663

4. Gautret P, Lagier J‐C, Parola P, et al. Hydroxychloroquine and azithromycin as a treatment of COVID‐19: results of an open‐label non‐randomized clinical trial [published online ahead of print March 20, 2020] Int J Antimicrob Agents. doi:10.1016/j.ijantimicag.2020.105949

5. Chen Z, Hu J, Zhang Z, et al. Efficacy of hydroxychloroquine in patients with COVID-19: results of a randomized clinical trial [published online ahead of print April 10, 2020]. medRxiv. doi:10.1101/2020.03.22.20040758

6. Rosenberg ES, Dufort EM, Udo T, et al. Association of Treatment With Hydroxychloroquine or Azithromycin With In-Hospital Mortality in Patients With COVID-19 in New York State. JAMA. Published online May 11, 2020. doi:10.1001/jama.2020.8630

7. Borba MGS, Val FFA, Sampaio VS, et al. Effect of high vs low doses of chloroquine diphosphate as adjunctive therapy for patients hospitalized with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection: A randomized clinical trial. JAMA Netw Open. 2020;3(4):e208857. doi:10.1001/jamanetworkopen.2020.8857

8. Geleris J, Sun Y, Platt J, et al. Observational study of hydroxychloroquine in hospitalized patients with Covid-19 [published online May 7, 2020]. N Engl J Med. doi:10.1056/NEJMoa2012410

9. Shi S, Qin M, Shen B, et al. Association of cardiac injury with mortality in hospitalized patients with COVID-19 in Wuhan, China [published online March 25, 2020]. JAMA Cardiol. doi:10.1001/jamacardio.2020.0950

10. Driggin E, Madhavan MV, Bikdeli B, et al. Cardiovascular considerations for patients, health care workers, and health systems during the COVID-19 pandemic. J Am Coll Cardiol. 2020;75(18):2352‐2371. doi:10.1016/j.jacc.2020.03.031

11. Chan KS, Lai ST, Chu CM, et al. Treatment of severe acute respiratory syndrome with lopinavir/ritonavir: a multicentre retrospective matched cohort study. Hong Kong Med J. 2003;9(6):399‐406.

12. Wongcha-um P. Cocktail of flu, HIV drugs appears to help fight coronavirus: Thai doctors. Reuters. Published February 2, 2020. Accessed May 14, 2020.

13. Cao B, Wang Y, Wen D, et al. A trial of lopinavir-ritonavir in adults hospitalized with severe Covid-19. N Engl J Med. 2020;382(19):1787‐1799. doi:10.1056/NEJMoa2001282

14. National Institute of Allergy and Infectious Diseases. A multicenter, adaptive, randomized blinded controlled trial of the safety and efficacy of investigational therapeutics for the treatment of COVID-19 in hospitalized adults. NCT04280705. Accessed May 14, 2020.

15. NIH clinical trial shows Remdesivir accelerates recovery from advanced COVID-19 [press release]. National Institutes of Health. published April 29, 2020. Accessed May 14, 2020.

16. ABC News. Dr. Anthony Fauci touts results of trial examining experimental COVID-19 drug treatment. Published April 29, 2020. Accessed May 14, 2020.

17. Wang Y, Zhang D, Du G, et al. Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial [published online April 29, 2020]. Lancet. doi:10.1016/S0140-6736(20)31022-9

18. Roston A, Taylor M. CDC removes unusual guidance to doctors about drug favored by Trump. Reuters. Published April 7, 2020. Accessed May 14, 2020.

19. FDA cautions against use of hydroxychloroquine or chloroquine for COVID-19 outside of the hospital setting or a clinical trial due to risk of heart rhythm problems. Updated April 30, 2020. Accessed May 14, 2020.

20. Russell B, Moss C, Rigg A, Van Hemelrijck M. COVID-19 and treatment with NSAIDs and corticosteroids: should we be limiting their use in the clinical setting? [published Mar 30, 2020]. Ecancermedicalscience. doi:10.3332/ecancer.2020.1023


Melissa: Hello. I’m Melissa, associate editor of The Dermatologist. Dr Silverberg is joining us to discuss evidence‑based medicine and the rise and fall of hydroxychloroquine as a therapy for COVID‑19. Dr Silverberg is an associate professor of dermatology and director of clinical research and contact dermatitis at George Washington University School of Medicine and Health Sciences.

Dr Silverberg: Today. I’m going to be discussing some of the data and the evolving story around hydroxychloroquine usage, potentially other antimalarials for COVID‑19. I think this is a very provocative topic and one that is rapidly evolving. I look at this at the evolution in three phases. One was the initial phase, which really was this…you know…some preliminary data led to huge optimism.

