Same comment I posted on the other thread relevant here:
Very promising early results. The caveats for non scientists to be aware of is that these are not randomized trials with a placebo and all - this is just through a mechanism called "compassionate use" that allows Gilead to give people the drug. Also, 68% of patients improved, the title implies that patients say a 68% improvement in their symptoms which is not the case, the 68% is the number of patients improved divided by total number of patients. To get a good picture of how much the drug improves outcomes compared to not having the drug administered, we need large placebo controlled randomized trials so we can get the statistical signal to see if it works and how well it works. As the paper states, those trials are underway.
> More than half of patients on mechanical ventilation were extubated (57 percent, n=17/30) and nearly half of all patients (47 percent, n=25/53) were discharged from the hospital following treatment with remdesivir.
Death rate for patients on ventilators has been something like 80 percent.
It's a nice signal, but yeah, theres a reason they're running a large randomized controlled trial right now for FDA approval - they arent likely to get approval just based on this.
"Results Of the 15 757 patients admitted, a total of 5183 (33%) received mechanical ventilation for a mean (SD) duration of 5.9 (7.2) days. The mean (SD) length of stay in the intensive care unit was 11.2 (13.7) days. Overall mortality rate in the intensive care unit was 30.7% (1590 patients) for the entire population, 52% (120) in patients who received ventilation because of acute respiratory distress syndrome, and 22% (115) in patients who received ventilation for an exacerbation of chronic obstructive pulmonary disease. Survival of unselected patients receiving mechanical ventilation for more than 12 hours was 69%. The main conditions independently associated with increased mortality were (1) factors present at the start of mechanical ventilation (odds ratio [OR], 2.98; 95% confidence interval [CI], 2.44-3.63; P<.001 for coma), (2) factors related to patient management (OR, 3.67; 95% CI, 2.02-6.66; P<.001 for plateau airway pressure >35 cm H2O), and (3) developments occurring over the course of mechanical ventilation (OR, 8.71; 95% CI, 5.44-13.94; P<.001 for ratio of PaO2 to fraction of inspired oxygen <100).
"Conclusion Survival among mechanically ventilated patients depends not only on the factors present at the start of mechanical ventilation, but also on the development of complications and patient management in the intensive care unit."
Yes, I think both NYC and Louisiana were seeing death rates above 80% for intubated patients, which is why they are starting to think they need to change the protocol.
In the remdesivir study, death rates on invasive ventilator patients was 13% which is much lower than expected at this point.
They could have selected withing ventilated patients for patients more likely to survive if they thought compassionate use might be most effective in patients that have a shot. They're running a large controlled trial so we should know more soon enough
It’s nowhere near as difficult to make as the article may imply, though. Yes, that first-generation route is quite a doozy, involving some harsh conditions (multiple steps at -78° C, which is actually not as difficult to carry out at the bench as it sounds) and nasty reagents, especially n-BuLi, which explodes into flames upon exposure to air. But it’s fairly easy to carry out on a small scale in a typical med chem lab. If you’re a medicinal chemist and need to make limited amounts, say, maybe in the hundreds of milligrams—which would probably be far more than enough to do preclinical studies—then that route will do. But it definitely can’t be scaled up easily to multi-thousand gallon chemical reactors, although like the article says, that’s where the process chemists come in. The second-generation route looks much better, and I’m sure it’s already being optimized even further.
It should be illegal to run a trial without a control group... For all we know, this trial was run in a good hospital and this drug actually hindered recovery rates. We just don't know.
You're literally throwing away human lives without learning anything that can be used to help other humans.
There are a lot of times that it isn't feasible, and they are running a large trial with a control group right now. The drug has a reasonable chance of working so people requested compassionate use, so the company figured they might as well track the compassionate use patients to get some data out of them. It's not like they're using this as a basis for getting FDA approval.
The primary endpoint of the trial of moderate COVID-19 patients originally looked at the proportion of patients discharged by day 14. The trial of severe patients was assessing the normalization of fever and oxygen saturation through day 14.
Now, Gilead has changed all of those details. The trial of patients with severe COVID-19 has swelled in size from 400 subjects to 2,400 participants. Gilead dialed the enrollment target in the clinical trial of moderate COVID-19 patients up from 600 to 1,600.
The trials have new primary endpoints, too. In each case, Gilead has changed the primary endpoint to focus on the odds of improvement on a seven-point scale that runs from death to not hospitalized.
As my advisor always used to say. If you didn't include a control, you didn't do an experiment. It's that simple. THe results are promising, but you have to keep in mind what that really means. It means, we haven't proven that it doesn't work. SO there is a chance that it works and we don't know how big that chance is. It also means that we haven't proven that it causes more harm than good. So there is a chance that it does indeed do more harm than good.
They also mean "advertising opportunity in the face of ignorance."
Due to general systemic issues of incentives and disincentives to research (especially tied to profit driven entities but its everywhere) that are growing ever more problematic, successes are often painted more vibrant and failures toned down, usually through loose/generalized language in summaries or by adding so many quantifiable qualifiers to the situation/experiment at hand, the results become nearly useless.
Research is not incentivized to give unbiased truth of failures the way science should be. They focus on optimistic results and try to gloss over negative results (have to pay rent and keep eating).
The truth usually requires reading through a full paper because while forming narratives is commonplace, pure academic dishonesty is not. Research results typically do not forge or falsify results.
Even in a humanitarian perfect situation where Gilead is pushing the supply of this out for free, I can assure you managers and purse holders want to see positive results. If you're a research scientist working at Gilead, I highly suspect you're going to be encouraged to show positive optimistic results, not problems/shortcomings (as in most business environments). If you do show negative results, they won't see the light of day if they're even read by management.
It's not like you can go buy the drug on the open market. The fda are the gatekeepers, and as gilead states in the press release, they are currently running a large phase 3 study with a control group etc to confirm these results
False equivocation. Remdesivir has been actively going through the drug approval process. P2 results for remdesivir indicate that side effects include liver issues. Contrast that with chloroquine which is nowhere near as studied (in context of covid) and is significantly more dangerous. There's a reason why the WHO put remdesivir as the most promising drug.
They usually have predetermined readout dates since p values can fluctuate throught data collection and prospectively defining those dates prevents you from being able to just pick the best time for a readout and skew results
Somewhat counterintuitively to frequentists, this is not a problem if they did Bayesian analysis. Many new trails for COVID19 therapeutics are actually reporting realtime Bayesian estimates of efficacy.
This press release has no medical significance. It is borderline irresponsible for Gilead to release such information, albeit they seem to have carefully qualified everything they say so as to avoid liability.
Irresponsible to who? This doesnt get the drug approved. They're running large phase 3s and all, which is what let's them put the drug on the market. In the mean time compassionate use data is useful to know.
Very promising early results. The caveats for non scientists to be aware of is that these are not randomized trials with a placebo and all - this is just through a mechanism called "compassionate use" that allows Gilead to give people the drug. Also, 68% of patients improved, the title implies that patients say a 68% improvement in their symptoms which is not the case, the 68% is the number of patients improved divided by total number of patients. To get a good picture of how much the drug improves outcomes compared to not having the drug administered, we need large placebo controlled randomized trials so we can get the statistical signal to see if it works and how well it works. As the paper states, those trials are underway.