Everyone is eager to find a cure for covid-19. Some people — notably President Trump, in his nightly press briefings on the pandemic — have touted the promise of hydroxychloroquine, an old malaria treatment. He’s suggested that patients should take it, asking, “What do you have to lose?”

Scientists in several countries have rushed to study the drug, with very mixed results. Many point to a recent study in France as evidence of its potential, but the quality of the study was low, with a small sample size (36 people) after questionable exclusion of certain participants from the data analysis. Other tiny studies have not identified positive effects. What we do know, though, is that this drug can have serious side effects, including heart arrhythmia, which can kill.

Even experienced scientists and clinicians tend to overestimate the promise of new drugs and minimize the risks. This is why randomized clinical trials are the gold standard in medicine, and why we use randomization even when testing treatments for patients who might be willing to try anything (those with late-stage cancers, say, or highly fatal infectious diseases like Ebola).

The question today is: How do we conduct such trials at pandemic speed? Officials, families of patients and the public are clamoring for results; researchers, even scrupulous ones, may find it tempting to jump the gun on announcing promising news. The answer is to stick with the gold-standard research methodology — no shortcuts — but look closely at each of the steps involved in running a trial, to see how to make them faster. The good news is that we can accelerate the process without compromising scientific standards. 

Particularly since most treatments for severe viral diseases are only mildly or moderately effective, it’s essential to follow a group of patients who are identical in every way, except one group gets the drug and the other doesn’t; ideally, neither the doctors nor the patients know who is in which group.

We have experience adhering to that level of rigor, even during frightening epidemics. During the 2014-2016 West African Ebola outbreak, an antibody-based treatment called ZMapp was tested in a randomized trial, involving 72 people, against the then-standard treatment, which included intravenous fluids to maintain hydration. Patients receiving ZMapp appeared somewhat more likely to survive, although the results were not strong. Then, in the 2018 Ebola outbreak in Congo, more drugs were added to another randomized trial, involving 681 patients. Thankfully, two of the drugs, REGN-EB3 and MAb114, each a combination of manufactured antibodies, showed stronger results than ZMapp. In contrast, the antiviral remdesivir did not appear to be effective, even though it had looked promising in the lab. By following the careful process of randomized studies, we now have at least two treatments that we feel confident improve survival and can be used for future Ebola patients.

In the covid-19 emergency, scientists are streamlining each step of the research endeavor. The first is to write the trial protocol explaining who is eligible for a study, how treatment will be provided, what data will be collected and how — and gain approval from ethics committees that protect patient safety. Rather than starting from scratch, investigators are adapting protocols from research on similar diseases, such as severe pneumonia. Partly because of work like this, the speed at which coronavirus treatment studies have been set in motion is astounding: Chinese researchers had begun more than 80 trials by mid-February, and more than 600 trials are underway worldwide: Investigators are exploring hydroxychloroquine and its close cousin chloroquine, HIV drugs, remdesivir (the drug that failed in the Ebola trials), anti-arthritis drugs, and more.

Once a trial is approved, researchers then enroll ill patients who could benefit from treatment and who consent to participate. Unlike in clinical trials for patients with cancer or heart disease — whose occurrence is by now predictable — a hospital’s ability to enroll covid-19 patients will depend heavily on the course of the epidemic. Since cases will occur in different areas at different times, it’s especially crucial  that hospitals work together to pool patient data. Such sharing — including from hospitals with only a handful of patients — allows us to quickly enroll the numbers needed to detect even small improvements in survival, and it means the trial can continue even if an outbreak is brought under control at a key research site.

This is the strategy being used by a National Institutes of Health trial, in which 40-plus U.S. sites are already participating, with more to come. Outside of the United States, the World Health Organization’s multicountry “Solidarity” trial will enroll patients in countries including Italy, Norway, Canada and South Africa, and many more plan to join.

Monitoring patient outcomes throughout a trial is an approach commonly used to produce rapid results. Once powerful evidence emerges of a drug’s effectiveness, researchers can stop the test early and declare the drug a winner. (It’s a balancing act, since early promising results could be a statistical fluke.) On the other hand, if a drug isn’t working at all or is unsafe, they can stop the trial early and move on to another candidate treatment. Ultimately, a committee of experts independent from the trial, tasked with overseeing it, makes the final call. If a drug is found to be effective, the results can be rapidly communicated to the public, and companies might start shifting toward production.

The final step is for regulatory agencies, such as the Food and Drug Administration, to review the evidence. “Emergency use authorizations” are specifically designed for expedited decisions such as those required during epidemics — although they must be granted cautiously. The FDA has issued an emergency use authorization for hydroxychloroquine and chloroquine to treat covid-19, but several former FDA leaders and other scientists have condemned that decision. “I understand the desire to find hope, but we need more evidence than is currently available before we encourage widespread use,” Margaret Hamburg, a former FDA commissioner and past president of the American Association for the Advancement of Science, told Science magazine. In contrast, the European equivalent of the FDA, the European Medicines Agency, allows the use of these drugs for covid-19 patients only in clinical trials and in other limited settings — a more evidence-based approach.

Though clinical trials may at times feel burdensome, they are crucial for protecting patient health and safety. Administering unproven drugs to severely ill people because they “have nothing to lose” does nothing to further our understanding of the best way to treat coronavirus patients. Fortunately, the pace of scientific research has never been faster than during this pandemic. Through innovation and collaboration, we are well positioned to identify the safe and effective treatments we need. But until such treatments emerge, we must remember the risks of new drugs, and not focus only on their promise.

Twitter: @nataliexdean