Cyril Geismar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Cyril Geismar has a total of 176 co-authors. Here are the top co-authors by number of shared publications:
- Robert William Aldridge
Centre for Public Health Data Science, Institute of Health Informatics, University College London, London, UK
39 shared publications - Sarah Beale
Centre for Public Health Data Science, Institute of Health Informatics, University College London, London, UK
36 shared publications - Ellen Fragaszy
Institute of Health Informatics, University College London, 222 Euston Road, London, NW1 2DA, United Kingdom, 44 02035495969
29 shared publications - Andrew Hayward
Institute of Epidemiology and Health Care, University College London, London, UK
27 shared publications - Annalan Mathew Dwight Navaratnam
IMS Epidemiology Unit, University of Cambridge School of Clinical Medicine, Cambridge CB2 0SL, United Kingdom
27 shared publications Jana Kovar
UCL, Inst Epidemiol & Hlth Care, London, England
25 shared publications
Latest Publications
- Open Access
- 27 October 2026
- Published by Springer Nature in Communications Medicine
Reconstruction of SARS-CoV-2 transmissibility within households in the UK Virus Watch Study
Background: Households provide an ideal setting to study SARS-CoV-2 transmission modes due to the close proximity and extended exposure among members. Understanding the patterns of who infects whom during infectious disease outbreaks is crucial for implementing effective control measures globally. Combining genomic data with epidemiological insights can help clarify and infer pathways of viral spread. Methods: We analyse SARS-CoV-2 whole genome sequences from 237 subjects across 162 households recruited to the Virus Watch Prospective Community Cohort Study ( n = 372 total participants). We consider minority variants as an indicator of within-host genomic diversity and incorporate these into a phylogenetic model to reconstruct the relationships between genomes, leading to a more accurate reconstruction of transmission than if only the consensus variation were used. Results: Here we show that of the 73 households with multiple infected participants, both consensus and within-host phylogenetic analyses identify eight households in which one or more secondary cases appear to arise from separate community introductions rather than within-household transmission. These households are excluded from the bottleneck and within-host transmission analyses. For the remaining 65 households, including minority variants when reconstructing phylogeny allows us to infer transmission chains that are otherwise uncertain with consensus variants only. Conclusions: Integrating within-host genetic diversity into a phylogenetic model alongside epidemiological data provides a robust approach to accurately identify true household transmission events, assess risk factors for infection acquisition and spread, and improve the estimation of the secondary attack rate
- 27 October 2026
- Published by Springer Nature in Communications Medicine
Open AccessReconstruction of SARS-CoV-2 transmissibility within households in the UK Virus Watch Study
Background: Households provide an ideal setting to study SARS-CoV-2 transmission modes due to the close proximity and extended exposure among members. Understanding the patterns of who infects whom during infectious disease outbreaks is crucial for implementing effective control measures globally. Combining genomic data with epidemiological insights can help clarify and infer pathways of viral spread. Methods: We analyse SARS-CoV-2 whole genome sequences from 237 subjects across 162 households recruited to the Virus Watch Prospective Community Cohort Study ( n = 372 total participants). We consider minority variants as an indicator of within-host genomic diversity and incorporate these into a phylogenetic model to reconstruct the relationships between genomes, leading to a more accurate reconstruction of transmission than if only the consensus variation were used. Results: Here we show that of the 73 households with multiple infected participants, both consensus and within-host phylogenetic analyses identify eight households in which one or more secondary cases appear to arise from separate community introductions rather than within-household transmission. These households are excluded from the bottleneck and within-host transmission analyses. For the remaining 65 households, including minority variants when reconstructing phylogeny allows us to infer transmission chains that are otherwise uncertain with consensus variants only. Conclusions: Integrating within-host genetic diversity into a phylogenetic model alongside epidemiological data provides a robust approach to accurately identify true household transmission events, assess risk factors for infection acquisition and spread, and improve the estimation of the secondary attack rate
- Open Access
- 1 June 2026
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 22 (6) , e1014271
A statistical framework for comparing epidemic forests
Inferring who infected whom in an outbreak is essential for characterising transmission dynamics and guiding public health interventions. However, this task is challenging due to limited surveillance data and the complexity of immunological and social interactions. Instead of a single definitive transmission tree, epidemiologists often consider multiple plausible trees forming epidemic forests . Various inference methods and assumptions can yield different epidemic forests, yet no formal test exists to assess whether these differences are statistically significant. We propose such a framework using a chi-square test and permutational multivariate analysis of variance (PERMANOVA). We assessed each method’s ability to distinguish simulated epidemic forests generated under different offspring distributions. While both methods achieved perfect specificity for forests with 100+ trees, PERMANOVA consistently outperformed the chi-square test in sensitivity across all epidemic and forest sizes. Implemented in the R package mixtree , we provide the first statistical framework to robustly compare epidemic forests.
