During the manual randomization phase:
"Our CRO did not have electronic randomization capability and instead distributed sealed envelopes containing patient randomization assignments to each site. During the manual randomization phase, the Company noted a large number of patients with renal failure being entered into the treatment arm versus control (57% vs 17% control, p=0.05) resulting in a much sicker treatment arm. Envelopes were audited and two sites most responsible for this bias were also found to have had envelopes
opened with no corresponding patient enrolled."
"After a thorough review by the SAB and external DSMB, they concluded that data from the manual randomization phase was biased by a compromise of randomization and unanimously recommended to reset the clock with electronic randomization. It was believed that the remaining 47 patients to be enrolled under electronic randomization could be sufficient to achieve the goals of the study. Statistically, electronic randomization corrected the randomization issue. Except for safety data, only data from the electronic randomization phase will be presented here."
This is the real story lhat is never told...