Home > Research > Publications & Outputs > Seasonal influenza in medical students
View graph of relations

Seasonal influenza in medical students: an outbreak simulation model based on a social network approach

Research output: Contribution to journalJournal article

Published

Standard

Seasonal influenza in medical students : an outbreak simulation model based on a social network approach. / Edge, Rhiannon; Heath, Joseph; Rowlingson, Barry; Keegan, Thomas; Isba, Rachel.

In: Lancet, Vol. 384, No. Supplement 2, S29, 19.11.2014.

Research output: Contribution to journalJournal article

Harvard

APA

Vancouver

Author

Bibtex

@article{946f5ff415bd46c99e8e312370ae284e,
title = "Seasonal influenza in medical students: an outbreak simulation model based on a social network approach",
abstract = "BackgroundThere is increasing interest in the effects of social networks on disease dynamics. We simulated the spread of influenza through a population of medical students where transmission was related to the social network structure and vaccination status of individuals.MethodsAll students at Lancaster Medical School, Lancaster, UK (n=253) were asked to rate the strength of their relationship with all other students from the medical school. Students also self-reported their influenza vaccination status. An individual-based outbreak model was developed using R statistical software. Using these data, combined with appropriate transmission parameters, we simulated an influenza outbreak and assessed the effects of preferentially vaccinating according to the social network analysis data. We ran the simulation 1500 times. Each simulation selected a random student to introduce the virus into the population. For each vaccination strategy, the likelihood of each individual being infected was calculated as a percentage of the number of times they were infected in the 1500 possible outbreaks.Findings215 students (85%) responded. Non-responders were assumed to have reciprocal relationships with responders; therefore it was possible to construct the entire medical student network. We found that the outcomes of vaccination strategies based on between-ness (the extent to which an individual lies between others in the network) and degree (the number of connections an individual has), which are both measures of connectedness, quickly converged. As more individuals were vaccinated, the likelihood of individuals contracting the infection tended to be similar, irrespective of vaccination based on between-ness or degree. After vaccination of an additional 8% of the population (20 students) the outcome of the experimental influenza outbreak was similar for both strategies.InterpretationOur results add to a small pool of evidence supporting targeting vaccination of individuals according to between-ness in an attempt to reduce the spread of influenza. However, in small, densely connected populations, vaccination according to degree might be preferential because of the rapid convergence and the relative ease of locating individuals with a high degree versus those with high between-ness. This study suggests that vaccination strategies that target highly connected individuals within a network might limit spread of infectious disease. Future work could include evaluating current vaccination approaches using social network analysis.FundingUniversity Hospitals of Morecambe Bay NHS Foundation Trust funded data collection.ContributorsRI conceived the programme of research of which this was a part. JH collected the data. RE was responsible for input and analysis of the data, supervised by RI and TK. RE developed the model and BR developed the software. RE wrote the abstract with input from RI and TK. All authors have approved the original version of the abstract for publication.Declaration of interestsWe declare no competing interests.",
author = "Rhiannon Edge and Joseph Heath and Barry Rowlingson and Thomas Keegan and Rachel Isba",
year = "2014",
month = nov
day = "19",
doi = "10.1016/S0140-6736(14)62155-3",
language = "English",
volume = "384",
journal = "The Lancet",
issn = "0140-6736",
publisher = "Lancet Publishing Group",
number = "Supplement 2",

