Replies

Not getting into whether it works or not but that Frenchie hospital’s website is being deliberately misleading with statistics. Not surprising.

At the bottom right of the page you’ll see two numbers. One is the total number of patients treated with the combination at that hospital which is 1291 at the moment. This number includes people treated for 1, 2, 3 days or more.

Below this is the number of patients who were treated with this combination for more than 3 days who died, ie, 4 days or greater. He’s not counting anyone who was treated with this combination with 1, 2, or 3 days who died.

He’s making comparisons between two numbers he should not be.

There’s clearly a reason for this. He’s including everyone in the first group but the second group is the people who are already more likely to survive since they survived in the hospital for 4 or more days. There are probably a lot of people in the 1, 2, and 3 days of treatment who died since they wouldn’t be admitted unless their situation was very serious anyway. This hospital needs to make honest comparisons considering people are holding out a lot of hope for this. This doctor has played similar games before.

Since he’s playing games with statistics, I also wonder whether he’s playing games with who gets the drug in the first place. If he doesn’t include the most serious patients who are likely to go on a ventilator and not survive, he can make himself look even better.

There are probably a lot of people in the 1, 2, and 3 days of treatment who died since they wouldn’t be admitted unless their situation was very serious anyway.

Hmmm. That's worth considering. I haven't found anywhere that this was stated either way, but now I know to look out for it.

What you're saying is that patient X arrives and is treated with the combo. Patient X dies that day, or the next, or the next. Patient X is therefore not counted in this particular statistic as a death because they didn't receive the full course of treatment. Do I have that right?

I'll keep an eye open for that from here out.

Common sense does tell me that the more critical you are when a drug is administered, the less chance it has of working. You can't reasonably expect somebody who is several hours away from death has near as much of a chance that any pill is going to work.

That's why I've sort of bought into the "prescribe immediately after positive confirmation even if symptoms are mild" argument.

Thanks for providing more to consider.

He's not playing games as far as I can tell. He is stating up front his tracking parameters and showing both scenarios right next to each other. How is showing the two data cohorts next to each other "playing games"?

The average length of time in the ICU is 12 days for COVID-19. The average length of time in the ICU for a patient who ends up dying due to COVID-19 is 28 days. So, yes, he is saying that if he gets a patient so far gone that he can't make it 3 days on this treatment (e.g. on the 24th day of stay in ICU), he is not going to count him in his efficacy cohort as having undergone the treatment. That's perfectly fine so long if that is clear up-front, which it is.

Further, this perfectly coincides with his samples he has shown in his previous studies where the hydroxychloroquine and azythromycin does not really begin to do all that better in reducing the virus markers compared to the control group until day 3 where it really begins to take effect.

I don't mind that you post a caveat. What bugs me is that you compose it like he was trying to hide something when he posted it in plain view. And to top it off, you act like you are in the know when you really don't know the basic facts.