Inquiry into stillbirths and infant deaths.

Although the data are not available, it seems that the risk to the fetus of asymptomatic reinfection is small,3" and certainly it is far less than that of primary rubella. In the rare cases of symptomatic reinfection viraemia may be presumed to have occurred, putting the fetus at appreciable risk.9 An antibody response may be confidently ascribed to reinfection if two or more previous positive results for rubella antibody have been obtained. One previous positive result must be treated with scepticism as errors may have occurred.12 Hence the Royal College of Obstetricians and Gynaecologists and a working party of the Public Health Laboratory Service advise that every woman should be tested for rubella anti-bodies in every pregnancy.6 A history of rubella immunisation also means that primary rubella is unlikely, but vaccination failures are well recognised.'2 If a patient has a rising titre of rubella antibody reinfection may be discriminated serologic-ally from primary rubella if specific IgM is not found or is found only in low concentration.91' 13 The recent finding of raised concentrations of specific IgM in confirmed reinfec-tions (unpublished observations) adds to the problem. If only late serum samples are available and there is no history of previous testing or immunisation then specific IgM (even at low concentration) may reflect recent reinfection or primary rubella some weeks previously. Recently developed assays that measure specific IgG subclasses and avidity may allow this distinction to be made, but the tests have not yet been validated in routine use. 416 A reliable and safe method of prenatal diagnosis of fetal infection would be valuable when there is diagnostic uncertainty or proved reinfection (although fetal infection does not necessarily mean fetal damage). Approaches include culturing amniotic fluid for rubella virus, which is time consuming and unreliable, and detecting specific IgM in fetal blood, which has to be delayed until 23-24 weeks ofpregnancy. Detecting the rubella virus genome and antigen in chorionic villus samples is promising but requires further assessment. 17 Correct diagnosis of primary rubella or reinfection in patients without symptoms may thus be difficult and lead to much anxiety about how to manage the pregnancy. It is thus essential for diagnosis that full information is given to the laboratory on symptoms, date and type of contact, previous rubella immunisation, and dates and results of previous tests for rubella antibody. ELISAs and an anti-IgM capture immunoassay. Clinical rubella with virus transmission to the …