Average weekly temperatures (in degrees Celsius) and monthly sunshine (in hours) for the UK over the same period were sourced from

the UK MET office information.23 The relationship between the weekly incidence of IPD and viral infections was initially analysed by calculating the Pearson and Spearman’s correlation coefficients, for the original and standardized datasets. The data were standardized in order to crudely remove the effect of the concurrent seasonality of the pathogens. For each weekly count, the data were standardized by subtracting the mean and dividing by the standard deviation of the counts for that week over all of the years of the study period (13 years), thus providing a measure of how the incidence for a particular week deviates from the average for that time of year. Three different regression models were investigated (Table 2). Two were ABT-737 mw additive models (a basic linear regression and an identity-linked negative binomial regression) and one was a multiplicative model (a log-linked negative binomial regression).

The negative binomial regression models were applied to account for over-dispersion of the dataset. The dependent variable was the incidence of IPD, with explanatory variables, the incidence of influenza and of RSV. Two additional explanatory variables, the UK mean weekly temperature23 and monthly hours of sunshine, were investigated in the models to adjust for the common seasonality of the pathogens. The models were applied to all ages and then to each age group individually, as well as to a range of lags (0–4 weeks). We estimated www.selleckchem.com/products/dabrafenib-gsk2118436.html the percentage of IPD cases that could be attributable to influenza and RSV. For the additive models, this was estimated by multiplying the virus’ case count with its regression coefficient. This determined the estimated number

of cases of IPD attributable to the virus and from which a percentage could be calculated. For the multiplicative model, the percentages of IPD cases attributable to influenza and RSV were estimated by multiplying the virus’ case count with its C1GALT1 fitted rate ratio (RR), (attributable percentage = case count × (RR-1) × 100).24 All analyses were carried out with STATA version 11.2 (StataCorp. 2009. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP). The common seasonal incidence of all three diseases, IPD, influenza and RSV can be clearly seen in this dataset (Fig. 1). Whilst IPD cases are reported all year round, there are distinct increases during winter months. For influenza, there are similarly timed peaks in reported incidence, but with fewer cases out of season. The same is true for RSV, with very few cases reported in the summer months and with large numbers of cases being reported, mainly in infants (Table 1), in the winter. Fig. 2a also displays the strong shared seasonality of the different pathogens. However, in Fig.