Abstract
Reconstructing the history of individual transmission events between cases is key to understanding what factors facilitate the spread of an infectious disease. Since conducting extended contact-tracing investigations can be logistically challenging and costly, statistical inference methods have been developed to reconstruct transmission trees from onset dates and genetic sequences. However, these methods are not as effective if the mutation rate of the virus is very slow, or if sequencing data is sparse. We developed the package o2geosocial to combine variables from routinely collected surveillance data with a simple transmission process model. The model reconstructs transmission trees when full genetic sequences are unavailable, or uninformative. Our model incorporates the reported age-group, onset date, location and genotype of infected cases to infer probabilistic transmission trees. The package also includes functions to summarise and visualise the inferred cluster size distribution. The results generated by o2geosocial can highlight regions where importations repeatedly caused large outbreaks, which may indicate a higher regional susceptibility to infections. It can also be used to generate the individual number of secondary transmissions, and show the features associated with individuals involved in high transmission events. The package is available for download from the Comprehensive R Archive Network (CRAN) and GitHub.
Highlights
The identification of transmission trees and transmission events during infectious disease outbreaks can lead to identifying factors and settings associated with subsequent transmissions1–4, describing super-spreading events5,6, or populations and areas more vulnerable to importations and transmission7–10, and quantifying the impact of control measures11,12
Genetic sequencing of pathogens have since become more common, and new tools such as the R package outbreaker2 were created to combine the timing of infection and the genetic sequences in order to improve the accuracy of inferred transmission trees13,14,18–20
The R package o2geosocial is a new tool for data analysis building upon the framework developed in outbreaker2
Summary
The identification of transmission trees and transmission events during infectious disease outbreaks can lead to identifying factors and settings associated with subsequent transmissions, describing super-spreading events, or populations and areas more vulnerable to importations and transmission, and quantifying the impact of control measures. The most straightforward approach to reconstructing who-infected-whom is to carry out patient interviews and establish the previous contacts to connect the reported cases. Statistical methods have been developed to infer transmission trees from routinely collected epidemiological data. The Wallinga-Teunis method was first developed to infer probabilistic transmission trees from onset dates and generation times in a maximum likelihood framework. Genetic sequencing of pathogens have since become more common, and new tools such as the R package outbreaker were created to combine the timing of infection and the genetic sequences in order to improve the accuracy of inferred transmission trees. The accuracy of these reconstruction methods relies on the proportion of sequenced cases, the quality of the sequences, and the characteristics of the pathogen. The measles virus evolves slowly, and sequences from unrelated cases can be very similar, which makes these methods ineffective for measles outbreaks
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
