🧵The TOGETHER trial purports to show that Ivermectin doesn't provide any benefit. It's given top billing in the NEJM and is now splashed everywhere on the medical news. It's a completely awful study that should not have passed review.
nejm.org/doi/pdf/10.105…
nejm.org/doi/pdf/10.105…
/2 The 1st and most damaging issue that the trial was conducted in an area where IVM use was already exceptionally high. They **DID NOT EXCLUDE** for prior IVM use. Only "known hypersensitivity" to tested drugs. Case closed. This trial cannot deliver interpretable results. 
/3a
The randomization of the TOGETHER trial was **NOT RANDOM.**
The trial protocol states: "Participants are randomly assigned with equal allocation (...) Allocation of participants to treatment arms is uniform across all concurrent interventions as well as placebo."
The randomization of the TOGETHER trial was **NOT RANDOM.**
The trial protocol states: "Participants are randomly assigned with equal allocation (...) Allocation of participants to treatment arms is uniform across all concurrent interventions as well as placebo."
/3b
Did the TOGETHER randomization process equally and randomly assign patients to various arms in equal measures at equal times?
No. Case closed. This trial didn't even follow its own protocols.
Did the TOGETHER randomization process equally and randomly assign patients to various arms in equal measures at equal times?
No. Case closed. This trial didn't even follow its own protocols.
/4
Finally, as an obvious nail in the coffin, if patients were randomly assigned, and the placebo was inert, and they had no reason to suspect they weren't getting the treatment they wished, there wouldn't be this ENORMOUS drop-out gap with only 40% adhering to placebo (WTF?):
Finally, as an obvious nail in the coffin, if patients were randomly assigned, and the placebo was inert, and they had no reason to suspect they weren't getting the treatment they wished, there wouldn't be this ENORMOUS drop-out gap with only 40% adhering to placebo (WTF?):
• • •
Missing some Tweet in this thread? You can try to
force a refresh