- 1 June 2026
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 22 (6) , e1014271
Open AccessA statistical framework for comparing epidemic forests
Inferring who infected whom in an outbreak is essential for characterising transmission dynamics and guiding public health interventions. However, this task is challenging due to limited surveillance data and the complexity of immunological and social interactions. Instead of a single definitive transmission tree, epidemiologists often consider multiple plausible trees forming epidemic forests . Various inference methods and assumptions can yield different epidemic forests, yet no formal test exists to assess whether these differences are statistically significant. We propose such a framework using a chi-square test and permutational multivariate analysis of variance (PERMANOVA). We assessed each method’s ability to distinguish simulated epidemic forests generated under different offspring distributions. While both methods achieved perfect specificity for forests with 100+ trees, PERMANOVA consistently outperformed the chi-square test in sensitivity across all epidemic and forest sizes. Implemented in the R package mixtree , we provide the first statistical framework to robustly compare epidemic forests.
- Open Access
- 15 December 2025
- Published by The R Foundation
linktree: Estimate Transmission Assortativity Coefficients
Estimates group transmission assortativity coefficients from transmission trees. Group transmission assortativity coefficients measure the tendency for individuals to transmit within their own group (e.g. age group, vaccination status, or location) compared to other groups. The package requires information on who infected whom, group membership for all individuals, and the relative sizes of each group in the population. For more details see Geismar et al. (2024) .
- 15 December 2025
- Published by The R Foundation
Open Accesslinktree: Estimate Transmission Assortativity Coefficients
Estimates group transmission assortativity coefficients from transmission trees. Group transmission assortativity coefficients measure the tendency for individuals to transmit within their own group (e.g. age group, vaccination status, or location) compared to other groups. The package requires information on who infected whom, group membership for all individuals, and the relative sizes of each group in the population. For more details see Geismar et al. (2024) .
- Open Access
- 23 October 2025
- Published by Springer Nature in Research Square
Reconstruction of SARS-CoV-2 transmissibility within households in the UK Virus Watch Study
Households provide ideal settings to study SARS-CoV-2 transmission due to close proximity and extended exposure among members. Understanding transmission patterns is crucial for implementing effective control measures. We analysed whole genome sequences from 237 subjects across 162 households within the Virus Watch Prospective Community Cohort Study (372 total participants). We incorporated minority variants as indicators of within-host genomic diversity into phylogenetic models to reconstruct viral relationships more accurately than using consensus sequences alone. In 73/162 households with multiple infections, phylogeny identified eight secondary cases resulting from separate introduction events rather than within-household transmission. For the remaining 65 households, including minority variants resolved transmission chains that were otherwise uncertain. Our approach demonstrates that integrating within-host genetic diversity into phylogenetic models alongside epidemiological data provides a robust method to accurately identify household transmission events, assess infection risk factors, and improve secondary attack rate estimation, essential for understanding viral spread dynamics.
- 23 October 2025
- Published by Springer Nature in Research Square
Open AccessReconstruction of SARS-CoV-2 transmissibility within households in the UK Virus Watch Study
Households provide ideal settings to study SARS-CoV-2 transmission due to close proximity and extended exposure among members. Understanding transmission patterns is crucial for implementing effective control measures. We analysed whole genome sequences from 237 subjects across 162 households within the Virus Watch Prospective Community Cohort Study (372 total participants). We incorporated minority variants as indicators of within-host genomic diversity into phylogenetic models to reconstruct viral relationships more accurately than using consensus sequences alone. In 73/162 households with multiple infections, phylogeny identified eight secondary cases resulting from separate introduction events rather than within-household transmission. For the remaining 65 households, including minority variants resolved transmission chains that were otherwise uncertain. Our approach demonstrates that integrating within-host genetic diversity into phylogenetic models alongside epidemiological data provides a robust method to accurately identify household transmission events, assess infection risk factors, and improve secondary attack rate estimation, essential for understanding viral spread dynamics.
