Breaking (research): Japanese data (Gibo et al.) gives STRONG signal (adds to BODY of evidence) that the Malone Bourla Bancel Pfizer Moderna Sahin Kariko et al. mRNA transfection shot causes cancer
using age-adjusted data; this is not a comparative effectiveness Randomized placebo controlled double/triple blind study, is associational/relationship & confounded but a strong temporal, biological
plausibility, potential biological gradient (dose response with boosters etc.) indicates all/most tenets of Bradford Hill are met as to causality when not using cause and effect research models. See below for list of Hill’s criteria for declaring causation or which strengthens the potential for a causal link. Note, a well conducted observational cohort study, proper statistical controls, procedural controls etc. can be more optimal than a poorly conducted small sample size RCT. Not because it is a RCT makes it purely gold standard for if the base sample size is small from which sequence generation resulted and the allocation/randomization to groups are small sized, then randomization could have broken down in terms of distributing confounding factors (unknown) near evenly across groups. We have found methodologically that sample size must approach 60 (at least 30 per group, trial arm) and above to allow for proper baseline balance of known and unknown distorting confounding variables. Above 30 n=30 per trial arm allows data to assume a normal population Gaussian bell-curve distribution (and above of course). Just some tips.
But note, a RCT is based on a highly selected study (inclusion criteria) and thus the population is highly selected and no matter how pure the methods are, the findings cannot be extrapolated or generalized to the general population. Only to the population under study. I did a lot of research work on the need for large, pragmatic, limited selection criteria research RCTs so that the findings could be extrapolated to the wider populations.
