WP2-T5 Descriptive epidemiology and definition of causal links in the context of mollusc mortalities using data from the database

Participants :

  • 5 - IRTA
  • 9 - CEFAS

The aim of this task is to provide the basis to establish causal factors for disease manifestation using longitudinal data on the occurrence of pathogens (i.e. prevalence) and mortalities collected in other WPs. This will be achieved by focusing on identifying the link between infection and mortality for each pathogen at each site being studied in distinct countries using a purely descriptive approach related to pattern and distribution. Consequently, the information collected for the database data fields will be considered by three main groups related to the variable factors concerning the host species, pathological agent and the environment of each site (see WP2, Task 4).

Methodologically, a classical approach will be undertaken and preliminary analysis will be related to showing any distribution of disease through the measurement of central tendency and variation. The incidence (timing) of disease will also be taken into account through an intersite comparison, which will describe the status of disease at the selected sites at a particular point in time or over a given period. The number of site surveys will determine the time orientated description.

Since risk factors such as high temperatures or salinity modifications may have causal link to disease or mortality, the environmental or site variables will be used additionally to identify differences between areas or whether local clustering occurs (in space or time). This will be analysed together with pathogen characteristics to determine any exposure differences that might affect disease manifestation, as well as a consideration of specific cyclical (seasonal) disease patterns. Data on other risk factors in addition to the presence of infectious agents is to be collected, it is important that within WP3 methods of data collection and storage are agreed and consistently implemented by all project participants.

Data will be collected in a number of countries by different participants. Firstly, it is important that the responsibility for devising data collection protocols, central collation and storage of these data is clear. Secondly, the statistical analysis of the entire dataset is also crucial (in terms of statistical power to detect associations and identify differences between regions, production systems etc.).

The identification of any potential determinants will be carried out and their possible relationship to disease manifestation (pathogen presence and mortalities) will be investigated using an analytical approach. This means it will be necessary to consider an event, condition or characteristic that plays an essential role in producing disease occurrence. Therefore, Task 5 will need to make statistical inferences related to causes for the diseases originating from the chosen pathogens in populations of three distinct bivalve species.

The data will be site specific (e.g. clustered by bay) and simple initial sample size calculations will be performed to determine how many sites need to be sampled to generate sufficient power to identify odds ratios >2 at P<0.05 for putative explanatory factors (e.g. presence of the agent) whilst controlling for confounding variables, such as associations related to strength, consistency, specificity, etc.

Multivariate statistical modelling will then be used to identify risk factors, and standard statistical packages (e.g. SAS or SPSS, Excel/PopTools) will be used in conjunction with the relevant site data from Task 4 and the descriptive epidemiology from Task 5.