If the mom is O and the baby is AB with a pos DAT, I guess there could be more than the usual circulating antibodies involved--both anti-A and anti-B (or more probably anti-A,. I guess it's more a matter of the same antibodies having more antigens to attack and the techs wondering if there is a specimen mix-up. Also, a baby can be homozygous for an antigen the mom has antibodies to if she is not the genetic mother. That's about all I have ever heard of about it. It might be more difficult to find units for an exchange transfusion that are compatible with the baby as well as the mom's antibodies--but that happens with genetic moms sometimes too.