Why are all multi-centre trial sites not the same? A network analytic approach

Method Assessment of a multi-centre trial, incorporating network data on communication and trust within the team, collected at discrete intervals, to gauge ‘team spirit’. Members within each centre will be asked to provide information on communication within the team network (e.g. email, face-to-face, etc.) regarding key aspects of support, organisation and resource sharing, both formal and informal. Data will be assessed by actor oriented longitudinal network analysis techniques implemented in SIENA. This is a modelling framework to assess the outcome variable of recruitment and its association with ‘team spirit’, as measured through network parameters for communication and interaction, whilst controlling for known network structures such as reciprocity (i.e. if A interacts with B, then B interacts with A), hierarchical formal structures, and density. The same model will be applied to each centre and parameter estimates obtained. These will, in turn, form a basis for meta-analysis over all trial sites to factor in centre characteristics (e.g. size), and determine if higher level characteristics have importance. Findings will provide key information for the planning and implementation of future trials.