- Open Access
- 10 October 2025
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 11 (1) , e85512
Correction: Comparison of Virus Watch COVID-19 Positivity, Incidence, and Hospitalization Rates With Other Surveillance Systems: Surveillance Study
- 10 October 2025
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 11 (1) , e85512
Open AccessCorrection: Comparison of Virus Watch COVID-19 Positivity, Incidence, and Hospitalization Rates With Other Surveillance Systems: Surveillance Study
- 8 October 2025
- Published by The R Foundation
pipetime: Tools to Time Pipe Operations
Enable users to measure and record the execution time of pipe operations (using |>) with optional logging to dataframes and output to the console.
- 8 October 2025
- Published by The R Foundation
pipetime: Tools to Time Pipe Operations
Enable users to measure and record the execution time of pipe operations (using |>) with optional logging to dataframes and output to the console.
- Open Access
- 29 September 2025
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 11, e69655
Comparison of Virus Watch COVID-19 Positivity, Incidence, and Hospitalization Rates With Other Surveillance Systems: Surveillance Study
Background: Effective disease surveillance is essential for understanding pathogens’ epidemiology, detecting outbreaks, and enabling timely public health responses. In the United Kingdom, large-scale studies, such as the Office for National Statistics COVID-19 Infection Survey (CIS), have monitored SARS-CoV-2 transmission but required significant resources, making them challenging to sustain when pandemic-specific funding ends and also in resource-limited settings. In contrast, the Virus Watch study, at lower cost, relied on self-reported and linked national testing data as well as symptomatic testing, while Severe Acute Respiratory Infections Watch (SARI) leveraged hospital data for cost-effective surveillance. Objective: This study aimed to evaluate the effectiveness of Virus Watch as a surveillance system in monitoring COVID-19 positivity, incidence, and hospitalization rates in England and Wales, using data from the CIS and SARI as benchmarks for comparison, while considering the key differences in the study designs, including recruitment strategies, incentives, and testing criteria. Methods: We used the Virus Watch prospective community cohort study to estimate COVID-19 positivity, incidence, and hospitalization rates in England and Wales from June 2020 to March 2023. Rate estimates were compared with CIS modeled positivity and incidence rates, and with SARI COVID-19 hospitalization rates. Global synchrony between datasets was measured using overall Spearman ⍴ and local synchrony using 9-week rolling Spearman ⍴. For England, comparisons with CIS estimates used Virus Watch rates calculated with and without linked national testing data. Positivity rates were also assessed overall and separately before and after the end of free national testing. Results: A total of 58,628 participants were recruited into the Virus Watch study, of whom 52,526 (89.6%) were resident in England and 1532 (2.6%) in Wales; region was missing for the remainder. Virus Watch–estimated COVID-19 positivity and incidence rates in England, calculated with and without linked testing data, showed strong global synchrony with CIS estimates (positivity ⍴: 0.91 and 0.90; both P <.001 and incidence ⍴: 0.92 and 0.90; both P <.001) and strong local synchrony (positivity ⍴: median 0.75, IQR 0.53‐0.85 and median 0.67, IQR 0.47‐0.83, and incidence ⍴: median 0.76, IQR 0.49‐0.88 and median 0.66, IQR 0.45‐0.82), despite having lower absolute values. Global and local synchrony of positivity rates were similar for periods before and after the end of free national testing, although the difference between Virus Watch and CIS estimates was greater post–free testing. COVID-19 hospitalization rates were also lower and less synchronized with SARI estimates. In Wales, Virus Watch estimates exhibited greater variability (positivity ρ: 0.75, P <.001; incidence rate ρ: 0.85, P <.001) and lower local synchrony (positivity ρ: median 0.61, IQR 0.34‐0.74, and incidence ρ: median 0.52, IQR 0.38‐0.71) compared to England. Conclusions: Our results highlight the effectiveness of the Virus Watch approach in providing accurate estimates of COVID-19 positivity and incidence rates, even in the absence of national surveillance systems. This low-cost method can be adapted to various settings, particularly low-resource ones, to strengthen public health surveillance and inform timely interventions.