Then there was the intermediate phase, which was more of a tempered enthusiasm, but there was a thought that, “well, even if the efficacy doesn’t work so well, but at least it’s safe.” Now, we’re getting into a more realistic, evidence‑based conversation that suggests that maybe it’s actually more harmful than good in certain circumstances.

Let me take you through some of the early data just so you get a sense of this because you’re going to hear a lot of mixed messages depending on at what stage you’re reading an article, where that was developed, what study is being reviewed, and how that all evolved.

The initial studies that indicated that hydroxychloroquine might be effective really had a number of limitations. It was a small cohort, very few subjects, a little bit of a mixed intervention.1 Some only got hydroxychloroquine, some got hydroxychloroquine plus azithromycin, and very a short observation period; you’re talking about they only followed them for about six days.2 Not a randomized study, a lot of bias that comes along with that, and really no great reporting on the effects on clinical evolution, on details about viral load, and other aspects of the disease, but very exciting.

It’s obviously a very scary and difficult time for everyone out there right now in terms of how to deal with this. This was a sorely needed positive data that gave us a lot of hope and enthusiasm. Then the second French study was a little larger but had no control arm. There were some issues there in terms of the quality of reporting that came up. Really didn’t show such amazing results, had some reduced viral load, but no other clinically relevant outcome.3 Now all of a sudden, this had more limitations in certain respects. All of a sudden, we have this tempered effect where, “Well, OK, the next study is not really showing such great results.”

Then you had a study that came out of China, randomized controlled trial there that looked at hydroxychloroquine.4 Some of them were in combination with other antiviral agents, which led to certain issues there. They found no significant differences in terms of viral carriage and transmission potential compared to whatever the local standard of care was.

All of a sudden, you have now a few studies that are showing maybe this doesn’t work quite so well. Then you have a subsequent study that came out and showed that there was a shorter time to clinical recovery. Really didn’t show much in terms of mortality improvement, but perhaps a little bit of a speedier recovery.5

The analogy that you might think of to what we already know about in the viral world would be something like Tamiflu for the flu where it doesn’t improve mortality, but it does at least hasten recovery. Perhaps it does speed up recovery.

There was also some issues there that came up in terms of generalizability with respect to the severity of the disease, and really all throughout this whole process. It’s unclear about when hydroxychloroquine, if it was effective at all, should be started.

You have this initial huge optimism. Lots of media reports, lots of discussion between myself and my colleagues about, “this might be the wonder drug. I’m going to stock up on Plaquenil and hydroxychloroquine. I’m going to have it there. I can go see my patients. I don’t have to worry about getting exposed because this is a wonder drug that’s going to fix everything.”

Then the data show, “well, no, maybe not,” because even in those studies that had positive results, really modest benefit, modest efficacy, really not dramatic effect sizes in terms of improving mortality, or even in terms of hastening recovery. Unclear benefit in a lot of these different studies.

You look at the data at that point, brings into step two of, “Well, OK, maybe it’s not this wonder drug, but I have patients who are sick. What do I do for a family member who’s sick? Do I do nothing? Well, it’s better than doing nothing. I might as well give them hydroxychloroquine, because we use it in dermatology quite a bit. It’s used a lot in rheumatology. It’s a safe drug. We don’t really see that many side effects in the real world. What do we have to lose? Even if it doesn’t work so well, certainly there’s no harm. We don’t have to worry about it.”

Then come several studies, which really start raising some concerns. Studies showing that there’s more cardiac toxicity, some studies even being halted due to cardiac toxicity in COVID‑19 patients.6-8 For context, we don’t see that much cardiac toxicity in the real world when we’re using it for our lupus patients and connective tissue disease, etc.

COVID‑19 is a different disease where there appear to be cardiac complications.9,10 The risk in this disease may be very different than what we’re used to in the dermatology setting. All of a sudden, there’s studies showing that there’s really cardiac toxicity here that we do have to worry about.

Then some say, “Well, I never see cardiac toxicity. Come on, this can’t be real. It must be a fluke. It’s not a real finding.” Now, you’re starting to see studies that come out, showing that there may actually be higher mortality rates related to the use of hydroxychloroquine to treat COVID‑19 patients.

Now, not only are we seeing studies showing that there may be no real demonstrable benefit, but they may actually cause more harm than good and kill more patients. There are severe limitations to some of these studies as well. Some of these are retrospective studies looking at real‑world data, but I don’t think that the signals can be discounted. We have to be really concerned here about the potential toxicities in this relatively new and unexplored disease. I look at this and we’re now at a new stage which says, “Oh boy, we actually don’t know what the right answer here is,” and our not so cautious optimism early on was a little bit premature.

