National burden estimates of hospitalisations for acute lower respiratory infections due to respiratory syncytial virus in 2019: the Burden of Influenza and RSV Disease (BIRD) study

Abstract

# Abstract # ## Background ## Respiratory syncytial virus (RSV) is the predominant viral pathogen associated with acute lower respiratory infection (ALRI) in children under five years of age (<5y). Little is reported on the national estimates of RSV-ALRI hospital admissions in <5y based on robust RSV epidemiology data. ## Methods ## We included data on RSV and ALRI hospital admission in <5y from systematic literature reviews (including unpublished data) and from clinical databases. We used two different methods, namely the rate-based method and the proportion-based method, to estimate national RSV-ALRI admissions in <5y in the year 2019 among countries with robust data. The rate-based method synthesized data on laboratory-confirmed RSV-ALRI hospitalisation rates using a spatiotemporal Gaussian process meta-regression (ST-GPR). The proportion-based method applied data on RSV positive proportions among ALRI to all-cause ALRI admission envelopes using a Bayesian regularized trimmed meta-regression (MR-BRT). Where applicable, we reported estimates by both methods in order to provide a plausible range for each included country. ## Findings ## A total of 334 studies and 1985 data points from 58 countries were included in the analysis, accounting for 58% of the <5y population worldwide. The number of the annual national RSV-ALRI hospitalisations in <5y ranged from 0·06 thousand (95% confidence interval [CI]: 0·02–0·14) in Iceland to 936·29 thousand (95% CI: 261·23–2151·83) in China. Despite great variation among countries, a high proportion of <5 of RSV-ALRI hospitalisations were in infants <1y in all countries (median proportion: 46%, interquartile range: 32–57). In most (76%) years, RSV-ALRI hospitalisation rate fluctuated between 80% and 120% of the country’s median yearly rate. General agreement was observed between estimates by the rate method and proportion method, with a few exceptions in India, Kenya, Norway and Philippines. The estimates by the rate-based method and by the proportion-based method could be extrapolated to a global estimate of 4·1 million (95% CI: 3·8–4·5) and 2·7 million 95% CI: 1·7–5·2) RSV-ALRI hospitalisations in <5y, respectively. ## Interpretation ## By incorporating data from various sources, our study provides robust estimates on the national burden of RSV-ALRI admissions in <5y. These estimates are important for informing policy for introduction of RSV immunisations and also serve as baseline data for assessing the effectiveness of implementing RSV immunisation programmes. ## Funding ## The Foundation for Influenza Epidemiology