- 29 September 2025
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 11, e69655
Open AccessComparison of Virus Watch COVID-19 Positivity, Incidence, and Hospitalization Rates With Other Surveillance Systems: Surveillance Study
Background: Effective disease surveillance is essential for understanding pathogens’ epidemiology, detecting outbreaks, and enabling timely public health responses. In the United Kingdom, large-scale studies, such as the Office for National Statistics COVID-19 Infection Survey (CIS), have monitored SARS-CoV-2 transmission but required significant resources, making them challenging to sustain when pandemic-specific funding ends and also in resource-limited settings. In contrast, the Virus Watch study, at lower cost, relied on self-reported and linked national testing data as well as symptomatic testing, while Severe Acute Respiratory Infections Watch (SARI) leveraged hospital data for cost-effective surveillance. Objective: This study aimed to evaluate the effectiveness of Virus Watch as a surveillance system in monitoring COVID-19 positivity, incidence, and hospitalization rates in England and Wales, using data from the CIS and SARI as benchmarks for comparison, while considering the key differences in the study designs, including recruitment strategies, incentives, and testing criteria. Methods: We used the Virus Watch prospective community cohort study to estimate COVID-19 positivity, incidence, and hospitalization rates in England and Wales from June 2020 to March 2023. Rate estimates were compared with CIS modeled positivity and incidence rates, and with SARI COVID-19 hospitalization rates. Global synchrony between datasets was measured using overall Spearman ⍴ and local synchrony using 9-week rolling Spearman ⍴. For England, comparisons with CIS estimates used Virus Watch rates calculated with and without linked national testing data. Positivity rates were also assessed overall and separately before and after the end of free national testing. Results: A total of 58,628 participants were recruited into the Virus Watch study, of whom 52,526 (89.6%) were resident in England and 1532 (2.6%) in Wales; region was missing for the remainder. Virus Watch–estimated COVID-19 positivity and incidence rates in England, calculated with and without linked testing data, showed strong global synchrony with CIS estimates (positivity ⍴: 0.91 and 0.90; both P <.001 and incidence ⍴: 0.92 and 0.90; both P <.001) and strong local synchrony (positivity ⍴: median 0.75, IQR 0.53‐0.85 and median 0.67, IQR 0.47‐0.83, and incidence ⍴: median 0.76, IQR 0.49‐0.88 and median 0.66, IQR 0.45‐0.82), despite having lower absolute values. Global and local synchrony of positivity rates were similar for periods before and after the end of free national testing, although the difference between Virus Watch and CIS estimates was greater post–free testing. COVID-19 hospitalization rates were also lower and less synchronized with SARI estimates. In Wales, Virus Watch estimates exhibited greater variability (positivity ρ: 0.75, P <.001; incidence rate ρ: 0.85, P <.001) and lower local synchrony (positivity ρ: median 0.61, IQR 0.34‐0.74, and incidence ρ: median 0.52, IQR 0.38‐0.71) compared to England. Conclusions: Our results highlight the effectiveness of the Virus Watch approach in providing accurate estimates of COVID-19 positivity and incidence rates, even in the absence of national surveillance systems. This low-cost method can be adapted to various settings, particularly low-resource ones, to strengthen public health surveillance and inform timely interventions.
- Open Access
Reconstruction of SARS-CoV-2 transmissibility within households in the UK Virus Watch Study
Households provide ideal settings to study SARS-CoV-2 transmission due to close proximity and extended exposure among members. Understanding transmission patterns is crucial for implementing effective control measures. We analysed whole genome sequences from 237 subjects across 162 households within the Virus Watch Prospective Community Cohort Study (372 total participants). We incorporated minority variants as indicators of within-host genomic diversity into phylogenetic models to reconstruct viral relationships more accurately than using consensus sequences alone. In 73/162 households with multiple infections, phylogeny identified eight secondary cases resulting from separate introduction events rather than within-household transmission. For the remaining 65 households, including minority variants resolved transmission chains that were otherwise uncertain. Our approach demonstrates that integrating within-host genetic diversity into phylogenetic models alongside epidemiological data provides a robust method to accurately identify household transmission events, assess infection risk factors, and improve secondary attack rate estimation, essential for understanding viral spread dynamics.
