With a prevalence at birth of 0.7% and 0.4% worldwide [1] and in Europe [2,3] respectively, congenital CMV infection (cCMV) is the commonest congenital infection and the leading non-genetic cause of sensorineural hearing loss and a major cause of neurological disability [4]. In Western European countries, where CMV seroprevalence is around 50% in young adults, half of cCMV infected neonates are infected following a maternal primary infection (PI) and the other half following maternal non-PI [2,3]. In women with PI, all long-term sequelae are seen following PI that occurred between 3 weeks prior to conception and 14 weeks. [5,6].

Despite the burden of cCMV, no Public Health body has ever recommended universal serology screening in pregnant women outside Greece. This was mainly explained by the absence of any validated treatment, the variability in CMV serological assays and the cost of screening. However, recent data from one randomized controlled trial, 2 confirmatory studies and one meta-analysis of individual data established the efficacy of early antiviral therapy to prevent vertical transmission from the mother to the fetus in women with PI in the first trimester of pregnancy (T1) [7], [8], [9], [10]. This major breakthrough should convince Public Health authorities to revise their position and consider moving toward universal serology screening in pregnancy. The question of the choice of the best algorithm for CMV serological screening in pregnancy will become crucial.

Diagnosis of PI is based on a combination of 3 serological markers: IgG, IgM and IgG avidity. A CMV serological screening strategy in pregnancy should have a high positive predictive value (PPV, be able to diagnose all cases of PI in T1 and in the periconceptional period from 3 weeks before to 14 weeks after the conception)[11] and a high negative predictive value (NPV, be able to exclude cases that happened neither in T1 nor in the periconceptional period). Two screening strategies are possible: the first one is to screen for IgG and IgM and then to perform IgG avidity to date the infection when IgM are present. The second is to screen IgG alone and to perform IgG avidity when positive. A single small study, including 310 cases, compared the these 2 strategies and concluded that the strategy based on IgM detection had better performance because with the second strategy 26% of all tested sera had intermediate avidity results and were therefore difficult to classify [12]. Since then, studies reporting CMV serology screening in pregnancy were based on IgG and IgM screening. However, the study by Mace et al, was done nearly 20 years ago, and used a first generation IgG avidity assay. Second generation avidity assays demonstrate better sensitivity and specificity to detect a recent PI [13], [14], [15]. We aimed to revisit the dogma of IgM screening and to compare this strategy to a strategy based on IgG alone followed by IgG avidity in cases with positive IgG using newer generation assays.