When we actually see the evidence, it doesn’t point to a very compelling story. As a clinician, of course, we can always prescribe what we feel is needed for our patients. In the critical care setting, there may be no other option to try. This is something we have to be very careful about. We have to be very careful about recommending for all patients without more clear-cut safety data and understanding of the risk‑benefit ratio.

Also, understanding that this perhaps false optimism about the value of hydroxychloroquine is causing the run on the drug. It caused the temporary inability for our patient who are on it for years to be able to get their prescriptions refilled and caused potential relapses and safety concerns for those patients.

There’s an overall take‑home message, because this is not just the story about hydroxychloroquine and COVID‑19. This evolution is really consistent with what we see all throughout medicine and the conversations about evidence‑based medicine. There are many who are on different sides of the coin with this. Some are big proponents of evidence‑based medicine. You can probably tell from my tone that I am a very loyalist, a big loyalist and fan of evidence‑based medicine. These principles evolved over the years for good reason. We have clear methodology developed for a reason.

In dermatology, we’re used to having diseases that sometimes don’t even have names, let alone evidence to support their treatment. We’re all too quick to pick treatment choices based on case reports or case series, or to move our own treatment decisions based on a single study, which is a big no‑no in the world of evidence‑based medicine.

What you see here is the value of evidence‑based medicine. You see why it is that we push for evidence‑based medicine principles throughout all of medicine and why we need more of a push even in dermatology. Sometimes that first study that gets published due to bias, due to other reasons isn’t replicable. Even if one or two studies are replicable, sometimes you find out about unexpected harms down the road with better studies. It’s a cautionary tale. But, where we’re at right now, with respect to the evidence for hydroxychloroquine in COVID‑19, is very mixed results.

For many patients, it may be better off doing nothing, and using the standard of care, and not pushing the wanting use of hydroxychloroquine without clear‑cut evidence. Of course, the beautiful thing and the challenging thing about evidence‑based medicine is it’s always evolving. There’s now over 70 trials for hydroxychloroquine being conducted globally. We will have an enormous amount of evidence, coming out soon enough that will help guide this story even more.

It’s important for us to understand that, and the way I think of it is, if I have a choice between social distancing, avoiding those potential risky exposures, versus taking my chances but keeping a prescription of hydroxychloroquine in my back pocket, because that’s going to somehow protect me it’s a very clear choice that the avoidance of exposure and the social distancing is the far more effective strategy.

Comes with its own challenges as we are all facing in dermatology and all throughout medicine. From a safety standpoint, there is no comparison that social distancing, and that avoidance of exposure, of high‑risk exposures especially in the healthcare setting, will provide a far greater protection than the modest, at best effect size of hydroxychloroquine in COVID‑19.

Melissa: Were there any other antivirals being investigated for COVID‑19?

Dr Silverberg: Even before hydroxychloroquine was being considered as this wonder drug for COVID‑19, there were a number of other agents that were similarly being touted as going to be the panacea for COVID. One was Kaletra, the HIV drug. There were a lot of reports coming out of Asia that the combination of Tamiflu plus Kaletra was really a magical cocktail and was saving lives.11,12

Then unfortunately, the randomized controlled trial that got published in New England Journal showed absolutely zero benefit.13 Again, similar story, our anecdotal experience in a disease that spontaneously improves quite a bit is such that: “Yeah, we don’t know when patients are getting better. Is it because of the drug or because they would have gotten better on their own anyway?”

Whereas, in the randomized controlled trials, which is, of course the gold standard for evidence, there really was no benefit. Right now, we’re not seeing any clear‑cut signals of efficacy for any of the existing HIV drugs, for Tamiflu, or any of the other ones out there.

Unfortunately, those are laid to rest. We’ll see more studies coming over the next few months that were already started and will get readout soon enough. At least from the anecdotal experience of those who are on the front lines, as well as the available data, there really doesn’t seem to be any value for those drugs.

I think the more interesting one, well maybe the more politically interesting one, is remdesivir, which is the newer kid on the block that was studied for COVID‑19. This has a little bit of an interesting story. The initial study really faced a lot of setbacks, challenges in terms of recruitment. No clear‑cut evidence of efficacy, and really was very disappointing.

Just a few days ago, there was a press release by Tony Fauci from the CDC that’s reviewing the recent results from a NIAID‑sponsored study, looking at remdesivir in COVID‑19.14 Keeping in mind that these are interim analyses, these are not final, there’s more data that needs to be seen. What they found was that there was a modest improvement in terms of time to recovery. A 31% faster time to recovery with remdesivir compared to the placebo group.15 In the press release, Dr Fauci emphasized to those methodologists out there, there’s a P value of less than 0.001.16

That’s nice, but if you ask any statistician, we don’t judge based on P values. We judge really based on effect size more than anything else. This effect size is really quite modest at best.

