When medical historians look back at the Covid-19 pandemic, they will reckon with how the United States, with its vast technological and scientific resources, stumbled so badly in the face of an emerging virus. They’ll wonder why the country responded so slowly, and why, in particular, it lacked adequate diagnostic tests for months after cases started to rise.

But they will also be baffled at something else: that it took so long to study new medicines to determine if they worked — even as the novel coronavirus, SARS-CoV-2, was killing 2,000 Americans a day.

Simply put, more people will die from Covid-19 because we cannot study drugs more quickly.



This does not mean that it would be better to provide people with unproven therapies — or that Covid-19 studies should be accelerated so fast that we draw the wrong conclusions or put people at risk. But our inability to start and run clinical trials faster — whether in normal times or in a pandemic — is a legacy of our decision not to develop the technologies and approaches that would make doing so easier.


The gold standard method of testing drugs is what is known as a randomized controlled trial, in which patients receive either an experimental drug researchers think will work or the standard care that is already available — a selection that is based, effectively, on a coin flip. Ideally, the trial is also double-blind, which means that patients who don’t get the drug get a placebo, and neither the patients nor their doctors know who got what.

Obviously, for the patients who receive a placebo, this can be a sacrifice. And it can be a hard one to make. 

But researchers have learned the hard way that this is often the only way to figure out if a treatment works. Randomized controlled trials (RCTs) have put a stop to routine mastectomies in breast cancer, the belief that common arrhythmia drugs saved lives, and a common knee surgery that was no better than physical therapy. A decade ago, Pfizer spent more than a billion dollars developing a heart drug that an RCT showed resulted in more patients dying.

So many experts see randomized controlled trials as critical at this time. 

“I just don’t understand why you would do a trial that’s not randomized,” said Ethan J. Weiss, an associate professor of cardiology at the University of California, San Francisco. “We have an incredible opportunity to do real trials in real endpoints in a record amount of time.”

But the process of designing and conducting studies takes time, and it’s even harder when the health care system is under siege from a surge of Covid-19 cases. 

Heroically, lots of studies are getting done in the face of Covid-19, and researchers are using innovative methods to do them, allowing studies to be modified quickly to include new treatments. The U.K.’s RECOVERY trial is testing treatments including hydroxychloroquine and a pair of HIV drugs. The most rigorous study of Gilead’s experimental Covid-19 drug, remdesivir, is a trial being conducted by the National Institute of Allergy and Infectious diseases. The World Health Organization is running a study that includes both remdesivir and hydroxychloroquine, among other drugs. The U.S. government’s Patient Centered Outcomes Research Institute is running a study to see if hydroxychloroquine prevents doctors from getting Covid-19.

But both remdesivir and chloroquine were identified as potentially active against SARS-CoV-2 in laboratory tests in February, and we still don’t know for sure if they work against the virus in people. Gilead should release some data on remdesivir this month in patients with severe disease, though that study compares two courses of remdesivir, not a control group. 

Part of the problem is technology. Electronic health record software in the U.S. is not set up to make clinical research faster and easier, Raj Mehta, a family medicine researcher at Florida Hospital Family Medicine Residency, wrote in an email. There is historical patient data, there are partnerships to use “big data” with machine learning and artificial intelligence. But all that offers little at the present time.

“We have billing claims as, absurdly, our only reliable and easily integratable national source of raw patient data,” Mehta wrote. “What we don’t have is anything useful to produce evidence-based medicine.

“The criticism may seem harsh, but if we could trade all the data silos, all the AI/ML efforts, & all the billing data, for a fully integrated, nationwide, RCT platform in EHRs, we would *all* do it in a heartbeat,” he said.

One reform packaged in a 2016 law that made changes to the Food and Drug Administration was “to push for randomized trials to answer questions quickly using electronic health records and claims data,” said Robert Califf, the former FDA commissioner who now heads clinical policy and strategy at Verily Life Sciences and Google Health. It’s still needed, he said.

“All you have to do is look at hydroxychloroquine, where you have all these observational studies that cannot possibly answer the question,” Califf said.

For years, there’s been talk about making the clinical trial process more standardized, and cheaper, so that the same rules would apply each time a study needed to be run. There’s even been discussion that what are known as pragmatic trials — large, simple, randomized studies in which less data are collected — might be conducted using electronic health records. But that hasn’t happened at the pace it should.

The reason involves another part of the problem. Clinical trials are principally run by drug and medical device companies in order to obtain regulatory approvals, with public health authorities only picking up the slack in rare examples. But the result is that we have not built a system that would make studies simpler; most patients have little opportunity to participate in research; and we are too slow to figure out what works.

What would the system look like if we fixed it? It would make it easier to study drugs for heart disease, where studies are so large and expensive that many companies don’t test their medicines. It would ease studies for rare cancers, which are currently problematic because the right patients are hard to find. And it could create a medical information superhighway that would power health care through the next century.

