(B-140) Estimating the burden of viral disease using the Comparative Risk Assessment framework from the Global Burden of Disease study – Two case studies assessing the impact of influenza and respiratory syncytial virus

Background: Viruses like influenza, respiratory syncytial virus (RSV), and dengue are widespread, significantly impacting populations worldwide. However, their non-specific symptoms and limited routine testing make accurate burden assessment challenging, despite surveillance systems like WHO GISRS and the PAHO/WHO Regional Dengue Program.

Objectives: We propose utilizing the Comparative Risk Assessment (CRA) framework, a key component of the Global Burden of Disease (GBD) study, to assess the impact of these viruses on population health, and present two case studies estimating the effects of population-level influenza and RSV activity on specific respiratory and cardiovascular outcomes.

Methods: The CRA framework calculates the population attributable fraction (PAF) for any exposure-outcome pair, that is, the percentage of disease cases linked to specific exposures. We adapted this approach from the GBD study to link population-level viral activity to health outcomes using time-series regression, a method common in environmental studies. The derived exposure-response relationships are then combined with exposure data to estimate attributable disease burden. In our first case study, we estimate influenza-attributable ischemic heart disease (IHD) deaths using surveillance and mortality data, extrapolating results to 204 countries for global weekly and annual estimates. In our second case study, we estimate RSV-attributable hospitalizations among older patients in 15 middle- and high-income countries, accounting for influenza, COVID-19, seasonal trends, temperature, and humidity.

Results: We found a strong link between population-level influenza activity and increased IHD deaths, estimating approximately 300,000 influenza-attributable IHD deaths annually. Similarly, population-level RSV activity was significantly with increased hospitalizations, varying by age and country. The highest RSV-attributable hospitalization rates were in those 75 and older, ranging from 111 per 100,000 in Argentina (respiratory hospitalizations) to 1,200 per 100,000 in New Zealand (cardiorespiratory hospitalizations).

Conclusions: Traditional methods like the Serfling model and analyses of medical records are not suitable for reliably estimating the impact of viral diseases due to their widespread presence and limited testing. Using the CRA framework adapted from the GBD study, we establish exposure-response relationships while accounting for confounders like climate and co-circulating viruses. As viral surveillance expands, this approach offers diverse applications, from localized burden assessments to scalable global estimates through data extrapolation.