There’s yet another new study out this week on hydroxychloroquine, the anti-malaria drug championed by President Trump (without evidence) as a cure for COVID-19. Bottom line: the hospitalized patients who took it didn’t seem to be any better off than patients who didn’t take the drug. Some patients had to stop taking the drug because of cardiac side effects.
The report isn’t published yet, and hasn’t gone through the standard review process, but doctors are starting to sort through its findings. It’s not a good sign for the drug. But this single study is not a killer blow, either.
That’s because a lone study or single clinical trial rarely offers incontrovertible evidence to disprove or bolster a claim — especially if, like this one, the study is small. Instead, it takes the buildup of data from multiple large trials and studies to guide medical decision-making. Everyone is desperate for answers around COVID-19, so there are more eyes on each new data point added to the pile. Watching that process in real time shows how messy science can be.
This particular hydroxychloroquine study included 181 people, which is small enough that scientists would hesitate to draw large conclusions from its results. It was also conducted with people who were already sick enough to be hospitalized. Other studies of the drug, which reached similarly dismal conclusions about how well it worked, were also done in hospitalized patients. But still other ongoing studies are testing how well it works in people who aren’t as sick, and whether it may prevent people who haven’t caught the virus yet from developing the most severe symptoms associated with COVID-19. With viral infections, earlier treatment tends to be better — that’s why people have to take Tamiflu right when they start feeling sick in order for it to work to treat the flu, for example.
Scientific research doesn’t usually provide yes-or-no answers. Instead, each new bit of evidence tilts the balance in one direction or another. While that’s happening, doctors make preliminary choices based on where they see that balance moving — as new data continues to be reported on hydroxychloroquine, some are deciding not to use it, while others may keep trying. They won’t be able to reasonably say they know for sure that it will or won’t work. It’s still a mostly open question, and decisions on care will still be made patient by patient for now.
Eventually, the balance of evidence will settle into a conclusion that experts have more confidence in. That might happen after researchers compile all the data from multiple small studies and analyze it as a group in a meta-analysis — that’s what might happen with the handful of small hydroxychloroquine studies that have been published already.
Even better, we might get more decisive data from larger studies. The World Health Organization’s Solidarity trial, for example, is testing multiple drugs (including hydroxychloroquine) in dozens of countries. Another study on the anti-viral remdesivir, which aims to include hundreds of patients, could also provide a clearer signal. Those types of studies take more time and resources, but they produce more conclusive results.
Patients are sick and dying from COVID-19 now, and there’s tremendous pressure to test drugs that might help them. In the United States, though, there has been limited coordination between groups running trials. It’s harder to understand where the balance of evidence lies when the research done is fragmented and disconnected, and when every researcher is asking a slightly different question.
“It’s a cacophony; it’s not an orchestra. There’s no conductor,” Derek Angus, chair of the department of critical care medicine at University of Pittsburgh School of Medicine, told The Washington Post. Francis Collins, director of the National Institutes of Health (NIH), said that a coordination plan was in the works.
Scientific research is hard under the best of circumstances. Finding answers that doctors and scientists trust is a long and frustrating process. It’s many times harder during an active public health emergency, when treating patients is a priority. The constant release of new information can feel like whiplash, as something that seemed like a solution one week turns out to be less useful the next. Reframing each bit of new data as piece of the puzzle, rather than an answer on its own, can help make sense of the flood.