(Total Views: 714)
Posted On: 01/27/2021 12:29:26 AM
Post# of 154036
Let's be clear about how little data is required for the significance calculation:
1) The 7-point ordinal scale (that is, a number from 1 to 7) is recorded at days 0, 14 and 28. Possibly at day 42. As death is 1 on the ordinal scale, it's not even a separate entry. That's 4 single-digit entries per patient.
2) The SOFA score, which is computed from 7 components, is recorded on days 0 and 14. Fourteen more entries.
3) We can't forget the variable saying whether they got leronlimab or not.
That's it, a grand total of 19 entries per patient is all that's needed for the significance calculation.
Now, obviously, there is an entire file about AEs and SAEs, with sufficiently detailed data to determine attribution. And we'll want to know what other treatments they received when for analysis.
But if they are chasing down "tens of thousands of data points," a very small fraction of those go to computing statistical significance.
1) The 7-point ordinal scale (that is, a number from 1 to 7) is recorded at days 0, 14 and 28. Possibly at day 42. As death is 1 on the ordinal scale, it's not even a separate entry. That's 4 single-digit entries per patient.
2) The SOFA score, which is computed from 7 components, is recorded on days 0 and 14. Fourteen more entries.
3) We can't forget the variable saying whether they got leronlimab or not.
That's it, a grand total of 19 entries per patient is all that's needed for the significance calculation.
Now, obviously, there is an entire file about AEs and SAEs, with sufficiently detailed data to determine attribution. And we'll want to know what other treatments they received when for analysis.
But if they are chasing down "tens of thousands of data points," a very small fraction of those go to computing statistical significance.
(7)
(0)
