CD-12 Expectations: A survey of iHangout CYDY Board Members
“The TOP LINE REPORT on what we expect from the TOP LINE REPORT”
Thirty-nine iHangout CYDY Board members provided their best estimates of deaths in each arm of the trial. Data were collected 2/11 through 2/14.
==> Only two of 39 projections non-significant at P=.05, only one more non-significant at P=.01.
==> The Median Case: 43 treatment deaths + 45 control deaths = 88 total; P=.0004, Relative risk 0.52 (95% CI 0.36 - 0.75), Absolute lives saved 15.6 per hundred treated.
Results for Individual Variables
Total deaths
Median total deaths were forecast at 89, with a tight intra-quartile range of 87 to 90. Only 2 respondents were above 92, and only 2 below 82. The distribution was skewed but smooth, with the long low-end tail reflecting those who believe substantial deaths occurred post day-28.
Deaths in the Control Arm
Median control arm deaths were forecast at 43, or 32.8%, with in intra-quartile range of 40 (30.5%) to 47 (35.9%). The lowest six estimates were 34, 36, 38 and three at 39. Thirty-nine is important because it allows 91 total deaths for significance, while 38 only allows 88. There was a long high-end tail to this distribution with six estimates of 54 or over.
Deaths in the Treatment Arm
Median treatment arm deaths were forecast at 45, or 17.2%, with in intra-quartile range of 38 (14.5%) to 47 (17.9%). Only four estimates were above 49, while the long low-end tail showed four projections below 31.
Relative Risk and Significance
Only two cases were not significant at p=.05, only one more non-significant at p=.01.
Median relative risk was forecast at 0.52, or a 48% reduction in likelihood of an individual dying versus the control group. The intra-quartile range is 0.59 down to 0.39. There was a long tail on this distribution with six respondents indicating a risk reduction of 2/3 or more.
Absolute Risk Reduction
Median case showed absolute lives saved at 15.6 per hundred treated: 32.8 in the control group down to 17.2 the treatment arm The interquartile range was 12.6 to 21.4. Five cases in the long tail showed lives saved per hundred treated at over 30.
(Note: I use "deaths per hundred treated" as quoting this in percentage points loses the fact these are lives!)
“The TOP LINE REPORT on what we expect from the TOP LINE REPORT”
Thirty-nine iHangout CYDY Board members provided their best estimates of deaths in each arm of the trial. Data were collected 2/11 through 2/14.
==> Only two of 39 projections non-significant at P=.05, only one more non-significant at P=.01.
==> The Median Case: 43 treatment deaths + 45 control deaths = 88 total; P=.0004, Relative risk 0.52 (95% CI 0.36 - 0.75), Absolute lives saved 15.6 per hundred treated.
Results for Individual Variables
Total deaths
Median total deaths were forecast at 89, with a tight intra-quartile range of 87 to 90. Only 2 respondents were above 92, and only 2 below 82. The distribution was skewed but smooth, with the long low-end tail reflecting those who believe substantial deaths occurred post day-28.
Deaths in the Control Arm
Median control arm deaths were forecast at 43, or 32.8%, with in intra-quartile range of 40 (30.5%) to 47 (35.9%). The lowest six estimates were 34, 36, 38 and three at 39. Thirty-nine is important because it allows 91 total deaths for significance, while 38 only allows 88. There was a long high-end tail to this distribution with six estimates of 54 or over.
Deaths in the Treatment Arm
Median treatment arm deaths were forecast at 45, or 17.2%, with in intra-quartile range of 38 (14.5%) to 47 (17.9%). Only four estimates were above 49, while the long low-end tail showed four projections below 31.
Relative Risk and Significance
Only two cases were not significant at p=.05, only one more non-significant at p=.01.
Median relative risk was forecast at 0.52, or a 48% reduction in likelihood of an individual dying versus the control group. The intra-quartile range is 0.59 down to 0.39. There was a long tail on this distribution with six respondents indicating a risk reduction of 2/3 or more.
Absolute Risk Reduction
Median case showed absolute lives saved at 15.6 per hundred treated: 32.8 in the control group down to 17.2 the treatment arm The interquartile range was 12.6 to 21.4. Five cases in the long tail showed lives saved per hundred treated at over 30.
(Note: I use "deaths per hundred treated" as quoting this in percentage points loses the fact these are lives!)
26
