Interesting post from Yahoo board. I don't do sta
Post# of 153822
(Total Views: 1094)
Posted On: 01/22/2021 4:31:28 PM
Interesting post from Yahoo board. I don't do stats and went to law school because there was no math, but this all seems good to me. Posted originally by Ross on Yahoo.
I would like to share some statistics with you that have a bearing on the likely outcome of the CD12 trial. I am a doctor who is part of a large hospital network. In our state right now, the most recent statistics show that 67% of the patients in the ICU are on mechanical ventilation. That would place them in the critical category for the CD12 trial. This is an important statistic because we know that the mortality rate in this population was reported to be as high as 80% in the beginning of the pandemic and is probably still at least 35% or higher. If we assume that most or all of patients enrolled in CD12 came from the ICU you can apply the following math:
394 patients in trial
67% critical and 33% severe
394 x 0.667 = 262 critical patients
394 x .333 = 132 severe patients
Now, conservatively assume 40% mortality rate in critical and 15% in severe
262 Critical x 40% mortality rate = 105 expected deaths
132 Severe x 15% expected mortality rate = 20 expected deaths
Total deaths should have been 105 + 20 = 125 for a 31.7% death rate if Leronlimab did not work at all and placebo group was the same as treatment group. Instead, we saw 87 deaths. Assume 90 total since people may have died after we were told 87. That means we had 35 fewer deaths than expected.
The ratio of LL to placebo was 2:1 in the trial, so the number of people treated was 262. Expected deaths in the treatment group is 262 x .317 = 83. However, according to our calculations, we had 35 fewer deaths than expected. If this difference was all due to Leronlimab, the actual deaths in the treatment group were 83-35=48. That would leave 42 deaths in the placebo group. These number represent a statistically significant difference with p<0.05.
Since these are conservative estimates of mortality in each group, this is another indication that we will get very good results from CD12 IMO.
I would like to share some statistics with you that have a bearing on the likely outcome of the CD12 trial. I am a doctor who is part of a large hospital network. In our state right now, the most recent statistics show that 67% of the patients in the ICU are on mechanical ventilation. That would place them in the critical category for the CD12 trial. This is an important statistic because we know that the mortality rate in this population was reported to be as high as 80% in the beginning of the pandemic and is probably still at least 35% or higher. If we assume that most or all of patients enrolled in CD12 came from the ICU you can apply the following math:
394 patients in trial
67% critical and 33% severe
394 x 0.667 = 262 critical patients
394 x .333 = 132 severe patients
Now, conservatively assume 40% mortality rate in critical and 15% in severe
262 Critical x 40% mortality rate = 105 expected deaths
132 Severe x 15% expected mortality rate = 20 expected deaths
Total deaths should have been 105 + 20 = 125 for a 31.7% death rate if Leronlimab did not work at all and placebo group was the same as treatment group. Instead, we saw 87 deaths. Assume 90 total since people may have died after we were told 87. That means we had 35 fewer deaths than expected.
The ratio of LL to placebo was 2:1 in the trial, so the number of people treated was 262. Expected deaths in the treatment group is 262 x .317 = 83. However, according to our calculations, we had 35 fewer deaths than expected. If this difference was all due to Leronlimab, the actual deaths in the treatment group were 83-35=48. That would leave 42 deaths in the placebo group. These number represent a statistically significant difference with p<0.05.
Since these are conservative estimates of mortality in each group, this is another indication that we will get very good results from CD12 IMO.
(13)
(0)