Saturday, October 3, 2020

Hydroxychloriquinone

 10/2/20 Reuters:

Combining trial data suggests hydroxychloroquine benefit

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." Along with the meta-analysis, he told Reuters, "This is extremely strong evidence of benefit." (https://bit.ly/2SlHEeE)

Remember Trump was called names when he touted this early on.

1 comment:

Tom Bridgeland said...

My general rule is that if it requires meta-analysis to find an effect, that effect is probably so minor as to be nearly insignificant in the real world.

If comparing nearly equally effective treatments, meta-ananlysis has a place. How did your doctors decide which blood thinner to use? Meta-analysis eventually showed a slightly lower risk of adverse side effects from some of the newer blood thinners over the tried-and-true Coumadin. If you have good insurance, you might go with Eliquis or Brilinta etc. If you are poor, use Coumadin if that's what your doctor says. The differences are pretty slight.