}

RIS

TY - JOUR

T1 - Seasonal influenza in medical students

T2 - an outbreak simulation model based on a social network approach

AU - Edge, Rhiannon

AU - Heath, Joseph

AU - Rowlingson, Barry

AU - Keegan, Thomas

AU - Isba, Rachel

PY - 2014/11/19

Y1 - 2014/11/19

N2 - BackgroundThere is increasing interest in the effects of social networks on disease dynamics. We simulated the spread of influenza through a population of medical students where transmission was related to the social network structure and vaccination status of individuals.MethodsAll students at Lancaster Medical School, Lancaster, UK (n=253) were asked to rate the strength of their relationship with all other students from the medical school. Students also self-reported their influenza vaccination status. An individual-based outbreak model was developed using R statistical software. Using these data, combined with appropriate transmission parameters, we simulated an influenza outbreak and assessed the effects of preferentially vaccinating according to the social network analysis data. We ran the simulation 1500 times. Each simulation selected a random student to introduce the virus into the population. For each vaccination strategy, the likelihood of each individual being infected was calculated as a percentage of the number of times they were infected in the 1500 possible outbreaks.Findings215 students (85%) responded. Non-responders were assumed to have reciprocal relationships with responders; therefore it was possible to construct the entire medical student network. We found that the outcomes of vaccination strategies based on between-ness (the extent to which an individual lies between others in the network) and degree (the number of connections an individual has), which are both measures of connectedness, quickly converged. As more individuals were vaccinated, the likelihood of individuals contracting the infection tended to be similar, irrespective of vaccination based on between-ness or degree. After vaccination of an additional 8% of the population (20 students) the outcome of the experimental influenza outbreak was similar for both strategies.InterpretationOur results add to a small pool of evidence supporting targeting vaccination of individuals according to between-ness in an attempt to reduce the spread of influenza. However, in small, densely connected populations, vaccination according to degree might be preferential because of the rapid convergence and the relative ease of locating individuals with a high degree versus those with high between-ness. This study suggests that vaccination strategies that target highly connected individuals within a network might limit spread of infectious disease. Future work could include evaluating current vaccination approaches using social network analysis.FundingUniversity Hospitals of Morecambe Bay NHS Foundation Trust funded data collection.ContributorsRI conceived the programme of research of which this was a part. JH collected the data. RE was responsible for input and analysis of the data, supervised by RI and TK. RE developed the model and BR developed the software. RE wrote the abstract with input from RI and TK. All authors have approved the original version of the abstract for publication.Declaration of interestsWe declare no competing interests.

AB - BackgroundThere is increasing interest in the effects of social networks on disease dynamics. We simulated the spread of influenza through a population of medical students where transmission was related to the social network structure and vaccination status of individuals.MethodsAll students at Lancaster Medical School, Lancaster, UK (n=253) were asked to rate the strength of their relationship with all other students from the medical school. Students also self-reported their influenza vaccination status. An individual-based outbreak model was developed using R statistical software. Using these data, combined with appropriate transmission parameters, we simulated an influenza outbreak and assessed the effects of preferentially vaccinating according to the social network analysis data. We ran the simulation 1500 times. Each simulation selected a random student to introduce the virus into the population. For each vaccination strategy, the likelihood of each individual being infected was calculated as a percentage of the number of times they were infected in the 1500 possible outbreaks.Findings215 students (85%) responded. Non-responders were assumed to have reciprocal relationships with responders; therefore it was possible to construct the entire medical student network. We found that the outcomes of vaccination strategies based on between-ness (the extent to which an individual lies between others in the network) and degree (the number of connections an individual has), which are both measures of connectedness, quickly converged. As more individuals were vaccinated, the likelihood of individuals contracting the infection tended to be similar, irrespective of vaccination based on between-ness or degree. After vaccination of an additional 8% of the population (20 students) the outcome of the experimental influenza outbreak was similar for both strategies.InterpretationOur results add to a small pool of evidence supporting targeting vaccination of individuals according to between-ness in an attempt to reduce the spread of influenza. However, in small, densely connected populations, vaccination according to degree might be preferential because of the rapid convergence and the relative ease of locating individuals with a high degree versus those with high between-ness. This study suggests that vaccination strategies that target highly connected individuals within a network might limit spread of infectious disease. Future work could include evaluating current vaccination approaches using social network analysis.FundingUniversity Hospitals of Morecambe Bay NHS Foundation Trust funded data collection.ContributorsRI conceived the programme of research of which this was a part. JH collected the data. RE was responsible for input and analysis of the data, supervised by RI and TK. RE developed the model and BR developed the software. RE wrote the abstract with input from RI and TK. All authors have approved the original version of the abstract for publication.Declaration of interestsWe declare no competing interests.

U2 - 10.1016/S0140-6736(14)62155-3

DO - 10.1016/S0140-6736(14)62155-3

M3 - Journal article

VL - 384

JO - The Lancet

JF - The Lancet

SN - 0140-6736

IS - Supplement 2

M1 - S29

ER -