It would certainly help with the next pandemic.

  • Matthew it is about time someone wrote about how incompetent medical science is. All you ever hear from the director of the FDA and Dr Fasuci is that it will take at least a year for a vaccine. Both of them say that everyday !!!. The American People want and need effective treatments NOW !!!!!.

  • Matthew, do you even realize what you cause with your misleading and catchy headline? You are blaming those who work under extreme circumstances in clinical research to contribute finding a solution for COVID. YES, I mean it- clinical research is NOT the cause for people dying, it is the SOLUTION!!!
    ‘Most patients have little opportunity to participate in clinical trials’ …what??? Do you know that 80% of clinical trials (in non-pandemic times) are delayed or fail due to the lack of volunteers needed for those trials? So, it is actually the other way around.
    ‘Simply put’ – you do not help with your article, you mislead. Just citing well renown capacities in the medical field doesn’t do the complexity of the topic nor the urgency of OUR fight as researchers justice – BTW, with the same struggles- barely providing enough PPE for our amazing research staff and our hero volunteers.
    I presume, that you are part of a lot of clinical research volunteer databases already, Matthew? It is time for ALL of us to contribute by volunteering and/or being part of the solution not the mongering.

  • “Ideally, the trial is also double-blind, which means that patients who don’t get the drug get a placebo, and neither the patients nor their doctors know who got what.” – All of us know that placebo has cured all diseases including stage four cancers. But none of us has cared to find out why and how it cures. So we are afraid of finding the truth.

    “There is historical patient data, there are partnerships to use “big data” with machine learning and artificial intelligence. But all that offers little at the present time.” – Did data prove anything at anytime? Data is false, meaningless, and cannot be useful. Billions of people knew that earth is the center of the universe. So much data from so many billions but were all proven wrong when Galileo showed a single proof with a single observation. We got scared and jailed him.

    Similarly, take as much data as you want, but you will never find that reincarnation is happening in our world or we are not our bodies, but we are souls. Data is based on real numbers. Real numbers are false because it is not an object of nature. Real numbers are points on a straight line, but there is no straight line in the universe, because all of objects in the universe are moving. Thus entire mathematics must be false. Data cannot help you to find truth. The process for finding the truth is somewhere else.

  • This article seems to fit the format that it’s all easily solved, that clinical trials are just too expensive and burdened by red tape because of stupid bureaucrats. What balderdash.

    And I don’t understand the use of one quote. Author implies that Dr. Carliff agrees with him. But I read it the other way around. That “Big Data” on a therapeutic agent, is hopelessly conflicted.

    Califf said. “All you have to do is look at hydroxychloroquine, where you have all these observational studies that cannot possibly answer the question”.

    And studies, observational?

  • Title: Ignorance: It’s not just in the Oval Office.

    Most researcher-MDs are ignorant. Hear me out! They believe “high quality trial” and “RCT” are synonymous. This core belief is actually unexamined canon and has dogma and blind faith as its foundations. It is FAR from true. They are merely partly overlapping sets. I’ll explain after noting briefly:

    Rolf is right; Bh is ignorant. But that could be no one’s fault. Bh is to be urged to get an education (or failing that, ignored). Just using “stuff”, without carefully studying outcomes, is a terrible idea; see the examples in the article for the results re. mastectomy, arrhythmia, knee surgery, Pfizer’s torcetrapib. A tragic current example is the massive, mostly unorganized and unstudied “compassionate” overuse of hydroxychloroquine which is already proving to have been a terrible idea.

    But Steve is NOT wrong!

    Steve, I think you meant to say:
    We make 1,000 doses of each of 10 promising vaccines. We divide 10,000 volunteers into 10 (undisclosed) groups of 1,000 people. Everyone in each group receives a dose of the vaccine assigned to that group. We see how the groups do. (I’ll add: Ideally the volunteers are in one of the groups at very high risk of death from COVID-19 and are paid and granted priority access to best available care if they fall ill.) This IS an ethical, relatively efficient, high quality trial that should start yesterday. (And be improved by an 11th group of 1,000 all getting a placebo.)

    I have discovered what I think is yet another hugely significant trial design improvement, but the math makes it hard to explain in an understandable way to most people. I need to work on it with a great clinical trial statistician and simulations to validate it and refine how to explain it to PIs, etc.

    PS: Historically, Kaiser has done well leveraging their EHRs to find good research data. Hopefully we’ll be seeing much more from them soon.

