(Total Views: 786)
Posted On: 07/01/2020 8:54:18 PM
Post# of 155798
NP's asking the FDA to defer the S/C interim analysis to include more patients is outstanding news. The jump in S/C enrollment to 120 was the biggest news today and presents a great opportunity.
Remember, statistical precision increases only with the square root of observations. So, compared to a 50 person study, 120 people is actually over halfway to the study's planned 390.
Put another way, if the width of the confidence interval for a 50 person study is "X," the width for a 120 person study is 65% of "X," and for a 390 person study is 35% of "X."
I can see the psychological argument for a 150 person study - 100 in the treatment arm - but past that it's diminishing returns. Based on run rates, that's before July 15th, implying last patient/last visit August 12th and reporting within two weeks thereafter.
In trade-off for about a six week delay, I think we not only get more certainty of statistically convincing results in this trial, but completion of a body of work on over 250 treated patients (75 EIND + 60 m/m trial + 100 s/c trial + 25 Mexico trial).
That body of work should make the FDA comfortable, and the American people confident in the FDA processes, in naming leronlimab as a standard of care for COVID-19.
Remember, statistical precision increases only with the square root of observations. So, compared to a 50 person study, 120 people is actually over halfway to the study's planned 390.
Put another way, if the width of the confidence interval for a 50 person study is "X," the width for a 120 person study is 65% of "X," and for a 390 person study is 35% of "X."
I can see the psychological argument for a 150 person study - 100 in the treatment arm - but past that it's diminishing returns. Based on run rates, that's before July 15th, implying last patient/last visit August 12th and reporting within two weeks thereafter.
In trade-off for about a six week delay, I think we not only get more certainty of statistically convincing results in this trial, but completion of a body of work on over 250 treated patients (75 EIND + 60 m/m trial + 100 s/c trial + 25 Mexico trial).
That body of work should make the FDA comfortable, and the American people confident in the FDA processes, in naming leronlimab as a standard of care for COVID-19.
(4)
(0)
