From what I remember (from the book Bad Blood), most patient uses fell into one of two categories: either they were drawing a large draw, running on their machi

From what I remember (from the book Bad Blood), most patient uses fell into one of two categories: either they were drawing a large draw, running on their machines and traditional machines to check correctness, and giving what they deemed correct from these (likely the traditional machine answer), or they were just outright running large samples on traditional machines with no attempt to do anything new. I think there were a few cases of patients getting the wrong results, maybe they were correcting later, but all in all it was not a huge issue for patients given how it was running in practice. This is not to excuse the practices in any way, but I do get why the focus in the trial was not on defrauding patients.