The long-term burden of congenital cytomegalovirus: Hospitalisation and mortality in a population-based matched cohort study.

Abstract

Congenital cytomegalovirus (cCMV) infection can result in considerable morbidity and mortality. However, the impact of cCMV on health system utilisation beyond infancy is poorly defined. Here we sought to describe the burden of cCMV health service use and mortality using linked population-based datasets. The design was a matched cohort study using record-linked administrative datasets. Participants included all children aged 0-15 years identified through the New South Wales (NSW), Australia, Admitted Patient Data Collection who were hospitalised with an admission code for cCMV infection between 2001 and 2011. Participants were then matched by age, sex and birth postcode quintile using Socio-Economic Indexes for Area, to children randomly selected from the NSW Perinatal Data Collection, excluding those with central nervous system infections or cerebral palsy. We calculated rate ratios (RR) for hospital admissions adjusted for preterm birth, the median length of stay, cost weights of admission and odds ratio (OR) for deaths. There were 130 children with cCMV matched to 2672 children. Neonates (aged <1 month) with cCMV had twice the rate of admissions (RR 2.4 95%CI 2.0, 2.8) and children aged ≥ 5yr to<15yrs with cCMV almost eight times the rate of admissions (RR 7.8 95%CI 5.1, 11.5). Children with cCMV had significantly higher cost weights of admissions and an increased risk of mortality (OR 18.4 95%CI 7.8, 43.6). Throughout childhood, children with cCMV had higher rates of hospital admissions, higher admission cost weights and an increased risk of mortality compared with matched peers without cCMV.