Bassler & Guyatt et al.: "Stopping randomized trials early for benefit and estimation of treatment effects: systematic review and meta-regression analysis"; supports Dr. Harvey Risch & points to fraud
by Paul Alexander
COVID vaccine estimates; see Figure 3 page 1186; very large over estimation (ratio of RR, 0.37; 95% confidence interval [CI], 0.31-0.44; P<.001) occurred in stopped trials with less than 200 events.
I conduct research with the people in this paper above.
The key is that stopping early for benefit and with outcome events less than 200 places your trial for serious risk of biased estimates of effect. They would be wrong. The estimates are wrong. So the 0.7% absolute risk reduction (ARR) in Pfizer’s legacy trial (remember, they deceptively reported relative risk reduction RRR of 95% when it should have been the ARR reported) will be basically zero (0) as was greatly over-estimated.
I join Dr. Risch as the guru, I mean his Brownstone article is off the chain as they say. So I wanted to dig into the evidence vault to show you he has serious technical underpinnings for his assertions. We modeled this in the past and showed he is correct as to small outcome event number.
Fraudulent and sub-optimal Pfizer and Moderna findings because the trial was stopped early for benefit and thus high-risk of stopping at a ‘random high’ and thus declaring benefit when the benefit may have disappeared had the study run to sample size; also, the number of events as per model above is less than 200 (legacy Pfizer was 162 placebo and 8 in treatment) and thus at heavy risk of over-estimation of treatment effect.
SOURCE:
https://pubmed.ncbi.nlm.nih.gov/20332404/
see Risch’s Brownstone paper that supports as well as is supported by this modelling paper of optimal outcome event number and as Risch argues, once outcome event number is small, then there is a likely breach of randomization of outcome events (focus should not only be on optimal randomization of the sample e.g. larger sample sizes), confounding, and we argue over-estimation of treatment effect (biased estimates of effect):