Abstract
BACKGROUND: Disability Weights (DWs) are crucial for assessing disease burden guiding public health decision-making. For emerging health threats such as COVID-19, the absence of relevant survey data from China has led to reliance on established DW values for specific symptoms in calculating the COVID-19 disease burden. However, these values have not been updated in real-time to reflect the ongoing mutations of the virus, potentially skewing the longitudinal estimation of COVID-19's burden and compromising the accuracy of public health interventions. METHODS: This study developed a real-time estimation framework using longitudinal internet survey data to track changes in DW distributions across different populations over time. These distributions were integrated into Monte Carlo simulations to model real-time disease burden, offering robust data to support evidence-based policy decisions and optimize resource allocation. RESULTS: Our analysis revealed substantial variation in DW distributions across symptoms. As populations experience multiple infections and the virus evolves, the COVID-19 disease burden has converged with, and in some cases fallen below, that of influenza's. Survey data suggested an average immunity interval of approximately five months between infections. Moreover, COVID-19 has profoundly reshaped healthcare-seeking behavior and consumption patterns, with individual lifestyle factors and pre-existing health conditions contributing significantly to infection severity. CONCLUSION: The real-time DW estimation method proposed in this study effectively and accurately reflects the dynamic changes in the COVID-19 disease burden amidst ongoing virus mutations, providing crucial reference data for the evaluation and formulation of public health policies. Furthermore, the study provides insights into the transmission interval of COVID-19 and behavioral changes during the pandemic, offering valuable insights for the potential outbreak of future "Disease X."