Write a comment

This review reflects comments and contributions from Femi Arogundade.

The study investigates factors influencing SARS-CoV-2 disease severity in Georgia, USA, considering vaccination status, viral variants, and clinical factors. Utilizing comprehensive clinical, demographic, and genomic data, the research emphasizes the critical role of vaccination, age, and underlying health conditions in determining disease outcomes. Analyzing data from 1,957 individuals, the study finds that age, underlying health conditions, and vaccination status significantly impact disease outcomes, with vaccinations offering consistent protection against severe outcomes for both Delta and Omicron variants. Despite limitations, the study offers valuable insights, particularly in a region with low vaccination rates, contributing to the broader understanding of SARS-CoV-2 dynamics. The methods are well-structured, ensuring transparency and ethical conduct, with rigorous statistical analyses supporting the robustness of the findings.

General Comments:

Positive Aspects of the Paper

• The study employs a comprehensive approach by integrating clinical, demographic, and genomic data, providing a holistic understanding of SARS-CoV-2 disease severity.

• The commitment to data transparency is commendable, with all sequence data available in NCBI and GISAID, fostering scientific collaboration and scrutiny.

• The study prioritizes ethical standards with institutional review board approval and a waiver of consent, ensuring privacy and adherence to ethical guidelines.

• The use of rigorous statistical methods, including multiple imputation for missing data and multinomial logistic regression for disease severity analysis, enhances the robustness of the findings.

• The detailed genomic analysis, including phylogenetic studies and viral metagenomic analysis, contributes valuable insights into the dynamics of viral variants and co-infections.

• The focus on a region with low vaccination rates, such as Georgia, provides real-world insights into the impact of vaccination in a population with unique demographic and healthcare characteristics.

• Definitions of variables, such as vaccination status and disease severity, are well-specified, contributing to the clarity and replicability of the study.

• The study transparently acknowledges limitations, such as sample bias and the retrospective design, providing a balanced interpretation of the findings.

• The study's emphasis on demographic and clinical factors influencing disease severity, alongside vaccination, offers practical implications for public health messaging and interventions.

• The study contributes valuable knowledge to the evolving understanding of SARS-CoV-2 dynamics, particularly in regions with low vaccine uptake and emerging variants.

Minor Comments:

• The integration of data from multiple sources, including electronic medical records and state databases, contributes to the richness and completeness of the dataset.

• The genomic analysis, including sequencing methods, phylogenetic studies, and viral metagenomic analysis, demonstrates a thorough and systematic approach to understanding viral dynamics.

• While the sequencing methods are well-described, additional details on the validation and accuracy of these methods would further strengthen the experimental aspect of the study.

• The viral metagenomic analysis provides valuable insights into co-infections, adding depth to the study's virological understanding.

• While the study focuses on Georgia, the findings may have broader implications, especially for regions with similar demographic and vaccination characteristics.

Major Comments:

• Include additional details on the validation and quality control measures employed in the sequencing process, enhancing transparency and confidence in the genomic analysis.

• How representative is the study population of SARS-CoV-2-positive individuals at Emory University Hospital to the broader population, especially considering the exclusion of partially vaccinated individuals and those reporting out-of-state residency?

• The reliance on data from multiple sources (GRITS, SENDSS, EMR) raises questions about data consistency and potential discrepancies. How were data discrepancies or missing information handled during the analysis?

• The waiver of consent should be discussed in terms of ethical implications. How was patient privacy and confidentiality ensured, especially considering the use of electronic medical records? Clearly articulate the rationale for obtaining a waiver of consent. This might include considerations of minimal risk to participants, impracticality of obtaining consent, or how obtaining consent might introduce biases to the study.

• The WHO clinical progression scale is utilized for defining disease severity. How consistently is this scale applied across healthcare settings, and is there a potential for variability in its interpretation?

• The use of a case-match basis from May 2021 to September 2021 and the inclusion of all positive individuals from October 2021 to May 2022 may introduce bias. How does this potential bias impact the generalizability of the findings? How did the study account for potential confounding factors in the analysis?

• The study acknowledges potential residual confounding due to its retrospective design. How confident are the authors that the adjustments made for variables such as age, demographics, pre-existing health conditions, and vaccination status adequately addressed potential confounders?

