I read with interest the prospective study by Skjöldebrand-Sparre et al. (Vol 107, April 2000)1 which found detectable parvovirus B19 DNA in 7/93 (7.5%) of all women with third-trimester intrauterine fetal death.
Before concluding that B19 was the likely cause of fetal loss in these seven cases, it is essential to know the prevalence of detectable parvovirus B19 DNA in placental tissue from normal births and the risk of fetal loss associated with B19 infection. The authors used placental tissues from 50 unselected first-trimester spontaneous abortions as controls and detectable B19 DNA was found in one. There are two methodological problems. First, the incidence of B19 infection is known to differ in the first trimester and the third trimester2. Placental tissues from normal births as controls would be more appropriate. Second, the small number of controls means that the 95% confidence interval for the prevalence is wide. The calculated upper 95% confidence limit is 6%.
A large prospective study in Spain found the incidence of acute B19 infection during pregnancy to be 3.7% and the incidence of fetal loss 1.66%3. Although the prevalence of placental parvovirus DNA in normal pregnancy was not directly measured in this study, the results suggest that it can be quite high and that fetal loss occurs only in a small proportion of infected mothers.
Hence, routine investigation of B19 polymerase chain reaction in all intrauterine fetal deaths as suggested by the authors might result in wrongly attributing fetal loss to the infection in a large number of intrauterine fetal deaths. A larger study of the prevalence of detectable B19 DNA in placental tissues from normal pregnancies is needed.