That is what we're saying. In fact, there were 28 former staff who came forward with evidence of protocol deviations. That's what they're called. That wasn't the only problem with the Canadian trial. In fact, they allowed women to be participants even if they had a known breast lump. Screening is for women with no lumps.
First of all, they allowed these women to participate. They were having trouble recruiting enough women for the study, and they actually approached breast surgeons to send patients to be in the study. The reason a woman goes to a breast surgeon is that she has a lump or a symptom.
First of all, they allowed these women to participate. What was supposed to happen was that every woman who came to participate—they were volunteers—got a clinical breast exam by a highly trained nurse, and then they would go to the coordinator, who would decide to put her in either the study group, where they got a mammogram, or the control group, where they didn't.
Nowadays, when we do these studies, the randomization is done by a central office, by a computer. In those days, the coordinators had a piece of paper in front of them with lines. The lines would say, “mammogram, control, mammogram, mammogram, mammogram, control, control”, and at the end of the sheet, you'd have an equal number of women in both arms.
What we know happened, because witnesses came forward and told us—and these are in three peer-reviewed published papers, by the way—is that the nurses would say, “This lady has to be in the mammogram group,” so the coordinator could write her name on the next available mammogram line, and then other women who came in later in the day could go in any blank lines that she had left. They didn't even have to make any erasures.
This was actually picked up in 1992, which was the first publication of the Canadian national breast screening study, because there was a significant imbalance of advanced cancers. In the very first year of the study, there were 25 advanced cancers, which they defined as a cancer with more positive lymph nodes in the armpit. There were 19 of them in the mammogram group and only five in the control group.
This was raised decades ago, and the principal investigators at the trial have denied it to this day. They claim that there was nothing wrong with the randomization. There was even a forensic—