Seasonality may be a proxy for some organisms whose incidence is affected by the environment in general and by climate in particular. This project aimed to systematically document the seasonality of pathogen-related infectious diseases reported in England and Wales to explore the possible influence of climate.
Human infectious disease data from Public Health England’s LabBase2 national surveillance database and modelled climate date from the Met Office were utilised, with linkage facilitated by the MEDMI project. The researchers conducted a time series analysis of 277 different pathogens (i.e. the top 75% in terms of total case count). Each organism’s time series was decomposed at weekly, monthly and quarterly periodicities, and forecasted using the Tbats package in R. Seasonality was detected using model fit statistics.
The results showed that whilst the majority of infectious disease organism serotypes displayed a seasonal component, only 36 serotypes displayed a moderate to strong correlation with climatic factors. This group included pathogens (e.g. Campylobacter) that have been known for some time to follow seasonal patterns, as well as other less-studied organisms or seasonal components with less frequently used climatic variables.
A key outcome of the project was a table of 36 pathogen serotypes with the details of their potential links with climate. In addition, an Rshiny app was developed to help with dissemination of results http://markcherrie.shinyapps.io/medmi_app. The user is able to filter the pathogens by seasonality, prevalence and serotype. Once an individual serotype is selected, a range of information is provided visually.