What they found was that the median time to recovery was 11 days for patients treated with remdesivir compared to 15 days for those on placebo, but actually a non‑significant, it was close to significant, but a non‑significant mortality benefit.17 This was a study that was decently powered. Really not incredibly impressive results.

It may be the best we have right now, it may have set the bar. When you’re looking at a modest benefit like this in terms of mortality and in terms of hastening recovery, again, this is similar to the story that we see with Tamiflu with the flu. It doesn’t really improve mortality much, but at least it does hasten recovery and improves the symptoms a little faster.

I look at these data, and I’m a little trepidatious because there’s so much optimism. Now there are so many folks out there who are pushing to open up every state government, open up every practice, and open up every retail outlet because “we have this drug that the CDC is showing us that works so well.”

If you actually look at the data, it doesn’t work so well. It’s modest benefit, at best. We’ll have many more studies to see results for. When you look at these kinds of studies and you see really these modest benefits, it’s actually concerning. I hope it is a first step to understanding how to treat this disease, and maybe we get even better drugs that are more effective. I don’t look at these results and say, “OK, the evidence here points to anything magical.”

Melissa: From an ethics standpoint, in the middle of a pandemic, where the question becomes, do I give a patient something or do I do the randomized controlled trial, do you have any thoughts on that part?

Dr Silverberg: One of the fundamental ethical questions that comes up in this type of situation. This is not just true in an outbreak setting. It can be true even in, let’s say, a chronic disease setting or a cancer setting where preliminary data about a new drug that might drastically reduce mortality. It’s based on preliminary data and you don’t yet have readouts from major trials.

Should you be allowed to use those kinds of drugs on a compassionate care basis? Should you be allowed to prescribe them off‑label with the hopes that it may do better than standard of care? It’s a very interesting ethical discussion.

To anyone who’s on the front line, I should say it’s not a clear answer. The clear answer is that the clinician has the license to do whatever they want to do, whatever they feel is best for their patients. To those clinicians who are on the frontlines, in areas that were just getting ravaged. They just could not keep up with it.

They were doing everything they could to try to keep patients off of ventilators and keep them out of the ICU. There are admittedly data, weak, but preliminary reports suggesting that using some of these treatments might save lives, it’s very reasonable to consider using those agents.

On the other hand, many providers said, “look,” and even the CDC’s recommendation is and the FDA’s recommendation is “please do not use hydroxychloroquine right now for COVID because there’s no evidence that it’s working now that we do have some evidence, and it’s just interrupting the supply chain for patients who are using it for diseases where we know it works.”18,19

There’s always going to be those folks who are hocks about evidence‑based medicine. My initial reactions to some of this evolving story was, “wait a second, let’s see some evidence here before we start singing these phrases, because we don’t really know how well it’s going to work.”

To the clinician who feels like they have no other choice, and they have a patient who is in critical condition, is rapidly deteriorating, there’s nothing else to offer, the studies show that giving systemic steroids was not a good option. Some studies implicating NSAIDS is not being the smart option, although that’s controversial.20

The feeling that, “well, my patient’s going to die… I can’t do nothing.” Others felt that, “but this may actually cause more harm than good. I would rather go with something...Wait until there is some evidence.” Many tried different approaches. I don’t think there’s a right or wrong here.

It’s easy for us to play Monday morning quarterback three months later, four months later as we look at the data coming out. It’s just, again, an important lesson that it always helps when we have the evidence because the conclusion may not always be what we think it is going into the exercise.

For those clinicians who they went with these drugs out of a sense of, in a sense, desperation because there was really nothing else to offer the patients, I think that they can feel comfortable with that decision. They can look back and say, “look, I did what I could best do for my patients at the time with the limited evidence.”

The question is, what do we do going forward? It’s a much more complicated discussion right now, because the initial enthusiasm has now been checked by some very weak or potentially even harmful data. We’re left with more support for those who say not to give, perhaps, than to give it. Of course, again, it’s left to the individual clinician’s discretion.

From an ethical standpoint, it is totally legitimate to reach for something in these kinds of urgent settings. Again, it’s just a fascinating evolution in the story when you watch evidence‑based medicines play itself out.

Melissa: That is really, really interesting. Thank you so much for answering that one and for talking to us today about hydroxychloroquine for COVID‑19.

Thank you for listening. Please leave any feedback such as comments or questions in the box below. We really appreciate anything you send us.   

References are listed at the top.

11 + 8 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.
Back to Top