• The study spans a period from May 2021 to May 2022. How did the potential seasonality of SARS-CoV-2, especially during the emergence of new variants, factor into the analysis and interpretation of the results?

• The study focuses on individuals who sought medical care, potentially skewing the population toward those with more severe diseases. How might this affect the generalizability of the findings to the broader population, including those with mild or asymptomatic cases?

• The study uses the Nextstrain web-based mutation calling tool for identifying SNPs and insertions/deletions. How robust is this tool, and were there any challenges or limitations encountered in accurately identifying mutations, particularly in the context of evolving variants?

• The study employs the Nextstrain SARS-CoV-2 Workflow for phylogenetic analysis. Could the authors elaborate on the specific parameters and considerations in the custom scheme used for subsampling, and how might this affect the representativeness of the selected sequences in the analysis?

• The authors subsampled the dataset to select sequences most genetically similar to their dataset. Could this introduce a bias, and how confident are the authors that the subsampling accurately represents the diversity and dynamics of SARS-CoV-2 sequences during the specified period?

• Were there any validation methods or external benchmarks employed to assess the robustness and accuracy of the phylogenetic findings, especially concerning the assignment of sequences to specific lineages and their implications for the study's conclusions?

• The paper aligns sequences with reference strains from Wuhan. Given the ongoing evolution of the virus, how might the use of these reference strains impact the interpretation of genetic changes, especially in the context of emerging variants?

• The paper mentions adjusting for significance with Bracken after the taxonomic assignment. Could the authors provide more details on the parameters and criteria used for this adjustment? How might this influence the identification of viral co-infections, and were any sensitivity analyses conducted to assess its impact?

• The minlen parameter during read trimming is set to 36. How might this impact the sensitivity of detecting viral co-infections, especially if there are short viral fragments present in the samples?

Comments on Statistical Analysis:

• While interpreting the results of the multivariable models, how confident are the authors in the causal inferences, considering the observational nature of the study? Were any causal mediation analyses or sensitivity analyses performed to explore potential mechanisms underlying the observed associations?

• Considering the study population and its characteristics, to what extent can the statistical findings be generalized to other populations with different demographic or healthcare access profiles? What limitations or considerations should be kept in mind when applying these findings to broader contexts?

• The study mentions significant differences between qRT-PCR testing platforms, necessitating control as a covariate. Could the authors elaborate on the nature of these differences and their potential impact on the results? How confident are the authors that controlling for this covariate adequately addresses any biases introduced by platform variations?

• While the use of multinomial logistic regressions to test the association with disease severity is appropriate, were there considerations for potential confounding factors that might influence the observed associations in the step-wise construction of multivariable models?

• Suggestions for Future Studies:

• Investigate the immune response dynamics among different age groups and individuals with various underlying health conditions. Understanding the nuances of immune responses can contribute to personalized vaccination strategies.

• Explore the impact of prior SARS-CoV-2 infections on disease severity and immune responses. This could provide a more comprehensive understanding of the interplay between natural immunity and vaccination.

• Investigate the role of specific genetic factors in influencing disease severity. A genetic association study might reveal insights into why certain individuals, despite being vaccinated, experience more severe outcomes.

• Conduct a longitudinal analysis to explore the durability of vaccine efficacy over an extended period. This could provide insights into the waning immunity and guide recommendations for booster doses.

• Investigate the role of specific genetic factors in influencing disease severity. A genetic association study might reveal insights into why certain individuals, despite being vaccinated, experience more severe outcomes.

• Explore the impact of social and behavioural factors on vaccination uptake. Understanding the barriers to vaccination and tailoring interventions can improve overall vaccine coverage.

• Assess the comparative effectiveness of different SARS-CoV-2 vaccines, considering variations in vaccine types, platforms, and formulations. This can guide vaccination strategies and inform future vaccine development.

• Integrate social determinants of health into the analysis to understand how factors such as socioeconomic status and access to healthcare influence disease outcomes and vaccination disparities.

• Conduct a comparative analysis of different SARS-CoV-2 sequencing methods to evaluate their sensitivity, specificity, and cost-effectiveness. This can inform best practices for genomic surveillance.

• Foster international collaboration for data sharing and analysis. Global datasets can provide a more holistic view of the pandemic, enabling a better-informed response to emerging challenges.

Competing interests

The author declares that they have no competing interests.