CDiddy,
Firs of all, this is a P2 trial.
I think they are taking it step by step. We see if it works -> It does -> We design quickly a P3 with fine-tuned parameters -> We do a P3 with much enhanced possibility of success -> We get approved.
Now, if we blow it out of the water with P2 we request EUA for LH treatment.
We have 100M affected by COVID-19, 30% might be LH as per Dr. Osterholm interview this morning: "this is going to transform medical practice."
Most insidious effect is brain-fog.
CYDY will be a player here.
OK. Now to Power of trial, which is not that important in P2 but nevertheless ... why did they lower the number of patients ??
To finish it ASAP and go on to P3. There will not be shortage of patients !!!. This is a different ball game.
Now, let's get technical. Can we do a P2 with adequate power with 50 patients ??
Yes, provided we have a very high effectivity. In this kind of trials there is not a dichotomous results (like live and death). Here we have a continuous response, for example in a scale to 1 to 10. All the patients result in a number and all the numbers are made to conform to a probability distribution.
Then one can make analysis on p-values and/or Power.
So, what is important is the difference of the averages of the measurements between the SOC and VX, how much the measurements scatter around the mean (called standard deviation) and what kind of reduction is measured between the groups (reduction on the average/mean).
For example, lets assume we will measure the patients symptoms in a scale from 0 to 10 and, at the end of the trial we expect that SOC mean is 6 with 1.4 standard deviations scatter. Also, we expect a 20% improvement with VX (4.8 mean).
How many patients do we need for a power of 90% ?? We need 58 patients.
This is just an example, I have no idea of what assumptions CYDY made. For the sake of having some fun, let assume we are expecting 15% improvement (5.1) mean. We would need 102 patients.
So, not easy to know why 50 patients as we HAVE to make assumptions on SOC results, VX results and standard deviations (data scattering).
My point is: I believe we are trying to do a quick-and-dirty trial, get data and then launch a bona-fide P3. In the mean time ascertain if VX works spectacularly or not and apply for EUA if this is the case.
Let's get started !!!!
Firs of all, this is a P2 trial.
I think they are taking it step by step. We see if it works -> It does -> We design quickly a P3 with fine-tuned parameters -> We do a P3 with much enhanced possibility of success -> We get approved.
Now, if we blow it out of the water with P2 we request EUA for LH treatment.
We have 100M affected by COVID-19, 30% might be LH as per Dr. Osterholm interview this morning: "this is going to transform medical practice."
Most insidious effect is brain-fog.
CYDY will be a player here.
OK. Now to Power of trial, which is not that important in P2 but nevertheless ... why did they lower the number of patients ??
To finish it ASAP and go on to P3. There will not be shortage of patients !!!. This is a different ball game.
Now, let's get technical. Can we do a P2 with adequate power with 50 patients ??
Yes, provided we have a very high effectivity. In this kind of trials there is not a dichotomous results (like live and death). Here we have a continuous response, for example in a scale to 1 to 10. All the patients result in a number and all the numbers are made to conform to a probability distribution.
Then one can make analysis on p-values and/or Power.
So, what is important is the difference of the averages of the measurements between the SOC and VX, how much the measurements scatter around the mean (called standard deviation) and what kind of reduction is measured between the groups (reduction on the average/mean).
For example, lets assume we will measure the patients symptoms in a scale from 0 to 10 and, at the end of the trial we expect that SOC mean is 6 with 1.4 standard deviations scatter. Also, we expect a 20% improvement with VX (4.8 mean).
How many patients do we need for a power of 90% ?? We need 58 patients.
This is just an example, I have no idea of what assumptions CYDY made. For the sake of having some fun, let assume we are expecting 15% improvement (5.1) mean. We would need 102 patients.
So, not easy to know why 50 patients as we HAVE to make assumptions on SOC results, VX results and standard deviations (data scattering).
My point is: I believe we are trying to do a quick-and-dirty trial, get data and then launch a bona-fide P3. In the mean time ascertain if VX works spectacularly or not and apply for EUA if this is the case.
Let's get started !!!!
13
