Genomics, social media and mobile phone data enable mapping of SARS-CoV-2 lineages to inform health policy in Bangladesh

The COVID-19 pandemic underscored the need for high-resolution data on SARS-CoV-2 spread, particularly in low- and middle-income countries (LMICs). Genomic surveillance, combined with population mobility data, has proven effective in high-income countries for implementing localized control measures. Bangladesh, an LMIC with a large and mobile population, faced significant challenges in controlling the pandemic. This study aimed to analyze the emergence and spread of SARS-CoV-2 lineages in Bangladesh by combining genomic data from the Institute of Epidemiology Disease Control and Research (IEDCR) and publicly available GISAID data with population mobility data from Facebook and mobile phone operators. Understanding the spread patterns was crucial for informing national health policies and optimizing intervention strategies. The study's importance lies in its potential to provide valuable insights into pandemic response in LMICs, where resources and infrastructure may be limited compared to high-income countries. The integration of multiple data sources offers a novel approach to track the complex dynamics of viral transmission and inform targeted public health interventions.

Existing literature demonstrates the effectiveness of combining genomic surveillance with population mobility data for tracking SARS-CoV-2 spread in high-income countries. Studies have shown how this approach enables the implementation of effective local control measures. However, similar research in LMICs like Bangladesh is limited. The emergence of SARS-CoV-2 genomic analyses from various countries has also driven the development of vaccines, emphasizing the critical role of genomic surveillance. While genomic surveillance is common in high-income countries, its application in LMICs remains crucial for guiding local health policies. This study builds upon previous research by using multiple sources of data to understand the spread in a low-income context where challenges regarding resource availability and data accessibility are significant.

This study employed a multi-faceted approach combining viral genomics and population mobility data. First, the researchers sequenced the complete genomes of 67 SARS-CoV-2 samples collected by the IEDCR in Bangladesh between March and July 2020. These data were combined with 324 publicly available GISAID SARS-CoV-2 genomes from Bangladesh during the same period. Phylogenetic analysis, including Bayesian time-scale phylogenetic analysis using BEAST, was conducted to determine the evolutionary relationships between the isolates, estimate the time of introduction into Bangladesh, and track the spread of different lineages. Second, population mobility data were obtained from two sources: Facebook's 'Data for Good' initiative, providing anonymized and aggregated data on user movements, and mobile phone operators, providing data on subscriber location changes. These datasets, beginning from March 2020, captured the movement of a large portion of the Bangladeshi population and allowed researchers to assess the impact of mass migration on the spread of SARS-CoV-2. Third, additional 85 genomes were sequenced between November 2020 and April 2021 to assess the emergence of Variants of Concern (VOCs) like Beta. The results were analyzed to connect the genomic evolution with population movement patterns to understand the role of mass migration in shaping the outbreak trajectory. Specifically, Nanopore MinION sequencing was used for genomic analysis, using established pipelines. The software used includes ARTIC-cg/biome-fpopulateSKP and Nextstrain. For phylogeographic analysis, TempEst and BEAST software were used. The statistical methods used for phylogenetic analysis involved building maximum likelihood and Bayesian time-scaled phylogenetic trees. Data analysis was performed with consideration for missing data, and confidence intervals were provided for mobile phone data analyses.

The study identified three dominant SARS-CoV-2 lineages (B.1.125, B.1.1, and B.1.36) circulating in Bangladesh during early 2020. Phylogenetic analysis suggested that these lineages were introduced from abroad, with B.1.125 being imported at least two weeks before the expansion of international air travel restrictions on March 21st. The analysis indicated that SARS-CoV-2 likely first emerged in mid-February in Bangladesh. Crucially, the mass migration from Dhaka at the end of March, coinciding with a national holiday, was strongly correlated with the rapid dispersal of these lineages across the country. Mobility data from Facebook and mobile phone operators confirmed this mass migration. Analysis of additional samples (November 2020-April 2021) showed that the Beta variant (B.1.351) was subsequently introduced and became dominant in Dhaka. This highlights the risk posed by mass migration and travel from urban hotspots to rural areas. The findings from the study emphasized the role of mass migration from Dhaka as the main driver for countrywide spread in 2020. Furthermore, the emergence of Beta VOC and its dominance in 2021 was identified as a new threat. The study clearly shows the importance of real-time genomic surveillance, especially in LMICs, for detecting and containing emerging variants. The combination of genomic data and population mobility data enabled detailed insights into the spread of SARS-CoV-2, and these findings have direct implications for informing government policies. This includes limiting within-country travel to control national case numbers.

This study demonstrates the value of combining genomic and population mobility data for understanding SARS-CoV-2 transmission dynamics, particularly in LMICs like Bangladesh. The findings highlight the interconnectedness of international travel, mass migration, and the spread of viral lineages. The rapid dissemination of SARS-CoV-2 lineages across Bangladesh following the national holiday underscores the challenges posed by population mobility in controlling pandemic outbreaks. The study’s strength is its multi-faceted approach, integrating genomics, social media data, and mobile phone data, generating valuable insights beyond what any single data source could reveal. The identification of the Beta variant and its rapid dominance in Dhaka serves as a case study for the need for continued genomic surveillance to monitor the emergence and spread of VOCs. The study’s findings have clear implications for public health policies, suggesting that strategies that focus on limiting travel and movement during outbreak periods might be more effective in LMICs. Future research could explore how these insights can be leveraged to enhance pandemic preparedness and response strategies in other LMICs.

This study's findings highlight the importance of combining genomic surveillance with population mobility data for effective pandemic response in LMICs. The study's methodology, integrating diverse data sources, provides a robust framework for understanding SARS-CoV-2 transmission dynamics. The identification of key lineages and variants, coupled with analyses of population mobility, enables informed policy interventions. The results provide essential insights into the challenges faced in controlling outbreaks in densely populated and highly mobile settings. Further research should explore the effectiveness of tailored public health interventions in LMICs, considering the unique challenges posed by mass migration and limited resources.

The study's reliance on data from specific sources (Facebook and select mobile phone operators) might limit the generalizability of findings. The representativeness of the sample locations for the entire country needs further consideration. While the study provides valuable insights into SARS-CoV-2 spread during specific periods, the evolution of the pandemic is ongoing, necessitating continuous monitoring. Further limitations include the availability of genomic data at the time and the limitations of the statistical methods used.