    • The human research has evolved tremendously over the past 100 years for the better of humanity. We should be careful not to repeat our mistakes from the past.
      The trial design, you describe is not new, it is called prospective randomized trial and medical communities are well aware and have performed numerous such trials. Even if you have perfect situation, starting a trial is not easy. Just to get approval, you have to right several pages to justify why, explain the risks benefits, send to a committee to reviewed and critique it. Get some funding, it is team effort and one person can not do it.
      As of today, we do not have vaccine, so you can not start a trial yesterday. Second this type of trial design does not answer the question if drug A is better than drug B, it is mostly for studying association of risk factors, like smoker vs non-smokers. Not to be skeptical but just say if 9K volunteers out of 10K who received the vaccine die, due to side effects of the vaccine then what?
      I do not agree we should bend the research rules for COVID19, as it will open the door for inhuman research in the future. None of the current electronic health records are meant for research, these are means to collect demographic and billing data.

  • Unfortunately, the American health system revolves around MONEY!
    Drug companies, insurance companies, hospitals etc.

  • I think the medical profession has a hard time breaking out of their current mode of thinking. They still want to treat covid like any other disease. They have to have their double-blind tests and anything else is anecdotal. Well when you have 2.6 million cases. You have enough statistical data to know if a drug is making a difference. Yet they insist they need to give some poor guinea pig patient a placebo just to see if he might die as part of their test. I can see this in an experimental drug but in drugs that have been around for 70 years you have all of the safety data you need. If there are drugs out there that might work and have been around for a long time and we know they are safe, then just use them. You have not come up with anything else and with the rate you are all moving this whole thing will be years in the past and the dead will already be buried before you find a treatment.

    • It’s clear that you have absolutely no understanding of any of the topics on which you’re speaking.

    • Rolf – Last week, I made a comment about the need for very widespread experimentation in both therapy and vaccines for this disease. Someone was very insulting towards me.
      Let me suggest some possible scenarios for you to see if they are flawed.

      Suppose we took the ten best vaccine candidates, and, instead of doing Phase 1, 2, 3 trials, we paid 10,000 people to get 1,000 of each vaccine, and then after a month, all get inoculated with the virus.

      Suppose 9 of what looked like the best candidates are not so good. They don’t work, and on top of that, some of the infected people are far worse off than if they had no vaccine -and, of course, many of them, most of them, never would have gotten sick anyway.
      But, for the sacrifice of those 9,000 people we get one good vaccine, and it works, and we get it six months sooner – or one month.

      Let’s supposed, with the virus still around, indefinitely, for the next two years, as a more or less steady rate, you have in the US – forget the rest of the world, which will of course also get the vaccine sooner – but just in the US, 1,000 people are dying every day, another 1,000 live but with damaged hearts, lungs, kidneys, and sometimes brains. They are pretty messed up in other words. On top of the 1,000 dying and 1,000 being left partial or complete invalids, you have at any one time at least many thousands who are too sick for work, for now, but they are small potatoes.
      Because the country can not go back to anything resembling normal, we are also at very high unemployment and losing our wealth daily, so the economic cost is another, say $1B per day.

      If skipping the Phase 1 and greatly shortening the Phase 2 trials savers you two months, 60 days, then the 9,000 people you sacrificed saved you 60,000 people and $60B.

      Of course I made these numbers up, but they are actually all very, very conservative. We have over 2,000 deaths per day now, and despite the screams to open the country, we can be nearly certain of that increasing if we do. The loss of $1B per day is a really minor recession, no one thinks it will be minor at this point.
      Any rational view of this tells you taking risks, hell, taking HUGE risks, to get vaccines and therapies sooner is the right thing to do, but the medical community can not adjust it’s thinking. How is the original poster wrong?

  • I am not sure why everyone is after remdesivir. It does not work. The data from compassionate use of remdesivir shows the 84% improvement and mortality of about 13% , which expected in COVID19 patients without any drug therapy, just supportive measure. On top of it, there were 23% adverse events from the drug itself. The duration to death was also as expected 9-15 days. So I don’t know why every one is asking for it? From pathological perspective, this virus does not cause wide spread infection in the blood, viremia. So in my opinion, any drug therapy that is aimed to kill this virus via blood stream concentration, will unlikely to work.

    • Remdesivir has been shown to inhibit Covid in-vitro; moreover its an RNA Polymerase inhibitor, so its a direct antiviral. Is this a silver bullet, no, but it should have some effect, perhaps in patients who have mild to moderate disease. The NEJM report you cite surprised many because these were the sickest group of pts but it was not a prospective randomized controlled trial, so results must be interpreted with caution; The body of evidence to date is largely observational, with (until today’s underpowered China trial) showing a positive signal. By end of May, we will have a more complete body of evidence that will be more definitive.

  • What I don’t understand is why these RCTs take so long to publish, especially for drugs that have already been trialed and found safe. Most of the remdesivir trials are for 10 days or less, why does it then take months to see the results, especially in times like these?

  • What I don’t understand is why these RCTs take so long to publish, especially for drugs that have already been trialed and found safe. Most of the remdesivir trials are for 10 days or less, why does it then take months to see the results, especially in times like these?

Comments are closed.

Sign up to receive a free weekly opinions recap from our community of experts.
Privacy Policy