Ok, I had NEVER heard of this and I'm an avid researcher. There is this phenomenom called the hook effect in which if the hcg is too fragmented, it doesn't show up on urine hcg tests. Also, the higher the hcg the more it can be fragmented. So, someone with twins and who is later in pregnancy is more likely to get a negative urine test. This web site explains it best:
"How can excess of a fragment of hCG cause a false negative result. This seems counterintuitive?
That's a good question. I think that most laboratorians are familiar with the concept of the so-called hook effect. This occurs in sandwich assays where there are two antibodies that act like bread, and they form a sandwich with antigen in the middle. When there is a great excess of the antigen and there is no wash step to take away the excess, then sandwich formation is inhibited. What you end up with is a bunch of open-face sandwiches, if you will. This is a wellknown phenomenon and can occur in both qualitative and quantitative hCG assays."
I know that's tough reading, but it comes down to how some women process hcg differently and if the parts of the hcg is fragmented and in different proportions than a urine test can come back negative.
Here's another part:
"When a negative result is suspected to be a false-negative, a simple dilution can be performed. The idea here is that as with the hook effect, you can get the variant antigen down to a concentration that does not block all the antibodies, and then a sandwich can form with the intact hCG."