Sensitivity and specificity of Rinne tuning fork test.

the site of the placenta. Administration ofanti-D immunoglobulin is routine after invasive procedures in Rh negative patients but is not appropriate for those already sensitised, in whom it would be ineffective.'5 In patients undergoing intra-uterine transfusions, however, fetomaternal haemorrhage may enhance maternal sensitisation. Not only might the disease be worsened in a subsequent pregnancy but also in the present pregnancy the rate of fall in fetal packed cell volume might be accelerated by an increase in maternal antibody titre, particularly after the first transfusion, when fetal erythropoiesis is generally not yet suppressed. Precisely estimating the volume of fetomaternal haemorrhage is difficult. In patients undergoing intra-uterine transfusions the Kleihauer-Betke test cannot be used because donor red cells cannot be identified in the maternal circulation. The increase in maternal a fetoprotein concentration can give a reliable estimate if fetal a fetoprotein concentration is also assayed. Possible sources of error are the estimated maternal blood volume, which cannot practically be assessed in each patient, and passage of a fetoprotein into the maternal circulation from fluids other than fetal blood, such as the amniotic fluid. This second possibility is unlikely as fetomaternal haemorrhage occurred most commonly when the site of sampling and transfusion was the placental cord insertion with an anterior placenta-that is, when the needle entered directly into the cord without any communication with the amniotic cavity. Moreover, any contribution of a fetoprotein from the amniotic fluid to the rise in maternal a fetoprotein concentration should be irrelevant as the concentration in fetal blood is more than 100 times that in amniotic fluid.'6 De Silva et al reported that 0-28 ml of red cells produced a secondary Rh immune response in all of their 14 volunteers'7; in our study the increase in maternal Rh (D) antibody concentration was clinically important when the volume of haemorrhage was greater than 1 ml. As the mean fetal packed cell volume at the first transfusion was 0-20 (range 0-08-0-31) the results were comparable. In some instances the rise in antibody titre was dramatic, and although the consequences were not important as far as the rate of fall in fetal packed cell volume was concerned, the course of the disease in a subsequent pregnancy will probably be affected. These results also suggest that sampling of fetal blood in patients with Rh isoimmunisation should not be routinely done for blood grouping early in pregnancy. It would be without consequence …