[Seroprevalence of rubella virus, varicella zoster virus, cytomegalovirus and parvovirus B19 among pregnant women in the Sousse region, Tunisia].

The aim of the study is to evaluate seroprevalence of rubella virus (RV), cytomegalovirus (CMV), varicella zoster virus (VZV), and parvovirus B19 (PB19) in 404 Tunisian pregnant women, and to determine reliability of maternal past history of eruption. Sociodemographic characteristics, risk factors, and past history of eruption were collected through a questionnaire. Serologic tests were performed using enzyme immunoassays. Risk factors were analyzed using univariate and multivariate logistic regression models. Seroprevalences were 79.7% for rubella, 96.3% for CMV, 80.9% for VZV, and 76.2% for PB19. In multivariate analysis, the number of persons per room (> 2) in the house during childhood was associated with CMV infection (P = 0.004), irregular professional husband's activity was correlated with VZV infection (P = 0.04), and an age of more than 30 years was associated with PB19 infection (P = 0.02). History of rubella, varicella, and PB19 infection was unknown for, respectively, 55.8%, 20%, and 100% of women. False history of rubella and varicella were found for 7.4% and 15% of women, respectively. The positive and negative predictive values (PPV and NPV) of rubella history were, respectively, 92.6% and 17.2%, and were, respectively, 84.9% and 20.9% for varicella history. Susceptibility to RV, VZV, and PB19 infection remains high in pregnancy in our population. Preventive strategies against congenital rubella must be reinforced. Vaccination against VZV should be considered in seronegative women. Systemic CMV screening is not warranted in our country where high immunity is acquired probably in childhood. Since maternal history of eruption is not reliable, we recommend serologic testing to determine immune status of women.