Open AccessReconstruction of SARS-CoV-2 transmissibility within households in the UK Virus Watch Study
Households provide ideal settings to study SARS-CoV-2 transmission due to close proximity and extended exposure among members. Understanding transmission patterns is crucial for implementing effective control measures. We analysed whole genome sequences from 237 subjects across 162 households within the Virus Watch Prospective Community Cohort Study (372 total participants). We incorporated minority variants as indicators of within-host genomic diversity into phylogenetic models to reconstruct viral relationships more accurately than using consensus sequences alone. In 73/162 households with multiple infections, phylogeny identified eight secondary cases resulting from separate introduction events rather than within-household transmission. For the remaining 65 households, including minority variants resolved transmission chains that were otherwise uncertain. Our approach demonstrates that integrating within-host genetic diversity into phylogenetic models alongside epidemiological data provides a robust method to accurately identify household transmission events, assess infection risk factors, and improve secondary attack rate estimation, essential for understanding viral spread dynamics.
- 5 March 2025
- Published by The R Foundation
mixtree: A Statistical Framework for Comparing Sets of Trees
Statistical framework for comparing sets of trees using hypothesis testing methods. Designed for transmission trees, phylogenetic trees, and directed acyclic graphs (DAGs), the package implements chi-squared tests to compare edge frequencies between sets and PERMANOVA to analyse topological dissimilarities with customisable distance metrics, following Anderson (2001) .
- 5 March 2025
- Published by The R Foundation
mixtree: A Statistical Framework for Comparing Sets of Trees
Statistical framework for comparing sets of trees using hypothesis testing methods. Designed for transmission trees, phylogenetic trees, and directed acyclic graphs (DAGs), the package implements chi-squared tests to compare edge frequencies between sets and PERMANOVA to analyse topological dissimilarities with customisable distance metrics, following Anderson (2001) .
- Open Access
- 2 December 2024
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 19 (12) , e0313037
Sorting out assortativity: When can we assess the contributions of different population groups to epidemic transmission?
Characterising the transmission dynamics between various population groups is critical for implementing effective outbreak control measures whilst minimising financial costs and societal disruption. While recent technological and methodological advances have made individual-level transmission chain data increasingly available, it remains unclear how effectively this data can inform group-level transmission patterns, particularly in small, rapidly saturating outbreak settings. We introduce a novel framework that leverages transmission chain data to estimate group transmission assortativity; this quantifies the extent to which individuals transmit within their own group compared to others. Through extensive simulations mimicking nosocomial outbreaks, we assessed the conditions under which our estimator performs effectively and established guidelines for minimal data requirements in small outbreak settings where saturation may occur rapidly. Notably, we demonstrate that detecting and quantifying transmission assortativity is most reliable when at least 30 cases have been observed in each group, before reaching their respective epidemic peaks.
- 2 December 2024
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 19 (12) , e0313037
Open AccessSorting out assortativity: When can we assess the contributions of different population groups to epidemic transmission?
Characterising the transmission dynamics between various population groups is critical for implementing effective outbreak control measures whilst minimising financial costs and societal disruption. While recent technological and methodological advances have made individual-level transmission chain data increasingly available, it remains unclear how effectively this data can inform group-level transmission patterns, particularly in small, rapidly saturating outbreak settings. We introduce a novel framework that leverages transmission chain data to estimate group transmission assortativity; this quantifies the extent to which individuals transmit within their own group compared to others. Through extensive simulations mimicking nosocomial outbreaks, we assessed the conditions under which our estimator performs effectively and established guidelines for minimal data requirements in small outbreak settings where saturation may occur rapidly. Notably, we demonstrate that detecting and quantifying transmission assortativity is most reliable when at least 30 cases have been observed in each group, before reaching their respective epidemic peaks.