This is the relevant sentence:
"More detailed information about the trial results, including more comprehensive data, will be available in a forthcoming report."
This looks more like a press release and contains no details that would let one evaluate how robust these results are.
They also include changes to endpoint metrics with P values. While it's not the full dataset, it's something.
"Preliminary results indicate that patients who received remdesivir had a 31% faster time to recovery than those who received placebo (p<0.001). Specifically, the median time to recovery was 11 days for patients treated with remdesivir compared with 15 days for those who received placebo. Results also suggested a survival benefit, with a mortality rate of 8.0% for the group receiving remdesivir versus 11.6% for the placebo group (p=0.059)."
,,Preliminary results indicate that patients who received remdesivir had a 31% faster time to recovery than those who received placebo (p<0.001)."
This sentence doesn't make any sense. They should have at least published a 95% confidence interval in the difference or ratio between time to recovery.
The conclusion you'd draw is the same using a P value and a confidence interval. Having a statistically significant P value, meaning a P value less than a given significance level, then implies that the confidence interval at that significance level does not include the null hypothesis value.
If your significance level is .05 (meaning you are willing to accept a 5% chance of a type 1 error in which you reject the null hypothesis incorrectly), you can use either a confidence interval or a p value. You could look at a 95% condfidence interval and see if the % faster time to recovery range includes the null hypothesis of zero, or you could calculate a P value to see if it is less than .05. Both of them will give you the same conclusion.
Based on the P values given, you know that the 95% confidence interval for % faster time to recovery will not contain 0% (nor will the 99% interval or the 99.9% interval). But, the % difference in mortality at a 95% confidence interval will contain zero, since p>.05.
I understand caution and skepticism...but how on earth is that the “relevant sentence” out of the whole article?
I would vote for the first one:
“Hospitalized patients with advanced COVID-19 and lung involvement who received remdesivir recovered faster than similar patients who received placebo, according to a preliminary data analysis from a randomized, controlled trial involving 1063 patients.”
So we know it was randomized with a control group and sample size of over 1000 patients. That’s a first step towards reaching statistical significance.
The adverse effects are concerning, however, the lancet trial was underpowered and had to be stopped early because there weren't enough patients [0]. That's why they say in their interpretation that even though they saw some numerical improvement in their results, larger trials are needed to evaluate statistical significance. The NIAID's trial is also a placebo-controlled randomized controlled trial and was adequately powered for reduction in recovery time, but didn't hit statistical significance at a p<.05 threshold for mortality rate.
Hi, I read about lancet trial in news actually ... it is not "powered" ... not enough patients. they say the wuhan virus went away before trial finish.
Just want to remind everyone to be consistent. President Trump touted Remdisivir while it was not approved for treating C19 and had no clinical trials.
Trump was especially irresponsible in touting hydroxychloroquine because he said it had very promising results, people have access to it now, there are no adverse effects, and he was going to let everyone get a prescription for it.
You're also incorrect, remdesivir had already started clinical trials in early March and patients were being enrolled. Trump made his comments ~2 weeks after they had begun.
>Trump was especially irresponsible in touting hydroxychloroquine because he said it had very promising results, people have access to it now, there are no adverse effects, and he was going to let everyone get a prescription for it.
Rheumatologist with 30 years of experience who, while skeptical of hydroxychloroquine's benefits for COVID19, has no problem with Trump's "what have you got to lose?" statement: https://www.youtube.com/watch?v=hRCG9RtFJ_Y
I found this news exciting/significant because this is the first time we've seen positive news about remdesivir in the context of a placebo controlled, randomized, double blinded trial. Unlike prior trials that either didnt have a control arm, or trials like the Lancet trial that was also released today that were underpowered, this trial shows a statistically significant effect by remdesivir. For context, the Lancet trial showed a numerical effect, but couldn't recruit enough patients and was ultimately underpowered, so it wasnt able to show statistical significance.
It's exciting that there's finally at least one drug with at least one fully powered thousand plus patient placebo controlled trial with a statistically significant impact. It shows the disease is drug gable and treatable, and is not invulnerable, and is an exciting win for science.
Something else I'm excited about is the potential to use Remdesivir as a preventative. Because it works by inhibiting viral replication, it could be used as a preventative for high risk individuals until a vaccine is developed or the disease is eradicated.
15 comments
[ 2.6 ms ] story [ 57.5 ms ] threadThis looks more like a press release and contains no details that would let one evaluate how robust these results are.
"Preliminary results indicate that patients who received remdesivir had a 31% faster time to recovery than those who received placebo (p<0.001). Specifically, the median time to recovery was 11 days for patients treated with remdesivir compared with 15 days for those who received placebo. Results also suggested a survival benefit, with a mortality rate of 8.0% for the group receiving remdesivir versus 11.6% for the placebo group (p=0.059)."
This sentence doesn't make any sense. They should have at least published a 95% confidence interval in the difference or ratio between time to recovery.
If your significance level is .05 (meaning you are willing to accept a 5% chance of a type 1 error in which you reject the null hypothesis incorrectly), you can use either a confidence interval or a p value. You could look at a 95% condfidence interval and see if the % faster time to recovery range includes the null hypothesis of zero, or you could calculate a P value to see if it is less than .05. Both of them will give you the same conclusion.
Based on the P values given, you know that the 95% confidence interval for % faster time to recovery will not contain 0% (nor will the 99% interval or the 99.9% interval). But, the % difference in mortality at a 95% confidence interval will contain zero, since p>.05.
I would vote for the first one:
“Hospitalized patients with advanced COVID-19 and lung involvement who received remdesivir recovered faster than similar patients who received placebo, according to a preliminary data analysis from a randomized, controlled trial involving 1063 patients.”
So we know it was randomized with a control group and sample size of over 1000 patients. That’s a first step towards reaching statistical significance.
[0]https://www.thelancet.com/journals/lancet/article/PIIS0140-6...
You're also incorrect, remdesivir had already started clinical trials in early March and patients were being enrolled. Trump made his comments ~2 weeks after they had begun.
Rheumatologist with 30 years of experience who, while skeptical of hydroxychloroquine's benefits for COVID19, has no problem with Trump's "what have you got to lose?" statement: https://www.youtube.com/watch?v=hRCG9RtFJ_Y
It's exciting that there's finally at least one drug with at least one fully powered thousand plus patient placebo controlled trial with a statistically significant impact. It shows the disease is drug gable and treatable, and is not invulnerable, and is an exciting win for science.
Something else I'm excited about is the potential to use Remdesivir as a preventative. Because it works by inhibiting viral replication, it could be used as a preventative for high risk individuals until a vaccine is developed or the disease is eradicated.