The yield of targeted examination for the detection of symptomatic congenital cytomegalovirus infection

Background: The incidence of congenital cytomegalovirus (CMV) infection in Israel is 0.7%. Only 10-15% are symptomatic. Valganciclovir has been shown to improve hearing and neurodevelopmental outcomes in neonates with symptomatic congenital CMV infection. Targeted examination of infants who fail routine neonatal hearing screening or have clinical or laboratory findings suggestive of symptomatic congenital CMV infection may be a cost-effective approach. Objectives: To assess the yield of targeted examination for the detection of newborns with symptomatic congenital CMV infection. Methods: A prospective observational study was conducted in 2014-2015 at two medical centers in northern Israel. Included were all newborns who were tested in the first 3 days of life by polymerase chain reaction (PCR) for urine CMV DNA (n=692), either for failure the hearing screening (n=539, 78%), clinical or laboratory findings suggestive of symptomatic congenital CMV infection, or primary CMV infection during pregnancy (n=153, 22%). Results: During the study period 15,433 newborns were born. The predicted rate of infection was 10-15% (symptomatic) of 0.7% of newborns, namely 0.07-0.105% or 10-15 infants. In fact, 15 infants (0.11%, 95% confidence interval 0.066-0.175) were diagnosed with symptomatic congenital CMV infection, 2/539 (0.37%) in the failed hearing group and 13/153 (8%) in the clinical/laboratory findings group. The incidence of symptomatic congenital CMV infection was within the predicted range. Conclusions: Targeted examination of only 4.5% (n=692) of newborns detected the predicted number of infants with symptomatic congenital CMV infection in whom valganciclovir therapy is recommended.