489 - Universal vs. Targeted CMV Testing in the NICU

Publication Number: 3471.489

Erin Shumans, Case Western Reserve University School of Medicine, Cleveland, OH, United States; El S. Healey, Case Western Reserve University School of Medicine, Westlake, OH, United States; Prabhu S. parimi, Case Western Reserve University School of Medicine, Cleveland, OH, United States

Background: CMV is the most common congenital infection in the US, and is the leading cause of non-syndromic sensorineural hearing loss. Many congenitally infected infants are asymptomatic, but 1 in 5 will have long-term sequelae such as hearing loss or developmental delay. Early treatment in symptomatic infants has been shown to improve long term outcomes. However, some symptoms of congenital CMV may be subtle, and require a high index of suspicion to test. There is not a consensus on who to test. In some states all infants are screened at birth, while in others a more targeted approach is used.

Objective: Our NICU started universal CMV screening in September 2024 while newborn nursery continued targeted testing. We aimed to determine if we should continue with universal screening.

Design/Methods: We designed a retrospective case/control study to evaluate differences in our unit pre and post universal screening. Inclusion criteria was any infant in NICU or newborn nursery from 1/1/2015 to 7/1/2025 with congenital CMV, plus 3 negative case-matched controls for each positive infant. Case matching was done on the basis of infant sex and gestational age at birth, and maternal race. Exclusion criteria was infants not tested for CMV. For each included infant, information about maternal biographical information, pregnancy history, infant symptoms, and infant anthropometrics was gathered. Logistical regression was used to determine the most sensitive predictors of CMV infection. Non-individualized yearly data was also pulled from the EMR including number of infants tested, number of infants with diagnosis codes for common CMV symptoms, and total number of births.

Results: During the years of targeted testing, the percentage of all births tested ranged from 2-8%. During the 2 partial years of universal testing in NICU only, the percentage tested ranged from 9.7-24%. During the years of targeted testing, the percent positive ranged from 0-0.2%. During the 2 partial years of universal screening, the percent positive was 0.15% and 0.16%. Logistic regression showed that maternal infection, hepatomegaly, elevated LFTs, and persistent thrombocytopenia were the most sensitive predictors of infant infection. Low birth weight, SGA, prematurity, and microcephaly were the most common symptoms tested for, but did not have the highest positivity rate. There were no asymptomatic positive infants in the NICU during the 10 months of universal screening.

Conclusion(s): Our NICU has a low rate of CMV, and all of the positive infants would have been captured by a targeted testing protocol. Therefore targeted testing is adequate in our NICU.

Figure 1
Graph with percent tested and percent positive for all births, and percentage of all births with diagnosis codes of several significant CMV symptoms.

Figure 2
Table with data from figure 1 - percent tested and percent positive for all births, and percentage of all births with diagnosis codes of several significant CMV symptoms along with positivity rates for those symptoms.

Figure 3
Graph of prevalence of each CMV symptom in positive and negative cohorts in our case-control groups.