In a series of randomized controlled trials, the malaria drug hydroxychloroquine did not show a statistically significant impact on the prevention or treatment of COVID-19. But when data from five of those trials were combined, researchers found that early use of the drug by people who were not hospitalized yielded a statistically significant 24% reduction in risk of infection, hospitalization or death.
“The meta-analysis pools together the studies and increases statistical power,” said Dr. Joseph Ladapo of the David Geffen School of Medicine at UCLA, coauthor of a report posted on Wednesday on medRxiv ahead of peer review. But a weakness of the meta-analysis, Ladapo acknowledged, is that infections, hospitalizations and deaths were grouped together into a “composite outcome.” Combining all those events into one big number makes it more likely researchers will find that treatment had a significant effect. Coauthor Dr. Harvey Risch of the Yale School of Public Health noted that seven nonrandomized controlled trials have also shown “statistically significant reduced risks with early outpatient use of hydroxychloroquine.”