Prenatal Screening for CMV Primary Infection: A Cost-Utility Model

Abstract

Objective: Congenital cytomegalovirus (CMV) infection is a major cause of deafness and neurodevelopmental disability in children. Our objective was to assess the cost utility of first-trimester serological CMV screening, compared to screening of high-risk pregnancies and no serological screening.

Design: A decision-analytic model was created to compare the cost utility of three strategies from a healthcare sector perspective: universal first-trimester serological screening, screening only of high-risk pregnant women (both including antiviral prophylaxis in cases of primary infection) and serological testing triggered by foetal morphological ultrasound (no CMV serological screening).

Setting: Canada.

Population: Hypothetical population of 80 000 pregnant women.

Methods: Probability, expected values and cost estimates were derived from published literature and local hospital and national insurance data.

Main outcome measure: Cost per maternal and infant quality-adjusted life year (QALY) lost.

Results: Universal serological screening was superior to both screening of high-risk women and no screening (utility of -0.42, -0.63 and - 0.87 QALY lost, respectively). Sensitivity analysis demonstrated that universal screening was the most cost-effective strategy regardless of the incidence of primary infection, the acceptability of amniocentesis and the efficacy of antiviral prophylaxis. In the Monte Carlo analyses, universal serological screening was the most cost-effective option in 96.36% of simulations. Universal serological screening would allow detection of 152 cases of primary maternal CMV infection and would prevent 29 cases of congenital CMV infection annually.

Conclusion: Our findings support the adoption of a population-based prenatal screening programme for the prevention of congenital CMV infection.

Keywords: cost utility; cytomegalovirus; prenatal screening.

FIGURE 1

Tornado diagrams. (a) Incremental cost‐utility ratio of universal CMV serological screening versus no serological screening. (b) Incremental cost‐utility ratio of CMV serological screening of high‐risk women versus no serological screening.

FIGURE 2

Impact of amniocentesis acceptability and positive serological screening rates on net benefit: Bivariate sensitivity analysis at a 50 000 CAD willingness‐to‐pay threshold.

FIGURE 3

(a) Monte Carlo analysis. LF: Life year. (b) Cost‐utility scatterplot. Dots represent combinations of incremental costs and quality‐adjusted life‐years from 10 000 Monte Carlo iterations for each strategy.