SARS-CoV-2 viral shedding and transmission during COVID-19 pandemic in Hong Kong

Abstract

The COVID-19 pandemic has highlighted the importance of leveraging surveillance data, especially viral testing data from infected individuals, to enhance the understanding of viral-host interactions, monitor epidemic dynamics and optimize control measures. Despite extensive research, significant knowledge gaps remain regarding how SARS-CoV-2 viral kinetics vary across individuals with different characteristics and how patterns of viral load in infected populations can be utilized to track real-time transmission across diverse epidemiological contexts.

By analyzing comprehensive viral testing data from COVID-19 patients in Hong Kong to model SARS-CoV-2 viral shedding kinetics from symptom onset, more clinically relevant insights were provided into how viral kinetics varied with patient demographics including age, disease severity, and vaccination status. Older patients and those with severe outcomes exhibited higher viral burdens for prolonged periods, though their viral burdens were not elevated soon after symptom onset. This suggested that these groups of patients might contribute differently to transmission and implied heterogeneity in immunological responses. Data from the largely infection-naïve population in Hong Kong prior to the Omicron wave along with detailed vaccination records enabled comparisons between infections caused by the ancestral strain and Omicron BA.2. Results revealed minor differences in intrinsic shedding kinetics between the two strains particularly during early days after onset, meanwhile emphasizing the substantial impact of COVID-19 vaccination in reducing viral burden even against the immune-escaping Omicron BA.2 variant.

The virological effect of oral antivirals, including molnupiravir and nirmatrelvir-ritonavir, was also evaluated, demonstrating significantly shorter shedding durations and lower viral burden associated with the treatments, particularly when nirmatrelvir-ritonavir was administered early. These findings support the role of antivirals in improving clinical outcomes and potentially lowering transmission risks. Notably, more pronounced effect of nirmatrelvir-ritonavir was found in unvaccinated than vaccinated individuals, suggesting complex interactions between vaccination, antiviral and viral kinetics. These findings emphasize the importance of initiating antiviral treatment early and consideration of patients’ immune background in optimizing the allocation and prioritization of antivirals.

Beyond individual-level shedding dynamics, the use of population-level viral load distributions was also explored as a novel tool for real-time epidemic monitoring. Analyses of early ancestral strain waves in Hong Kong demonstrated a strong correlation between daily population viral load distributions and COVID-19 transmission risks. This enabled accurate estimation of the effective reproduction number using viral load data obtained from syndromic and targeted surveillance, providing a timely and practical method for tracking transmission even in resource-limited settings. Systematic evaluations across different surveillance scenarios and pathogen characteristics further highlighted the adaptability of this method and identified key factors influencing its performance.

Findings presented in this thesis offer valuable insights into SARS-CoV-2 viral kinetics and their associations with host, viral and intervention-related factors, provide evidence on the effectiveness of vaccination and antivirals to guide resource allocation and prioritization, and demonstrate the feasibility of using population-level viral load data for real-time monitoring on the virus transmission in the community. Collectively, my research on SARS-CoV-2 highlight the important role of detailed surveillance data in enhancing pandemic preparedness and informing effective public health responses to infectious diseases.