Cost-effectiveness of Rotavirus vaccination in Vietnam

Abstract

Background: Rotavirus is the most common cause of severe diarrhea leading to hospitalization or disease-specific death among young children. New rotavirus vaccines have recently been approved. Some previous studies have provided broad qualitative insights into the health and economic consequences of introducing the vaccines into low-income countries, representing several features of rotavirus infection, such as varying degrees of severity and age-dependency of clinical manifestation, in their model-based analyses. We extend this work to reflect additional features of rotavirus (e.g., the possibility of reinfection and varying degrees of partial immunity conferred by natural infection), and assess the influence of the features on the cost-effectiveness of rotavirus vaccination.

Methods: We developed a Markov model that reflects key features of rotavirus infection, using the most recent data available. We applied the model to the 2004 Vietnamese birth cohort and re-evaluated the cost-effectiveness (2004 US dollars per disability-adjusted life year [DALY]) of rotavirus vaccination (Rotarix) compared to no vaccination, from both societal and health care system perspectives. We conducted univariate sensitivity analyses and also performed a probabilistic sensitivity analysis, based on Monte Carlo simulations drawing parameter values from the distributions assigned to key uncertain parameters.

Results: Rotavirus vaccination would not completely protect young children against rotavirus infection due to the partial nature of vaccine immunity, but would effectively reduce severe cases of rotavirus gastroenteritis (outpatient visits, hospitalizations, or deaths) by about 67% over the first 5 years of life. Under base-case assumptions (94% coverage and $5 per dose), the incremental cost per DALY averted from vaccination compared to no vaccination would be $540 from the societal perspective and $550 from the health care system perspective.

Conclusion: Introducing rotavirus vaccines would be a cost-effective public health intervention in Vietnam. However, given the uncertainty about vaccine efficacy and potential changes in rotavirus epidemiology in local settings, further clinical research and re-evaluation of rotavirus vaccination programs may be necessary as new information emerges.

Figures

Figure 1

Model schematic. This figure presents the schematic of the natural history of rotavirus infection. Natural infection by wild type rotavirus can provide protection against subsequent infections with a varying degree of immunity depending on the number of previous infections as well as ages at infection.

Figure 2

Selected model-predicted health outcomes. The upper panel presents the cumulative probabilities of the primary and subsequent rotavirus infections over time under no vaccination. The lower panel shows the cumulative age distribution of hospitalized cases associated with rotavirus gastroenteritis (observed data versus model estimates).

Figure 3

Selected results of one-way sensitivity analyses. This graph presents selected results of univariate sensitivity analyses from the societal perspective. The x-axis represents the ranges of the incremental cost-effectiveness ratios of infant rotavirus vaccination in Vietnamese children when the baseline estimates of several key parameters were varied over plausible ranges. The vertical line represents the base case incremental cost-effectiveness ratio of rotavirus vaccination.

Figure 4

Cost-effectiveness acceptability curves. This figure summarizes the results of probabilistic sensitivity analysis from the societal perspective. The curve shows the probabilities that rotavirus vaccination is cost-effective at varying cost-effectiveness threshold ratios.