Burden of paediatric respiratory syncytial virus disease and potential effect of different immunisation strategies: a modelling and cost-effectiveness analysis for England

Abstract

Background: Vaccines and prophylactic antibodies against respiratory syncytial virus (RSV) are in development and likely to be available in the next 5-10 years. The most efficient way to use these products when they become available is an important consideration for public health decision makers.

Methods: We performed a multivariate regression analysis to estimate the burden of RSV in children younger than 5 years in England (UK), a representative high-income temperate country, and used these results to assess the potential effect of different RSV immunisation strategies (targeting vaccination for infants, or pregnant women, or prophylactic antibodies for neonates). We did a cost-effectiveness analysis for these strategies, implemented either separately or concurrently, and assessed the effect of restricting vaccination to certain months of the year.

Findings: We estimated that RSV is responsible for 12 primary care consultations (95% CI 11·9-12·1) and 0·9 admissions to hospital annually per 100 children younger than 5 years (95% CI 0·89-0·90), with the major burden occurring in infants younger than 6 months. The most cost-effective strategy was to selectively immunise all children born before the start of the RSV season (maximum price of £220 [95% uncertainty interval (UI) 208-232] per vaccine, for an incremental cost-effectiveness ratio of £20 000 per quality-adjusted life-year). The maximum price per fully protected person that should be paid for the infant, newborn, and maternal strategies without seasonal restrictions was £192 (95% UI 168-219), £81 (76-86), and £54 (51-57), respectively.

Interpretation: Nearly double the number of primary care consultations, and nearly five times the number of admissions to hospital occurred with RSV compared with influenza. RSV vaccine and antibody strategies are likely to be cost-effective if they can be priced below around £200 per fully protected person. A seasonal vaccination strategy is likely to provide the most direct benefits. Herd effects might render a year-round infant vaccination strategy more appealing, although it is currently unclear whether such a programme would induce herd effects.

Funding: UK National Institute for Health Research.

Figure 1

Outcomes attributable to respiratory syncytial virus by month of birth (A) General practitioner (GP) consultations, (B) admissions to hospital.

Figure 2

Cases of respiratory syncytial virus averted and costs or QALYs saved for different vaccination strategies with complete vaccine efficacy Data given per 100 annual births for (A) general practitioner (GP) consultations averted,(B) hospital admissions averted, (C) deaths in hospital averted, (D) health-care costs saved, (E) QALYs saved, and (F) maximum cost-effective price (MCEP) of vaccination strategy. M=maternal immunisation strategy. N=newborn passive immunisation strategy. C=infant strategy, N+C=newborn and infant strategies. ICU=intensive-care unit. QALY= quality-adjusted life-year.

Figure 3

Ten most cost-effective periods over which to offer newborn vaccination or prophylactic antibodies against respiratory syncytial virus Red shading shows the birth months of the children that are most cost-effective to protect via a newborn strategy. Prices are per fully protected child for vaccination or prophylactic antibodies. All of the ten most cost-effective programmes involved protecting newborn infants for 5 months or less of the year, the top nine protected for 4 months or less, and the top eight recommended protecting children born in November.

Figure 4

Sensitivity to model parameters of cost-effectiveness calculations for respiratory syncytial virus vaccination (A) Infant strategy. (B) Newborn infant strategy. (C) Maternal strategy. Bars show by how much the maximum cost-effective price changes from its base case level when model parameters are varied. Changing the discounting strategy, or excluding deaths from the model, had little effect on the maximum cost-effective price (MCEP) estimates. Similarly, modifying the costs or quality-adjusted life-years (QALYs) associated with each health-care outcome had little effect on model estimates.