There's been a lot of focus on—and confusion about—the potential of hydroxychloroquine in the #COVID19 pandemic.
Though these are important studies and I'm glad they’ve been published, I'd like to see more rigorous studies on hydroxychloroquine from randomized, controlled trials.
One major potential limitation of observational studies involves "residual confounding"—despite efforts to correct for baseline differences, it’s possible sicker patients were more likely to get the medication, which could make it appear ineffective even if it is helpful.
Given the limitations of these studies, there is no definitive evidence you're more likely to die if you get hydroxychloroquine but I do think it makes it unlikely that it will be a dramatically effective treatment, as I told @greta. bit.ly/2XVRjuC
Another new study was published yesterday in @NEJM of 821 people exposed to the virus who were randomized to hydroxychloroquine or a placebo. nyti.ms/2AFu5Rm
Unlike the observational studies, it was randomized and double-blinded. It didn't show that hydroxychloroquine had a measurable effect in preventing symptomatic infection compared to placebo.
The NEJM study also has important limitations. Lack of testing capacity means that most patients with symptoms weren’t tested, and there was no way to see if the medication could reduce the rate of asymptomatic infection.
But the results suggest that hydroxychloroquine is unlikely to be dramatically effective when given as post-exposure prophylaxis.
No study published to date has looked at the possibility that hydroxychloroquine taken before exposure could reduce the risk of infection or of serious illness. That’s why it’s so important that rigorous, well-designed, controlled trials get done.
