Evolving cohesion metrics of a research network on rare diseases: a longitudinal study over 14 years

Research collaboration is necessary, rewarding, and beneficial. Cohesion between team members is related to their collective efficiency. To assess collaboration processes and their eventual outcomes, agencies need innovative methods - and social network approaches are emerging as a useful analytical tool. We identified the research output and citation data of a network of 61 research groups formally engaged in publishing rare disease research between 2000 and 2013. We drew the collaboration networks for each year and computed the global and local measures throughout the period, relating connectivity with the normalized impact. Although global network measures remained steady over the whole period, the local and subgroup metrics revealed a growing cohesion between the teams. Transitivity and density showed little or no variation throughout the period. In contrast the following points indicated an evolution towards greater network cohesion: the emergence of a giant component (which grew from just 30% to reach 85% of groups); the decreasing number of communities (following a tripling in the average number of members); the growing number of fully connected subgroups; and increasing average strength. Normalized citation measures, although ranking the member groups as above average impact for Spanish biomedical output, do not support correlation between collaboration and citation impact. The Spanish research network on rare diseases has evolved towards a growing cohesion - as revealed by local and subgroup metrics following social network analysis. This greater cohesion does not translate into a greater impact, as measured by the citation analysis of the papers.