Plasma exchange in severe rhesus haemolytic disease.

SIR,-In his review of maternal serum afetoprotein (AFP) screening in pregnancy (17 May, p 1199) Rodney Harris concludes that one potential benefit of population screening might be the identification of fetuses at risk of spontaneous abortion, low birth weight, or perinatal death. He bases this conclusion on the increased incidence of these problems among women who have a high maternal serum AFP concentration but a normal result on amniocentesis. The possibility that the prenatal diagnostic intervention may have caused the subsequent problems is discounted by reference to the significantly better pregnancy outcome of women who have amniocentesis for other reasons. An outcome measure which receives no attention in Harris's review, but which I have found to be significant in differentiating women with high serum AFP concentrations from those with other indications for amniocentesis, is the woman's emotional reaction to the diagnostic intervention.' The data in the table are based on interviews with 90 women who had a normal result on amniocentesis. Although nearly all women were worried about