Cost-effectiveness of rotavirus vaccination in children under five years of age in 195 countries: A meta-regression analysis

Abstract

Background: Rotavirus caused an estimated 151,714 deaths from diarrhea among children under 5 in 2019. To reduce mortality, countries are considering adding rotavirus vaccination to their routine immunization program. Cost-effectiveness analyses (CEAs) to inform these decisions are not available in every setting, and where they are, results are sensitive to modeling assumptions, especially about vaccine efficacy. We used advances in meta-regression methods and estimates of vaccine efficacy by location to estimate incremental cost-effectiveness ratios (ICERs) for rotavirus vaccination in 195 countries.

Methods: Beginning with Tufts University CEA and Global Health CEA registries we used 515 ICERs from 68 articles published through 2017, extracted 938 additional one-way sensitivity analyses, and excluded 33 ICERs for a sample of 1,418. We used a five-stage, mixed-effects, Bayesian metaregression framework to predict ICERs, and logistic regression model to predict the probability that the vaccine was cost-saving. For both models, covariates were vaccine characteristics including efficacy, study methods, and country-specific rotavirus disability-adjusted life-years (DALYs) and gross domestic product (GDP) per capita. All results are reported in 2017 United States dollars.

Results: Vaccine efficacy, vaccine cost, GDP per capita and rotavirus DALYs were important drivers of variability in ICERs. Globally, the median ICER was $2,289 (95% uncertainty interval (UI): $147-$38,993) and ranged from $85 per DALY averted (95% UI: $13-$302) in Central African Republic to $70,599 per DALY averted (95% UI: $11,030-$263,858) in the United States. Among countries eligible for support from Gavi, The Vaccine Alliance, the mean ICER was $255 per DALY averted (95% UI: $39-$918), and among countries eligible for the PAHO revolving fund, the mean ICER was $2,464 per DALY averted (95% UI: $382-$3,118).

Conclusion: Our findings incorporate recent evidence that vaccine efficacy differs across locations, and support expansion of rotavirus vaccination programs, particularly in countries eligible for support from Gavi, The Vaccine Alliance.

Keywords: Cost-effectiveness analysis; Meta-regression; Rotavirus vaccine.

Fig. 1

Study flow diagram of study selection in logistic and meta-regression models. * For articles with more than one ratio, some ratios may be excluded and others included. CEA = cost-effectiveness analysis, ICER = incremental cost-effectiveness ratio, QALY = quality-adjusted life-year, DALY = disability-adjusted life-year.

Fig. 2

Observed log-ICER vs fitted log-ICER overall and by GBD super-region. Fig. 2legend. Observed log-ICER vs fitted log-ICER overall and by GBD super region. The minimum and maximum values for Peru (PER), Bangladesh (BGD), and Cameroon (CMR) are shown to highlight differences in observed log-ICERs across studies within a given country. ICER = incremental cost-effectiveness ratio.

Fig. 3

Predicted ICERs from meta-regression analysis by country.Fig. 3legend. (A) Predicted ICERs from meta-regression analysis by country in 2017 USD per DALY averted; (B) Predicted ICERs relative to seven categories of GDP per capita ranging from <0.5 to >3.0 times GDP; C) GDP category in which the upper bound of the 95% UI falls. ICER = incremental cost-effectiveness ratio, UI = uncertainty interval, DALY = disability-adjusted life-year.