Hitting the primary endpoint led to the interruption of recruitment. However, it is a clinically weak composite endpoint of ER observation (6 hours) or hospitalization.
Näytä tämä ketju
A draft Fluvoxamine TOGETHER preprint is out. The trial was stopped too early, is broadly inconclusive and frustratingly underpowered.https://www.medrxiv.org/content/10.1101/2021.08.19.21262323v1 …
-
-
-
This is a regrettable departure from the Covid associated hospitalization or death, harder, primary endpoint used by TOGETHER in the previous Hydroxychloroquine/Kaletra trials.
Näytä tämä ketju -
Composite endpoints are generally bad, confusing causal matters. They are terrible when they focus on arbitrary and soft endpoints like "ER observation". What causes "ER observation"? The patient objective clinical condition? His personal level of panic? Triage saturation?
Näytä tämä ketju -
Expect nonsense and that seems to be what you get when you deconstruct the composite in that study: Fluvoxamine reduces massively ER visits.. ...but has no statistically significant impact on hospitalizations. Sigh...pic.twitter.com/6NVyK2UpK1
Näytä tämä ketju -
Can there be a rational explanation for patients on FLV to clog up ER 5x less than the placebo arm? Fluvoxamine reduces anxiety. Are patients (typically 6 days into their treatment) less prone to panic and unnecessary ER rushes? Are they calmer and trigger less observation?pic.twitter.com/iisCFyktqw
Näytä tämä ketju -
That being said, there is an objective benefit to less ER visits, even if they are psychological in motivation: less crowding of the medical service, less stress on the medical staff and less money spent.
Näytä tämä ketju -
Moving now to the hospitalized patients. The 21% Fluvoxamine reduction points to efficacy but lack statistical power (p=.17). What a dreadful decision to stop the trial at such an inconclusive stage.
Näytä tämä ketju -
All the hospitalization secondary endpoints miss statistical significance: • Length of stay • Days on mechanical ventilation • Death, with a trend to benefit (2.3% vs 3.3%, p=.26)
Näytä tämä ketju -
-
Underpowering means the primary endpoint forest plot is basically useless.pic.twitter.com/p9a0YKJluT
Näytä tämä ketju -
Combining this study with the previous ones on the hospitalization endpoint leaves us with still very promising but inconclusive trial evidence. More data is needed.pic.twitter.com/24i8itn2aH
Näytä tämä ketju -
If the Fluvoxamine mechanism of action is antiviral, intervening only 9 days into the disease on average (5 incubation + 4 symptomatic) might squander most of the therapeutic benefit. The trial protocol might be the source of the limited efficacy, rather than the drug.
Näytä tämä ketju -
The Brazilian P.1 variant might be another unknown, detrimental factor. Finally, treatment adherence was very poor (73%) and could explain a lot. The only Per Protocol primary number is much stronger: 0.34, CI 0.20-0.54. Would be great to have that for the secondary endpoints.
Näytä tämä ketju -
In conclusion: an underpowered mixed bag, pointing to efficacy/upside only on use. Another likely illustration of how critical promptness of treatment is in this viral disease, like in any other viral disease. A medical fact incomprehensibly suppressed during the pandemic.
Näytä tämä ketju
Keskustelun loppu
Uusi keskustelu -
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.